Science for Cultural Heritage: Technological Innovation and Case Studies in Marine and Land Archaeology in the Adriatic Region and Inland

VII INTERNATIONAL CONFERENCE ON SCIENCE, ARTS AND CULTURE

SCIENCE FOR CULTURAL HERITAGE Technological Innovation and Case Studies in Marine and Land Archaeology in the Adriatic Region and Inland

This page intentionally left blank

www.ecsac.eu

VII INTERNATIONAL CONFERENCE ON SCIENCE, ARTS AND CULTURE

SCIENCE FOR CULTURAL HERITAGE Technological Innovation and Case Studies in Marine and Land Archaeology in the Adriatic Region and Inland

August 28-31, 2007 • Veli Lošinj, Croatia

Editors

M. Montagnari Kokelj M. Budinich University of Trieste, Italy

C. Tuniz ICTP, Italy

World Scientific NEW JERSEY

•

LONDON

•

SINGAPORE

•

BEIJING

•

SHANGHAI

•

HONG KONG

•

TA I P E I

•

CHENNAI

Published by World Scientific Publishing Co. Pte. Ltd. 5 Toh Tuck Link, Singapore 596224 USA office: 27 Warren Street, Suite 401-402, Hackensack, NJ 07601 UK office: 57 Shelton Street, Covent Garden, London WC2H 9HE

British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library.

SCIENCE FOR CULTURAL HERITAGE Technological Innovation and Case Studies in Marine and Land Archaeology in the Adriatic Region and Inland Veli Lošinj, Croatia 28–31 August 2007 eds. C. Tuniz, M. Montagnari Kokelj and M. Budinich Copyright © 2010 by The Abdus Salam International Centre for Theoretical Physics

ISBN-13 978-981-4307-06-2 ISBN-10 981-4307-06-8

Printed in Singapore.

EH - Science for Cultural Heritage.pmd

1

1/4/2010, 2:30 PM

v

INTRODUCTION M. BUDINICH1, M. MONTAGNARI KOKELJ1, C.TUNIZ2 1 University of Trieste and 2ICTP, Trieste – Italy

The Lo!inj series of Conferences (see http://www.ecsac.eu/) focus on interdisciplinary themes, with the aim to promote a forum involving researchers, scholars and students from natural and social sciences and humanities. This interdisciplinary dialogue and the active involvement of local communities and non-specialists, with particular attention to young people, is one of the main objectives of the European Centre for Science Arts and Culture (ECSAC, http://www.ecsac.eu/), the main organizer of the conferences. ECSAC has its headquarters in Mali Lo!inj, Croatia, while the Managing Committee and the Secretariat are based in Trieste, Italy, hosted by the Consortium for the Promotion of Study and Research in Physics, at the ICTP – the Abdus Salam International Centre for Theoretical Physics. ECSAC’s President, Paolo Budinich, at the same time President of the Trieste International Foundation for the Progress and Freedom of Science, has been one of the key figures in the process started in the 1960s to develop the “Trieste System” as an international institutional scheme that promotes advanced knowledge involving also scientists and students from developing countries. ECSAC, ICTP and the Trieste University, promoters of the Lo!inj Conference, belong to the Trieste System, as well as SISSA / ISAS - the International School for Advanced Studies, AREA Science Park and the Consortium mentioned above, who all supported the project. Besides these partners, the institutional representatives of UniAdrion - the Virtual University of the Adriatic-Ionian Basin (a net including both ECSAC and Trieste University) and IAEA - the International Atomic Energy Agency, as well as scientific scholars from many European countries have guaranteed the international dimension of the Lo!inj events. The 7th Lo!inj International Conference on Science, Arts and Culture was titled “Science for Cultural Heritage: Technological Innovation and Case Studies in Marine and Land Archaeology in the Adriatic Region and Inland”. It was inspired by recent studies on the ancient bronze Apoxyomenos, found in 1999 in Veli Lo!inj waters, carried out by the Hrvatski Restauratorski Zavod

vi

(Croatian Institute of Restoration) in collaboration with Opificio delle Pietre Dure, the famous restoration centre based in Florence, Italy.. The Conference was held on 28 - 31 August in Lo!inj and involved 50 scientists, archaeologists and restoration experts mainly from north-eastern Italy, Carinthia, Slovenia and Croatia, but also from other prestigious European centres. The aim of the conference was to discuss the contribution of physics and other sciences in archaeological research and in the preservation of cultural heritage. Considering that the mission of ECSAC is to promote the interaction among the diverse cultures of the peoples from the lands on the Adriatic and Ionian seas, the major themes were related to the history and pre-history of this region, from greek-roman archaeology on the eastern Adriatic coasts to the palaeoanthropology of the Neanderthals of the Vindija caves in Croatia, from the Roman city of Aquileia to the pleistocenic cave of Homo heidelbergensis in the Karst of Visogliano (Trieste), from the Roman ship Julia Felix of the Grado lagoon to the ancient bronze Apoxyomenos of the Veli Lo!inj waters. A variety of scientific disciplines provide tools and methods that are crucial to reconstruct humanity’s past and to preserve material remains that witness the evolution of human culture. Geology reconstructs the history of terrestrial environments, critical for the evolution and dispersal of humans. Chemistry explains reactions that modify materials left by human activities, including the destructive effects of pollution. Biology has a critical role in archaeology, particularly with the recent progress in the analysis of DNA in ancient organic materials. Physics has a special role in archaeology and cultural heritage, providing a variety of non-invasive analytical methods that can characterise ancient materials. These methods include new microscopes based on synchrotron radiation, high-energy ions, neutrons, lasers and other radiations or particles that can detect in-situ the structure and composition of art objects and archaeological remains, with resolution at molecular and atomic level. High resolution satellite imagery allows the prospection of archaeological sites over distances of several kilometres. Also the time dimension is critical in archaeology. Physicists have developed clocks based on natural radioactivity that provide chronologic information on time scales from centuries to million years, from the Middle Ages to the Pleistocene. Finally, the analysis of certain isotopes such as nitrogen, calcium and strontium provide information that can be used to reconstruct diets, diseases and migrations of ancient human populations. The conference has offered the opportunity to promote the debate among scholars and students from natural sciences, humanities and social sciences,

vii

considering the impact of the most recent scientific and technologic developments in applications to art and archaeology.

Acknowledgements We would like to extend our warmest thanks to the Mayor of Mali Lo!inj Gari Cappelli who always helped and sustained the organization of the series of conferences on Science and Culture (see http://www.ecsac.eu/) in the beatiful island of Lo!inj. This year in particular we would like to thank the Town Council of Veli Losinj who offered the newly refurbished theater, now Kulturni Dom, of Veli Losinj. This conference wouldn't have been possible without the unbelievable dedication, kindness and patience of Morena Petrich who was responsible for the secretariat and for all the organizative matters.

This page intentionally left blank

ix

CONTENTS Introduction Program

v xiii

Archaeological and Archaeometric Data in the Study of the Athlete of Croatia M. Michelucci

1

Ion Beam Techniques for Analysis of Cultural Heritage Objects: Collaboration between the Ruđer Bošković Institute and the Croatian Conservation Institute S. Fazinić, I. Božičević, Z. Pastuović, M. Jakšić, D. Mudronja, K. Kusijanović, M. Braun and V. Desnica

15

Study by Mobile Non-Destructive Testing of the Bronze Statue of the “Satiro” of Marsala G. Guida, D. Artioli, S. Ridolifi and G. E. Gigante

23

Archaeometric Measurements with PIXE in Slovenia Ž. Šmit In Situ Chemical Composition Analysis of Cultural Heritage Objects Using Portable X-Ray Fluorescence Spectrometry D. Wegrzynek, E. Chinea-Cano, A. Markowicz, S. Bamford, G. Buzanich , P. Wobrauschek, Ch. Streli, M. Griesser, K. Uhlir and A. Mendoza-Cuevas Integrated Geophysical Techniques for the High-Resolution Study of Archaeological Sites M. Pipan and E. Forte Thermoluminescence Dating and Cultural Heritage M. Martini and E. Sibilia

31

41

55

69

x

New X-Ray Digital Radiography and Computed Tomography for Cultural Heritage F. Casali, M. Bettuzzi, R. Brancaccio and M. P. Morigi Cosmic Rays for Archaeology G. Giannini

85

100

Some Examples of Examination, Characterisation, Analysis & Conservation Techniques Dedicated to Archaeological Artefacts J. L. Boutaine

108

Presentation of DEMGOL: Online Etymological Dictionary of Greek Mythology E. Pellizer

125

Building Up an Archaeological Restoration & Conservation Department in Friuli-Venezia Giulia F. Lo Schiavo

131

Relative Sea Level Changes by Using Archaeological Markers: The INTERREG Italia-Slovenia Project “Alto Adriatico” S. Furlani, F. Antonioli and R. Auriemma

139

Digitization and Multispectral Analysis of Artistic Objects: Exemplary Cases and Web Documentation G. Maino and S. Massari

164

Actuopalaeontology: A Polyfunctional Tool for Archaeology G. Bressan, G. Fonda, S. Kaleb, R. Melis, M. E. Montenegro, P. Mourguiart, N. Pugliese, R. Riccamboni, A. Russo, N. Sodini and G. Tromba

172

Robotics Tools for Underwater Archaeology G. Conte, S. Zanoli, D. Scaradozzi and L. Gambella

187

Accelerators and Radiation for Art and Archaeology C. Tuniz

195

The 14C Contribution to the Protohistory of Friuli (North-Eastern Italy) P. Càssola Guida

211

xi

Serpentinite Shaft-Holed Axes in the Caput Adriae: Preliminary Results and Perspectives Based on X-Ray Computerized Microtomography F. Bernardini, E. M. Kokelj, N. Sodini, D. Dreossi, S. Favretto, G. Demarchi, A. Alberti and F. Princivalle

226

Mummies – A Special Report Results of CAT Scan Analyses of Egyptian Mummies in the Civico Museo di Storia ed Arte of Trieste M. V. Torlo

238

ANGLE Software for Semiconductor Detector Gamma-Efficiency Calculations and Possibilities for Its Applications to Cultural Heritage Objects Characterization S. Jovanovic and A. Dlabac Hominid Fossils as Universal and National Cultural Heritage: An Essay on Past and Present Attitudes Towards the Ownership of Hominid Fossils and the Question of Repatriation P. V. Tobias

247

258

This page intentionally left blank

xiii

PROGRAM OF THE CONFERENCE International Conference "Science for Cultural Heritage" Veli Lo!inj, Croatia, August 28-31, 2007

Tuesday, AUGUST 28, 2007 14.00 – 16.00 16.00 – 16.25

Registration Welcome and Greetings of the Authorities Chairperson: C. Tuniz (ICTP) P. Budinich, ECSAC M. Mu!i" , President of Council of Mali Lo#inj T. Morin, Local Community G.C. Ghirardi, Consortium for Physics M. Montagnari Kokelj, University of Trieste

The state of archaeometry in the Caput Adriae regions Chairperson: M. Montagnari Kokelj (University of Trieste) 16.25 – 16.50

16.50 – 17.15

17.15 – 17.40 17.40 – 18.05

M. Michelucci, former Director Archaeological Section Opificio delle Pietre Dure, Florence Archaeological and archaeometrical data in the study of the Athlete of Croatia S. Fazini", Ru$er Bo#kovi" Institute, Zagreb Ion Beam Techniques for Analysis of Cultural Heritage Objects: Collaboration between the Ru!er Bo"kovi# Institute and the Croatian Conservation Institute G. Gigante, University La Sapienza, Roma New archaeometric approaches to study large bronze statues %. &mit, University of Ljubljana Archaeometric measurements with PIXE in Slovenia Welcome Cocktail in Punta Hotel

xiv

Wednesday, AUGUST 29, 2007 Scientific methods and techniques: new perspectives Chairperson: C. Tuniz (ICTP) 9.00 – 9.25 9.25 – 9.50 9.50 – 10.15 10.15 – 10.40 11.15 – 11.40

D. Wegrzynek, IAEA In situ chemical composition analysis of cultural heritage objects using portable X-ray fluorescence spectrometry M. Pipan, University of Trieste Integrated geophysical techniques for the high-resolution study of archaelogical sites M. Martini, AIAr Luminescence dating techniques for the cultural heritage F. Casali, University of Bologna New X-ray Digital Radiography and Computed Tomography for Cultural Heritage G. Giannini, INFN Trieste Cosmic Rays for Archaelogy

12.00 – 16.00

Archeological boat tour to the Athlete of Croatia finding site

16.25 – 16.50

J.-L. Boutaine, C2RMF, Paris Some exemples of examination, characterisation, analysis & conservation techniques dedicated to archaelogical artefactcs N. Sodini, Sincrotrone Trieste – University of Firenze Characterization of archeological wood by means of X-ray computed micro-tomography A. Mereu and R. Riccamboni, University of Trieste, DiSGAM Pen-based Technology in geosites field surveying: mapping, drafting and data collecting using a Tablet PC

16.50 – 17.15 17.15 – 17.40

xv

Archaeometry and society Chairperson: M. Michelucci (University of Pisa) 17.40 – 18.05 18.30 – 18.55

18.55 – 19.20

E. Pellizer, University of Trieste The other “Great Code” – Heritage of ancient Greek and Roman culture in an on-line, multilingual, dictionary F. Lo Schiavo, Superintendent Cultural Heritage Regione Friuli Venezia Giulia Building up an archaeological restoration & conservation department in Friuli-Venezia Giulia R. Costa, University of Trieste, Facoltà di Architettura Unesco Chair Trieste The master plans for Aquileia and non-invasive surveys

Thursday, AUGUST 30, 2007

Marine Archaelogy: the sea and its relics Chairperson: M. Budinich (University of Trieste) 9.00 – 9.25

9.25 – 9.50 9.50 – 10.15

10.15 – 10.40

S. Furlani et alii, University of Trieste Relative Sea Level Changes by using archaeological markers: the INTERREG Italia-Slovenia Project “Alto Adriatico” S. Massari, ENEA Digitization and multispectral analysis of artistic objects: exemplary cases and web documentation G. Bressan, S. Favretto, G. Fonda, S. Kaleb, R. Melis, M.E. Montenegro, N. Pugliese, R. Riccamboni, A. Russo, N. Sodini, G. Tromba, F. Vita (ATA Group) Actuopalaeontology: a polyfunctional Tool for Archaeology G. Conte, Polytechnic University of Marche Robotics tools for underwater archaelogy

xvi

Land Archaelogy: Prehistoric sites and materials Chairpersons: G. Boschian (University of Pisa) and A. Coppa (University La Sapienza, Roma) 11.00 – 11.25 11.25 – 11.50

11.50 – 12.15

G. Boschian, University of Pisa Homo heidelbergensis environment and behaviour: the Visogliano site (Trieste) A. Coppa, University La Sapienza, Roma A new classification approach: Neural networks analysis by using the Self-Organizing Maps (SOMs) applied to human fossil dental morphology C. Tuniz, ICTP Dating human dispersal and impact during the Pleistocene

Land Archaelogy: Ancient humans Chairpersons: G. Boschian (University of Pisa) and A. Coppa (University La Sapienza, Roma) 16.00 – 16.25 16.25 – 16.50 16.50 – 17.15

17.15 – 18.40

18.40 – 19.05

P. Càssola Guida, University of Udine The 14C contribution to the protohistorical reconstruction in north-eastern Italy F. Bernardini et alii, University of Trieste Greenstone artefacts in prehistory: preliminary results and perspectives based on X-ray computerized microtomography M. Vidulli, Civic Museums, Trieste Mummies – a special report results of CAT scan analyses of egyptian mummies in the 'Civico Museo di Storia ed Arte' of Trieste (italian: Tomografia assiale computerizzata delle tre mummie egizie dei Civici Musei di Storia ed Arte di Trieste) S. Jovanovic, University of Montenegro Possibilities of ANGLE software for semiconductor detector gamma-efficiency characterization in neutron activation analysis of cultural heritage objects E. Cacciatore, Centro Regionale per la Progettazione e il Restauro della Regione Siciliana, Palermo Villa romana del Casale of Piazza Armerina. Restoration and Management

xvii

Friday, AUGUST 31, 2007 Round Table: Archaeometry: from research to high formation Chairperson: M. Martini (AIAr) 9.00

F. Lo Schiavo, Superintendent Cultural Heritage Regione FVG G. Gigante, University La Sapienza, Roma M. Montagnari Kokelj, University of Trieste M. Tenconi, University of Padova C. Tuniz, ICTP Trieste F. Bradamante, University of Trieste J.-L. Boutaine, C2RMF, Paris G.C. Ghirardi, University of Trieste A. Coppa, University La Sapienza, Roma G. Giannini, University of Trieste R. Costa, University of Trieste Students of Uniadrion Universities

This page intentionally left blank

1

ARCHAEOLOGICAL AND ARCHAEOMETRIC DATA IN THE STUDY OF THE ATHLETE OF CROATIA MAURIZIO MICHELUCCI

A few days after the bronze known as the Athlete of Croatia emerged from Lo!inj waters (fig. 1), the Hrvatstki Restauratorski Zavod (HRZ, the Croatian restoration institute) took the lead of the restoration project, involving in it also the Archaeological Division of the ‘Opificio delle Pietre Dure’ (OPD, a special institute of the Italian Ministry for Cultural Heritage).

Figure 1

OPD performed the preliminary inspections and gamma-graphic analysis in the training pool of Lo!inj Maritime Police (fig. 2) where the bronze was initially placed. These preliminary surveys (fig. 3) confirmed the good preservation of the artefact under the thick layer of organogenic concretions1 (fig. 4). Successively the long and difficult restoration process was performed in ! Photos kindly granted from journal “Archeologia Viva”, OPD, HRZ and by M. Michelucci.

2

Figure 2

Figure 3

1

See A. A LDROVANDI, S. PORCINAI, Indagine gammagrafica, in “Apoxyomenos - L’Atleta della Croazia”, edited by M. Michelucci, Firenze 2006, pgs. 110-112.

3

Figure 4

the well equipped HRZ laboratories in Zagreb by Giuliano Tordi – who in the meantime retired from OPD – and Antonio "erbeti#, from HRZ. During the entire restoration process all operations were preceded and accompanied by a considerable number of chemical, physical and metallurgical analysis on samples of the statue performed in OPD scientific laboratories. Analysis focused on alteration and corrosion products as well as on the metallic alloy, especially in those areas which were more likely to be of interest for the casting techniques and the successive refinements processes underwent by the statue before its loss, as it clearly appeared since the beginning of the restoration process2. OPD sent some of the organic remains found in the interior of the statue to the Beta Analytic Laboratories of Miami (Florida) for 14C analysis. The results of this work are known: the big and beautiful bronze statue entirely freed from marine concretions was presented to the general public in 2

OPD - HRZ collaboration was constant and fruitful along the whole, demanding, restoration process, thanks to good relations, not only on the scientific plan, but especially of reciprocal esteem and friendship, between the superintendants heading OPD – G. Bonsanti and successively C. Acidini – and the Direction of the restoration works, i.e. F. Meder, HRZ Director, and M. Domijan, Chief Conserver of the Croatian Ministry of Culture.

4

two big expositions in Zagreb and Florence – the latter in the prominent site of Palazzo Medici Riccardi (fig. 5) – together with a richly illustrated catalogue. The Florence exposition also concisely presented results of the scientific analysis.

Figure 5

This is one of the not-so-frequent cases in which the humanistic science of archaeology and the so-called ‘hard’ sciences happily and tightly joined their efforts in the critical interpretation of the found data and the subsequent analysis. There are some uncertainties still to be resolved by further experimentation and research, but this joint effort lead to common results on the statue meaning and dating, provenience, final destination and loss circumstances – data of paramount importance for the history of art, metallurgy and restoration. I will remember here some of them: the statue is surely a late republican Roman copy (someone prefer to say replica) of a Greek original. This dating results from a series of data critically considered. Alloy analysis on a series of duly mapped samples (fig. 6) didn’t give univocal data: as one can easily see from the attached table (fig. 7), those performed with the SEM/EDS method in non corroded areas indicated a percentage of lead between 1.6% and 3.3%, while those performed with the ICP/AES method on the same samples were

5

definitely higher (16.2% to 20%). As it should always be the case, chemical analysis data need an adequate interpretation to assess their validity, especially

Figure 6

Figure 7

6

when applied to cultural heritage elements. As it is well known, the ICP/AES method cannot discriminate among areas more or less affected by corrosion, as the results refer to the average value measured in the entire sample. This means that the higher solubility – and dispersion – of copper corrosion by-products in marine waters tend to indicate higher concentrations of lead than copper, altering the data of the original alloy composition. The SEM/EDS analysis is more selective and allows more precise investigations in areas not touched by corrosion processes; it consequently produces more reliable data on the original alloy composition. This is confirmed by the fact that SEM/EDS analysis on highly corroded samples – and thus affected from dispersion of copper alteration by-products – gave results similar to those of the ICP/AES method, indicating high percentages of lead3. It would be interesting to verify if previously published alloy composition data of bronze statues rescued from sea bed have allowed for these factors and methodological considerations. It is known that, since the middle of the Republican Era and for the entire Imperial Age, the alloy of big and small bronze statues in Rome was added with a considerable amount of lead in order to lower both fusion temperature and costs even if that implied a consistent degradation of the artefact quality, making it brittler. This was not true in archaic and classic Greek production nor in the first Hellenistic age, a technical fact that has considerable consequences on dating. We saw that the alloy composition of the Athlete of Croatia contains a low percentage of lead; if this scientific finding had been uncritically associated to the artistic style of the work, we would have been brought to consider the artefact as a Greek original of about 360 BC. However, like in a police investigation, other archaeological, analytical and technical factors have been taken into consideration, and their union brought to a much later dating of the bronze fusion. The data obtained from the 14C analysis of three organical finds discovered in the interior of the statue indicate chronological intervals going from the end of the 2nd century BC to the second half of the 2nd century AD. The latter datum comes from a burnt wooden splinter coming from the right leg – providing an important contribution to the dating of the shipwreck – while the former (110 BC to 70 AD) refers to a peach stone bitten by a small rodent, abundant remains of whose burrow have been found4.

3

4

See C. G. LALLI, G. LANTERNA, D. PINNA, S. PORCINAI, M.RIZZI, I. TOSINI, Indagini diagnostiche, In “Apoxyomenos...”, pgs.117-118 See Appendici scientifiche, ibidem, pgs. 121-123.

7

The fact that mice could have made their burrow – the lining has also been found, and was made of an herb that palaeobotanical analysis revealed to be typical of a ruin-like environment – within the statue indicates that the statue was abandoned, full of gaps and leaning on the ground. And it could not have been otherwise since a vast portion of the back of the right thigh – which sustains the whole weight of the statue – had a big gap due to a casting fault (fig. 8) making it impossible for the statue to stand.

Figure 8

The whole casting didn’t turn out well: this is testified not only by the above mentioned conspicuous gap on the supporting leg, but also by the large number of patches – some of which irregularly trapezoidal in shape and huge in size – applied on gas pockets, gaps, and craters widespread on the whole body surface (fig. 9).

Figure 9

The good composition of the bronze alloy heavily clashes, and with full evidence, with the nasty yield of the casting. All these data seem to have little in

8

common, but they assume particular relevance if combined with other general archaeological and historical information: 1. in Greek sculpture, good bronze alloys, preserving the classic age composition, are rarely assessed even in the early imperial age (some bronze portraits – certainly of imperial age – have been cast with this alloy, almost without lead); 2. the usage of big, convex patches of irregular shape to correct casting faults in bronze statues spreads in the 1st century BC. 3. it is assessed that in the late 1st century BC, continental Greece experiences a generalized economic and social crisis as a consequence of Sulla’s plunders, repeated civil wars and lack of investments; on the other hand the Micrasiatic area undergoes economic expansion – its big towns, former capitals of the Hellenistic reigns, would have become, with a growing urban development, the metropolis of the middle and late Roman empire; 4. the only bronze replica – almost of the same size of the Lo!inj statue – comes from a gymnasium of the late 1st century AD in Ephesus, Anatolia (fig. 10).

Figure 10

9

If we consider all these data and refer them to our statue, we get a general picture that brings to the following conclusions: the casting of the statue was poor because they wanted to use a good alloy which required a blast furnace reaching very high temperatures in an era when this happened rarely and only for smaller statues – like portraits, where the limited volume eased the casting operations: their technique, means, costs and maybe also their knowledge were and proved to be inadequate to succeed in the ambitious project of making a replica of an entire, big statue of the 4th century BC using the same bronze alloy of the original. The results was a statue full of faults which could not even stand because the supporting leg – the right one – could not carry all its weight due to a big casting gap on the thigh. At the end of the assembling and welding process of head and limbs, the statue, possibly rejected by the client, had to be laid down on the store floor of the foundry. There, after a while, a family of mice must have made its burrow within the statue itself, already partially freed by its casting earth. Mice are responsible for the indentations on the peach stone 14Cdated from 110 BC to 70 AD – more probably around the average dating of 20 BC. Clearly, the casting of our statue must have occurred earlier – but not too earlier – in a lapse of time going indicatively around the middle of the 1st century BC, in the Roman late Republic Era. As for the foundry, its original cultural background must have been undoubtedly Greek, but it should be geographically linked to one of the big cities of Asia Minor, in consideration of the historical and economical arguments mentioned above5. The presence of wooden splinters and carbon remains in the right leg and armpit should be attributed to one or more restoration works, which the poor 14C dating fix around 100 AD. Maybe, after many years of abandonment, the statue had found – or was about finding – a buyer in the northern Adriatic Sea and they tried to make it stand and give it a certain stability, consolidating the better they could (at the time) the elements that more needed it. Through the imperfect casting gaps they put a pin inside the right armpit and a wooden prop in the right leg in order to reinforce and support the welding of large plugs in two areas where the gaps and the concentration of casting faults did not grant enough stability. The traces of burning and even of carbonisation found on these wooden elements are the result of the welding of those patches, which should have had supporting as well as esthetical functions. From the examination of some samples, it has been possible to assess that the larger patches were made of a bronze alloy almost identical to that of the body of the statue (fig. 7). The fact that most of them have been lost – leaving in view the embeddings on the

5

See M. MICHELUCCI, Prima del naufragio, in “Apoxyomenos...”, pgs. 60-61.

10

surface of the statue where they were situated – can be explained by the poor staying power of the soft solder used to applying them. Very likely the statue was shipped from Asia Minor to the northern part of the Adriatic Sea after these restoration and make-up processes. The Adriatic was the sea Romans feared most, as it can be seen in some excerpts from the poems of Catullus and Horatius, and indeed, slightly before reaching its final destination, the ship carrying our beautiful, but heavy bronze statue, run into a storm near the isle of Lo!inj and had to throw it overboard to lighten and try to escape from the wreck (fig. 11). This expedient was rather common, as it can be deduced from sources describing similar cases. No traces have been found of the ship and we cannot know if their expedient had the desired result or if the ship wrecked some miles away, under the strength of the East-North-East winds. The Apoxyomenos, however, had to wait almost 2000 year to see the light again and

Figure 11

be admired on its pedestal and undergo a new – and much more effortdemanding – restoration work (fig. 12).

11

Figure 12

Is everything clear, then? Some important uncertainties still remain, where archaeometric studies cannot help. The main doubt is what the Athlete is actually doing: everybody agrees that the statue shows a winner, with the strigil in his right hand after the match, but as the strigil did not survive, since the finding of the statue of Vienna, in Ephesus in 1896, archaeologists are split in supporting two incompatible hypothesis on the represented action: some believe that the athlete is wiping the oil and sweat off his left wrist with the strigil – hence he would be an Apoxyomenos, i.e., literally “the scraping one” – while others believe that he is cleaning the inside of the strigil with his thumb after having cleaned the entire body6. Some technical tests performed in Zagreb putting a modern copy of the strigil in the right hand of the statue (fig. 13) would need further analysis, as they were made with only one standard strigil, and not with several models of different shape and size – whose existence is attested around the middle of the 4th century BC. It also seems – at least from the images that have been published 7 – that the strigil they put in the right hand was not turned in all possible positions.

6

7

Even the two archaeologists that studied accurately the statue for the Florence exhibit have different opinion on this subject: see N. CAMBI, L’Atleta che pulisce lo strigile and V. SALADINO, L’Atleta con lo strigile, in “Apoxyomenos...”, pgs. 21-51. See “Apoxyomenos...”, fig. 27.

12

Figure 13

Up to this moment, however, these tests did not provide irrefutable results, maybe also because of the slightest movements of the arms due to the fall or – like in the case of the statue of Vienna – to a restoration highly admired at the time, but today to be considered as inadequate, if not even damaging to the artefact. However, some useful elements for solving the problem come from historical sources and... logic. It is evident that this statue must have been rather famous at the time it was casted – in the 4th century BC – and in all ancient times if at least seven big and small replicas of different materials survived till our days. While ancient authors indicate two other statues of athletes wiping themselves with the strigil, i.e. Apoxyomenoi, made by Daedalus from Sicyon (still in Asia Minor!) and Polyklitus the Younger, prior to the more famous statue of Lysippus there are no indications of one or more models of statue showing athletes cleaning their strigil. Even if the ex silentio argumentation is not conclusive, considering the incompleteness and defectiveness of the ancient sources we possess, this fact is nevertheless very meaningful. Another

13

consideration could be added: if our athlete cleaned the strigil with his left hand, it would have been left-handed, a less probable hypothesis. The problem is not definitely solved but strong evidence make us believe that our Athlete – as well as his twin of Vienna – is indeed an apoxyomenos, even if we will never know if it comes from an archetype of Daedalus from Sicyon or Polykleitos the Younger 8 . It was for this reason that, in spite of some authoritative opinions saying the contrary, the statue of the athlete recovered from the sea of Lo!inj got this title in his presentation to the public in the Zagreb and Florence exhibits, beautiful again after a long and hard restoration9 (fig. 14).

8 9

See M. MICHELUCCI, in “Apoxyomenos...”, p. 19. In both exhibits the exposing criteria established by Branko Sila$in were absolutely innovative and – in my opinion – brilliant: the bronze statue – immerged in a pure white and almost blinding surrounding, with no shades at all – appeared as timeless and spaceless. Breaking the traditional concepts of illumination, using a strong, diffused light coming from indefinable surfaces, the many casting imperfections and the corroded areas of the surface almost disappeared, making pop up the exceptional image as a whole. Appreciated by the public, this exposing criterion has been often (but not always!) criticised by archaeologists, professionally interested in examining the technical and execution details and the casting faults more than the statue as a whole.

14

Figure 14

15

ION BEAM TECHNIQUES FOR ANALYSIS OF CULTURAL HERITAGE OBJECTS: COLLABORATION BETWEEN THE RU!ER BO"KOVI# INSTITUTE AND THE CROATIAN CONSERVATION INSTITUTE* STJEPKO FAZINI!†, IVA BO"I#EVI!, "ELJKO PASTUOVI!, MILKO JAK$I! Division of Experimental Physics, Ru!er Bo"kovi# Institute, Bijeni$ka c. 54, Zagreb, Croatia DOMAGOJ MUDRONJA, KATARINA KUSIJANOVI!, MARIO BRAUN Croatian Conservation Institute, Gr"kovi#eva 23, Zagreb, Croatia VLADAN DESNICA Laboratory for Science and Technology in Art, Department for Conservation and Restoration, Academy of Fine Arts, Ilica 45, 10000 Zagreb, Croatia Collaboration between the Croatian Conservation Institute (CCI) and the Laboratory for Ion Beam Interactions of the Ru%er Bo&kovi' Institute (RBI) in the field of analysis of cultural heritage objects is reviewed. This collaboration is based on applications of ion beam analytical methods for characterization of inorganic pigments, alloys and other materials in paintings, statues and other objects of cultural value that are under restoration/conservation process by the CCI specialists. Elemental composition of samples is determined by using ion beam analysis techniques such as Particle Induced XRay Emission (PIXE) and Rutherford Backscattering Spectrometry (RBS). The work is performed at the ion microprobe and other end-stations installed with the RBI Tandem accelerator facility. The methodology of work will be shortly presented and illustrated by one example of recently performed work.

1. Introduction The main analytical laboratory in Croatia dedicated to the analysis of cultural heritage and art objects operates within the Croatian Conservation Institute (CCI). Scientific analysis of artistic and cultural heritage objects is organized by the CCI through their Natural Science Laboratory with the main purpose to *

†

The work presented has been partially supported by the EU FP6 RBI-AF and IAEA CRO13050 projects. Corresponding author, e-mail:[email protected].

16

enable their better restoration and/or conservation. Chemical analysis of art objects may be an essential step in estimating their authenticity, origin and age; and for selecting appropriate restoration or conservation protocol. Due to a high historic and/or artistic value of such objects, one of the most important criteria for selection of analysis techniques is its capability for non-destructive investigation. The non-destructive character in this case means that the analysis process does not alter nor modifies in any way the investigated area. For objects or samples too large or too delicate and for artifacts on which sampling is not possible/allowed one has to apply a further criterion – that of non-invasiveness. Under the non-invasive term it is understood that any sampling or dismounting of the artifact is avoided and more generally that its integrity and its environment are preserved. Furthermore, since artifacts can often exhibit extremely different material composition and structure, analytical methods allowing sensitive and multielemental characterization are highly required. The CCI natural Science Laboratory has been equipped with microscopes enabling them to perform various microscopy techniques based on investigation of objects by using visible, infrared or ultraviolet light. In addition, CCI is equipped with the X-ray imaging system and with modern portable X-ray Fluorescence Spectrometer (XRF) used for in-situ chemical analysis of surfaces. Access to ion beam micro analytical techniques is provided to CCI through already long and successful collaboration with the Laboratory of Ion Beam Interactions (LIBI) of the Rudjer Bo&kovi' Institute (RBI), where elemental composition is analyzed by using ion beam analysis (IBA) techniques, such as Particle Induced X-ray Emission (PIXE) and Rutherford Back-scattering (RBS). One of the first successful collaborations between the LIBI staff and the CCI was between 1985-86, when the RBI participated in a project named “Secret paintings of Josip Ra(i' and Miroslav Kraljevi' – analysis by physical and chemical methods”. The project goal was to perform scientific analysis of paintings done by two important Croatian painters that were active at the period from the end of the nineteenth to the beginning of the twentieth century, to identify the methods and materials used by artists, to compare them, and to present results to the general public. It included analysis by X-rays, infra-red and UV light. RBI analyzed elemental constituents of pigments on 28 paintings. Analysis of pigments was very important part of the overall analysis. The project ended up with an exhibition (that presented all the scientific work) which was held at the Modern Gallery in Zagreb, between 6th March and 6th April 1986. Since then, a continuous collaboration between LIBI and CCI exists, where LIBI provides supplementary analysis of cultural heritage objects by using ion

17

beam analytical techniques. Short overview of the analytical methods used by LIBI will be shortly presented and illustrated by one example of recently performed work. 2. Ion beam analysis at RBI Through the Laboratory for Ion Beam Interactions, the Division of Experimental Physics of the Rudjer Boskovic Institute operates and maintains the Tandem Accelerators facility that physically consists of two electrostatic accelerators, associated beam lines and measurement end-stations. The facility is used for research and applications by various clients/collaborators in a range of fields, including nuclear and atomic physics, applications in materials science and development of advanced materials, archaeology and characterisation of cultural heritage objects, etc. The basis of CCI and LIBI collaboration is in the use of ion beam analysis techniques, such as Particle Induced X-ray Emission (PIXE) and Rutherford Back-scattering (RBS) for analysis of elemental composition of samples (objects). Nuclear reaction products

Charge pulse

! " rays

Recoil nuclei

Ion beam

Transmitted particles

X-rays

Forward scattered particles

Backscattered particles Secondary electrons

TARGET Light

Figure 1. Interaction of ion beams with target materials.

When energetic ion beam (of several MeV energy) hits a target (see Figure 1), individual ions penetrate through the target material in their incident direction, gradually losing energy until they stop at certain depth, which is typically at the order of 0.1 mm (depending basically on the ion initial energy and the target material). One of the most probable processes which takes place along the path of incoming ion (proton) is its scattering with electrons. As a result of this scattering, atoms are ionized, with electrons being ejected from atoms. The ionized atom tends to return to its original state, by filling created vacancy with electrons from the outer shells. The excess energy will be given to an emitted photon (X-ray emission) or an electron (Auger electron emission).

18

The energy of the emitted X-ray depends on the atom type and allows elemental characterization when x-rays are detected with appropriate detector. This is the basis of PIXE spectroscopy1. PIXE is multi-elemental non-destructive detection method capable to measure concentrations of elements from Na to U with the sensitivity down to the ppm scale. Another processes occur in the interaction of the primary ion beam with the material in the target (see Figure 1), resulting with the emission of gamma rays, secondary electrons, nuclear reaction products, or for example backscattered primary ions. Detection of these backscattered primary ions form the basis of the Rutherford Backscattering Spectrometry (RBS)2, which is useful method for determination of elemental concentration depth profiles in thin films, and is very often performed simultaneously with PIXE. If gamma ray detector is available, then one can at the same time employ the so called Particle Induced Gamma ray Spectrometry (PIGE) which is particularly useful for detection of some light elements3. Three end-stations have been used in the analysis of cultural heritage objects. Two of these end-stations are designed for small samples that has to be analyzed in vacuum. The first one is the general purpose IBA end-station (Figure 2a). This end station is equipped with the vacuum chamber and detectors for PIXE, PIGE and RBS measurements. Samples have to be inserted in the vacuum chamber, and are limited in size to several cm in diameter. The ion beam has dimensions between 1 and 5 mm, and this determines the size of the area to be analyzed.

Figure 2. Two end-stations for small samples for analysis in vacuum. The left side (a) shows the general purpose ion beam analysis end-station, while the ion microprpbe end-station is shown on the right (b).

The second end-station is the ion microprobe (Figure 2b), which is an instrument designed to focus ion beams to micrometer dimensions. The nuclear

19

microprobe facility at RBI has been in operation since 1991 and since then significant improvements in ion beam focusing, beam intensity, and detection systems have been made4. The recent upgrade, in which quadrupole dublet has been extended to quintuplet configuration, enabled focusing of protons and heavier ions to less then one )m. Focused beam can be scanned over the sample surface, covering analysis areas of up to about 1x1 mm. Samples for analysis have to be inserted in the vacuum chamber which is seen in the centre of the Figure 2b. Inside the chamber detectors for x-rays, scattered, recoiled and transmitted particles are located. As x-ray detector a 10 mm2 silicon drift detector (SDD) is used, which can, contrary to the conventional Si(Li) detectors, work with high count rates (more than 10000 cts/s) and does not need liquid nitrogen for cooling. Samples size can range from micrometer dimensions, up to 1-2 cm. The data acquisition system SPECTOR developed at the Institute5 enables recording of all signals from detectors together with the ion beam position in the moment of a particular event. This allows display of two dimensional images of information given by a particular detector. The third end-station is dedicated for measurements on samples which cannot be exposed to vacuum and cannot be sampled, i.e. for non-invasive and non-destructive analysis of objects. This is so called in-air end station, where ion beam is extracted through thin foil in the air, and an object for analysis is exposed to the ion beam in air. It is equipped with the sample stand that allows fine XYZ sample translation and positioning, and with Si(Li) x-ray detector for PIXE measurements.

3. Example: Analysis of pigments at the ion microprobe Non-destructive IBA methods can be applied for analysis of various materials, which are of interest to conservators during their work: paintings and pigments, illuminated parchments, metals and alloys, gold and jewelry, ceramics and glass, bones (posthumous remains) and objects made of bones, transcripts, wooden objects etc. As the main application of this method is directed to the characterization of materials, i.e. qualitative and quantitative determination of elemental composition, the results may yield diverse information about the state of a certain sample/object, its elemental constituents, help to clear the provenance question and enlighten the phases of production, and finally help conservators to choose the best methods and materials for work with the artifacts. The following example is given to illustrate the use of the ion microprobe for analysis of pigments.

20

The same conventional samples of paint layers cross sections embedded in polyester resin as prepared for optical microscopy investigations can be used for the analysis using the focused proton beam from the ion microprobe, with the objective to identify paint layers. During the last decade we have analyzed hundreds of such samples in relation to different projects6,7. Figure 5 shows typical samples with several paint layers as seen by optical microscope.

Figure 5. Typical samples as prepared for Optical Microscopy are suitable for nuclear microprobe investigation. The areas showed cover 0.4x0.6 mm. Minimal size of a sample needed is several )m.

Figure 6. elemental maps of Hg, S (+Pb+Hg), Fe, Si, Pb, Cu, Ca and Al in one of the microsamples taken during restoration from one of the paintings of the polyptych originating from the church ''Gospa od $unja'' at the island of Lopud near Dubrovnik.

Figure 6 shows elemental maps of Hg, S (+Pb+Hg), Fe, Si, Pb, Cu, Ca and Al in one of the microsamples taken from one of the three restored paintings, parts of a composition originating from the church ''Gospa od $unja'' at the island of Lopud near Dubrovnik. The whole composition is attributed to the

21

painter Matej Jun(i', based on known manuscript from 1452, where it was written that four Lopud inhabitants ordered and three months later paid to master Matej Jun(i' preparation of an altar with 12 paintings8. Figure 7 shows three investigated paintings.

Figure 7. Three investigated paintings.

Twenty one microsamples have been taken and analyzed together with insitu surface analysis done by using portable XRF. As a result, a number of pigments have been identified and information about restoration activities in the past were obtained. The following pigments were found in analysed samples: white: - lead white (2PbCO3.Pb(OH)2) - gipsum (CaSO4.2H2O) - barium white (BaSO4) – 19th century red:

- cinnabar (HgS) - red ochre (Fe2O3 x nH2O) - bolus (Al2O3 x SiO2 +Fe2O3) - organic red - alizarin (Al(OH)2) - minium (Pb3O4)

- azurit (2CuCO3 x Cu(OH)2) - (CaCuSi4O10) - ultramarin (2Na2Al2Si2O6 x NaS2) brown and yellow: - Iron oxide (brown and yellow ochre) black - organic black (C) metals: - gold (Au) blue:

22

- (Cu+Zn) – in compound characteristic for 19th century. Two dimensional maps of elemental composition confirmed that paintings were over painted with several layers, and that the last layer was done in 19th century, as can be concluded from the fact that some pigments and materials discovered at the beginning of 19th century were used. High concentration of calcium and sulfur proved the usage of gypsum like filling in the base, where the presence of aluminum and silicon indicated addition of some silicates, while the high content of lead shows extensive use of lead white in combination with other pigments like azurite. References 1. S. A. E. Johansson and J. L. Campbell, PIXE: A novel technique for elemental analysis, John Wiley & Sons, New York (1988). 2. W.K. Chu, J.W. Mayer, M.A. Nicolet, Backscattering spectrometry, Academic Press, New York (1978). 3. A. Savidou, X. Aslanoglou, T. Paradellis, M. Pilakouta, Nucl. Instr. and Meth. B152, 12 (1999). 4. M. Jak&i', I.B. Radovi', M. Bogovac, V. Desnica, S. Fazini', M. Karlu&i', Z. Meduni', H. Muto, ". Pastuovi', Z. Siketi', N. Skukan, T. Tadi', Nucl. Instr. and Meth. B261, 541 (2007). 5. M. Bogovac, I. Bogdanovi', S. Fazini', M. Jak&i', L. Kukec, W. Wilhelm, Nucl. Instr. and Meth. B89, 219 (1994). 6. S. Fazini', ". Pastuovi', M. Jak&i', M. Braun, D, Krsti', D. Mudronja, Utilization of Accelerators, Proceedings of an International Conference, IAEA Proceedings Series, STI/PUB/1251, IAEA, Vienna, (2006). 7. V. Desnica, K. $kari', D. Jembrih-Simbuerger, S. Fazini', M. Jak&i', D. Mudronja, M. Pavli(i', I. Perani', M. Schreiner, Appl. Phys. A92, 19 (2008). 8. J. Belamari' et al., The Gotic Century on the Adriatic Painting in the perspective of Paolo Veneziano and his followers, Gallery Klovi'evi dvori, Zagreb, Croatia (2004).

23

STUDY BY MOBILE NON DESTRUCTIVE TESTING OF THE BRONZE STATUE OF THE “SATIRO” OF MARSALA. GIUSEPPE GUIDA, DOMENICO ARTIOLI Istituto Superiore per la Conservazione, MiBAC STEFANO RIDOLIFI and GIOVANNI E. GIGANTE Dipartimento di Energetica, Sapienza Università di Roma

The bronze statue called the “il Satiro danzante” was found underwater in the Sicily channel in the 1997. The restoration at the Istituto Centrale per il Restauro was very long and careful, assisted by diagnostics procedures in order to help the conservative choice in the consolidation of the artefact and to know the age and provenance of the statue. The results on the alloys composition and structure are reported in this paper.

Introduction The study by means of destructive and non-destructive methods of investigation of the bronze statue of the “Satiro” had been possible because it was in restoration for a long period of time in the l’Istituto Centrale per il Restauro [1]. A complete diagnostic examination of a work under restoration is a practice becoming common, with some difficulties, in the conservative restoration of very relevant works [2]. The relative high cost of diagnostics limits, instead, the use in all cases, with remarkable risks and limitations overall for the aspects concerning the monitoring and maintenance of restored work. An extended program of investigations hasn’t, in fact, the only finality of conservation, but also of a better knowledge of restored work under the point of views of its material consistency and of a more precise historical identification [2]. In the case of a work, as the Satiro, accidentally recovered in the Sicily channel (outside by an historical context to which refer, then with a provenance and more uncertain historical context) the exams oriented toward the knowledge can have a greater importance then in other cases. This is the reason because in this paper the aspects of knowledge and identification of the work under study will be discussed in more details. Among the different approaches to the study of ancient metals the use of in situ non invasive techniques is rather recent thank to the growing potentialities

24

of mobile systems of non destructive testing and to the possibility to perform metallographic exams in situ; this kind of exam allow to study the production technology and to identify the techniques used in the casting a big bronze statue [4]. The non destructive approach has been privileged, but it is not to be neglected to verify always the results with destructive tests, making microsapling in hinder parts of the artifact. In the case of the Satiro the main open arcaeometric problems are: a) the age of the artefacts, b) its provenance, c) the casting techniques and the used surface treatments. 1. Measuring techniques and experimental apparatus A short description of experimental methods and apparatus is given in the next paragraphs. 1.1. In Field EDXRF spectrometer The Field Portable Energy Dispersive X-Ray Fluorescence (FP-EDXRF) technique is based on excitation of samples with X-rays and measurements of the energy of the secondary X-rays emitted by the samples themselves. The energy of secondary (also known as characteristic) X-rays depends on the chemical elements present in the sample being examined while the intensity of the energy is proportional to the abundance of the element under scrutiny. These surveys may be carried out prior or during the restoration work; however, as the methodology is totally non-invasive, it may be applied for purely informative purposes, regardless of intervention. The penetration of X-rays varies from a mere few (as in the case of gold) to several hundred microns (as in the case of light-weight matrix elements, for example those containing relevant amounts of organic compounds). The EDXRF examination is capable of detecting the composition of a metal alloy, in fact the high atomic numbers and the density of metal alloy facilitate the production of fluorescent X-rays of enough energy to be detected, even using a low intensity sources. The fluorescent lines emitted by all elements compounding the alloy appear within the spectrum, whereas low atomic-number elements are absent or minoritary in the matrix. A typical EDXRF-system is composed of three parts: a) an X-ray tube; b) an X-ray detector with electronics; c) an acquisition system with a multichannel analyser. The X-ray tube works at 30 kV and 0,1 mA. It is a light (air cooled) tube (less than 2 Kg of weight). The detector used a Silicon Drift Detector ( SDD) detector having a energy resolution of 139 eV at 6,4 keV. The detector is cooled with a Peltier build in circuit [5].

25

1.2. Metallography Metallography is the technique of preparing a metal surface for analysis by grinding, polishing, and etching to reveal microstructual constituents. After preparation, the sample can easily be analyzed using optical or electron microscopy. A skilled technician is able to identify alloys and predict material properties, as well as processing conditions and corrosion process due to the exposition to the different environments [6]. Metallographic specimens were "mounted" using a hot compression thermosetting epoxy resin. Mounting a specimen provides a safe and ergonomic way to hold a sample during the grinding and polishing operations. After mounting, the specimen is wet grounded to reveal the surface of the metal. The specimen is successively grounded with finer and finer grades of silicon carbide paper to remove damage from sectioning and then from each grinding step. After grinding the specimen was polished with a slurry of alumina, silica, or diamond on a napless cloth to produce a scratch-free mirror finish, free from smear, drag, or pull-outs and with minimal deformation remaining from the preparation process. After polishing, certain microstructural constituents can be seen with the microscope, e.g., inclusions and nitrides. Finally in order to reveal crystal structure (apart the non-cubic ones) were used suitable chemical or electrolytic etchant. 2. Results and discussion The starting point in the discussion on the obtained results is that, as is very common in the classic roman period, the statue is built soldering together different pieces. A careful examination during the restoration allow us to establish that a) head, b) chest, c) two legs, d) two arms are joint together by welding. The strong corrosion of the surface forbids the non destructive analysis of the alloys directly on the artefact after a scratching of the patina, that is a common practice on ancient bronzes that do not showing yet a thick degree of corrosion, as it is in the case of Satiro. It was chosen then to do few samples, however doing a mapping of the artefact surface with the aim to put in evidence superficial degradation phenomena of the alloy. 2.1. Alloys composition In table I are shown the results obtained on the samples, obviously done on the different pieces compounding the Satiro, in figure 3 are shown the withdrawals points. Taking only into account the results to be refer to the six pieces the first observation is that the alloys shown a similar composition, all featuring a high

26

lead concentration (14-21%) which, as will be confirmed following by the metallography, undergo a globular segregation. The head is constituted by high tin (11%) concentration alloy, in comparison with the other parts (4-6%), probably an intentional choice to obtain a better fluidity of the alloy. After all the hair alloy look similar, characterised only by lower lead concentration probably a liquefaction due to different behaviour during the melting and solidification phases. The first hair sample belongs to a piece melted apart. The lower iron concentration, the only marker allowing us to assess the purity of the row materials used, may suggest a greater care in the material selection, compared to that done for the other parts. The high artistic quality of the head, requiring a grater care, could therefore find verification. Table 1. Results of the EDXRF analysis on withdraw samples of the alloy from the Satiro. Position of the withdrawals head (rear inner folding in the right)

Cu

Sn

Pb

Fe

68,5

11,2

20,4

0,5

hair (inner side of the head)

76,6

9,8

13,6

0,4

hair (big piece detached)

73,8

9,3

16,9

0,4

left thigh (near the edge)

73,9

4,4

21,0

0,7

right thigh (external side )

70,4

6,4

21,7

1,4

left arm (inner side)

70,7

12,9

15,4

1,0

left arm (inner side)

72,2

12,7

14,1

1,0

right arm (external side)

69,1

8,7

21,1

1,1

right arm (inner side)

72,7

5,2

20,1

2,1

left leg (external side)

73,3

9,6

16,3

0,7

left arm (on the soldering)

49,5

7,4

42,5

0,6

right arm (on the soldering)

69,1

5,5

24,5

0,9

The sample taken from the welding show an increase of lead concentration, very evident in one of the two samples, pointing out the use of a copper alloy with a higher lead concentration. The results of the metallographic study of the samples taken by the head (figure 1), the left arm and the left thigh clearly showed a crystalline microstructure constitute by a dendritic array tending to a polygonal shape with the evidence of sliding planes. All the samples show a much corroded surface; this corrosion goes deep into being of an inter-dendritic type. In the case of the head it is in the interval 300-500 micron and for the left arm reach the 800 micron. Lead rounded globes of variable size and inclusions of different kind are always present. Table II shows the results of the analysis of the alloys obtained on the three metallographic samples; they are quite different from those on the withdraw samples. The superficial treatment of this sample could alter the results, for ex-

27

ample for the presence of lead globe. The obtained result confirms however that the alloys of the Satiro have a high lead concentration and a tin one in the normal range for an artistic bronze.

Figure 1. Images of metallographic sample taken from the head, right 100 !, left 200!. Table II . Results of EDXRF analysis on the three metallographic samples. Withdrawal position for metallography

Cu

Sn

Pb

Fe

Head (withdrawal from the nape)

80,7

8,2

10,9

0,2

Left arm

70,5

9,4

20,2

0,2

Left leg

71

8,5

20,5

0,2

Figure 2 . Left an example of corrosion phenomenon (green patina), left an example of sea origin concretion.

Figure 3. Position of some measuring points on the left leg of the Satiro.

28

2.2. Results of the superficial mapping with a EDXRF spectrometer With the aim to characterise the different corrosion phenomena on the Satiro surface and, maybe, identify the hexogen material adhering to it, (figure 2), a systematic surface mapping was carried out. The points were chosen using the visible alterations and their colour with the help of the restorers. In the three following figures (figure 3-5) the position of 26 points is shown and in table III there is a short description of their visual aspect.

Figure 4. Position of some measuring point on the chest, head and back of the Satiro There are several and useful results of the mapping that are not possible to include in this discussion. It was possible to identify particular alteration processes, such as that of point 18 (figure 6) in which the presence of iron and manganese allow to attribute the black colour to a digenesis phenomenon with minerals of the sandy floor.

29

Figure 5. Position of some measuring points on the hip and thigh of the Satiro. 3500

7000 Counts/channel

3000 2000 1000

Sn K!

Counts/channel

4000

Fe K !

Pb L"

2000

5000

Pb L" Pb L!

Pb L!

2500

6000

Cu K!

Mn K!

8000

Cu K !

Mn K!

3000

0 0

2

4

6

8 10 12 14 16 18 20 22 24 26 28 30 Energy (keV)

1500

Fe K !

1000

Sn K!

Sr K !

Cu K" Ni K !

V K!

Pb M

Sn K"

Zn K !

Fe K"

Ca K !

500

0 0

2

4

6

8

10

12

14

16

18

20

22

24

26

28

30

Energy (keV)

Figure 6. Two spectra of the left leg, point 18 colour dusty dark-black,

References 1. You can see a video in WWW pages at: http://www.youtube.com/ watch?v=aKiwY1pvEsU

30

2. G.E. Gigante, S. Ridolfi , G. Visco , G. Guida, “Appraisal of the new approach to the archaeometric study of ancient metal artifacts by the use of movable EDXRF equipments”, proceedings of della “International Conference on Araechaeometallurgy in Europe, II 293-302 ISBN 88-85298-50-8 (2003). 3. R.Cesareo, G. E. Gigante, A.Castellano, J.S. Iwanckyk, “Portable Systems for Energy Dispersive X-Ray Fluorescence Analysis”, Encyclopaedia of Analytical Chemistry, R. A. Meyers Ed, ed. John Wiley & sons, 1332713338 (2000). 4. R. Cesareo, G.E. Gigante, P. Canegallo, A.Castellano, J.S. Iwanczyk and A. Dabrowski Nuclear Instruments and Methods in Physics Research B. 380, 440-445, (1996). 5. G. E. Gigante., R. Cesareo, Radiation Physics and Chemistry, 51(4), 689700, (1998). 6. G.Giardino, G.E. Gigante, G. Guida and R. Mazzeo, Sciences of Conservation & Archaeology, Shanghais, 10, 58-64 (1998). Table III – Description of measuring points in the mapping on the Satiro surface N°

Visual identification

Short description of obtained results

1 2 3 5 6 7 8 9 10

Near-white clear dusty green Green Near-white clear dusty green Nearly unpatinated iron grey Clear dusty green Blue-grey Grey-brown (clean spot) Clear brown Thick emerald green

11 12 13 14

Thick average green Dark glossy green Blue-grey (below calcareous concretions) Olive-green

Lead remarkable superficial enrichment Lead remarkable superficial enrichment Lead and zinc superficial enrichment Lead remarkable superficial enrichment Lead, iron and tin superficial enrichment Lead remarkable superficial enrichment Lead remarkable superficial enrichment, iron and vanadium Iron and vanadium remarkable superficial enrichment Lead, superficial enrichment, strontium, calcium, potassium and vanadium Lead, superficial enrichment, iron, manganese and vanadium Lead superficial enrichment Lead superficial enrichment

16 17 18

Inner point Dark-brown Dusty dark-black

19 20 21 22 23

Remedial plug Basin welding Cleaned area Black point on the chin Grey-brown point on thecheek Basin welding on the neck Outer spot of the welding Eye

24 25 26

Lead superficial enrichment, calcium, manganese, iron and vanadium Lead, zinc, tin remarkable superficial enrichment Calcium, iron and strontium Lead remarkable superficial enrichment, calcium, potassium, manganese, iron, zinc and strontium Calcium, potassium, iron and strontium Lead remarkable superficial enrichment Lead superficial enrichment, vanadium Lead remarkable superficial enrichment, iron enrichment Lead and iron remarkable superficial enrichment, calcium, vanadium, strontium Lead and iron remarkable superficial enrichment, zinc Lead and iron remarkable superficial enrichment, zinc Calcium, iron, strontium

31

ARCHAEOMETRIC MEASUREMENTS WITH PIXE IN SLOVENIA !. "MIT University of Ljubljana, Faculty of Mathematics and Physics, Jadranska 19, SI-1000 Ljubljana, Slovenia, and Jo!ef Stefan Institute, Jamova 39, POB 3000, SI-1001 Ljubljana, Slovenia The in-air proton beam of the tandem accelerator of the Jo#ef Stefan Institute is used for systematic investigation of the archaeological artifacts and objects of art. The review guides through the results obtained by the PIXE and PIGE methods on the investigation of medieval glass, metal objects of the Roman and medieval period, numismatics and painting pigments. The differential PIXE method for the evaluation of concentration profiles is briefly introduced.

1. IBA methods Ion beam analytical provides efficient methods for the analysis of cultural heritage objects. Irradiation with a particle beam produces nearly negligible irradiation damage, and the techniques based on the particle beam in the air allow simple handling of the objects irrespective of their size. Simultaneous measurements with different types of detectors enable a large range of elements that can be analyzed simultaneously: the method of proton induced X-rays (PIXE) can detect elements from about aluminum to uranium; the characteristic K-shell X-ray lines are used for the elements up to about tin, and the L-shell Xray lines for heavier elements. The limitation at the lightest elements is due to the X-ray absorption in the air and the inactive parts of the detector. Using helium flush in the interacting region and a thin-window X-ray detector, the light elements up to sodium can easily be detected, reaching oxygen and carbon with special experimental set-ups. However, the excited X-rays in this case originate from the very surface layer about one micrometer thick, which could easily be altered by corrosion and other chemical effect. It is then more convenient to analyze the light elements by nuclear reaction analysis; in order to overcome the Coulomb barrier of the nucleus, these have to be performed by projectiles of higher energies that penetrate deeper into the target. The method of proton-induced gamma-ray emission (PIGE) is based on the detection of

32

gamma rays excited by inelastic proton scattering or in resonant reactions via the compound nucleus. The elements between sodium and silicon emit gamma rays of energies between a few 100 keV and 2 MeV, which can efficiently be measured by intrinsic germanium detectors. The lightest elements emit more energetic gamma rays of a few MeV energy; the scintillation detectors provide a better counting efficiency in this energy range. Though all ion beam methods are limited to the sample surface (which requires special care for determining the bulk concentrations), they provide unique possibilities for studies of the surface structures of the objects. Rutherford backscattering analysis (RBS) can reveal the surface layers of heavy elements, while the light elements (notably hydrogen) can be determined by elastic knock-out (or recoil) by heavier projectiles (ERDA); both methods can be performed by an external particle beam in helium atmosphere. Depth profiling can also be made by resonant nuclear reactions. The deepest concentration profiles can be reached by differential measurement, i.e. by varying the projectile energy or the incidence angle, thus reaching different depths of the target. First archaeometric measurements at the Jo#ef Stefan Institute were performed occasionally, using an old Van the Graaff accelerator and a vacuum measuring chamber; they include an analysis of a series of Celtic coins [1] and studies of usewear layers on Mesolithic stone tools [2]; microbeam techniques were also used, in laboratories abroad [3]. In 1997, a new 2 MV Tandetron accelerator became operative at the Institute. Archaeometry was among the new applications, but since then in collaboration with the National Museum of Slovenia. The joint research is funded in various forms by the Slovenian Ministry of Science. Informal collaboration also involves other institutions, including the National Gallery of Slovenia, Slovenian Academy of Sciences and the Faculty of Arts. The experimental techniques involve mainly the PIXE-PIGE method in air, but also the micro beam was used; presently the RBS/ERDA techniques are used in helium atmosphere. The materials studied are glass, metals and painting pigments. The research politics is to perform measurements that would help answering some historic questions; though sporadic service measurements asked by the museum conservators are also carried out. 2.

Analysis of glass and ceramic materials

Excavations in Ljubljana provided a large amount of glass, characterized typologically as glass in the Venetian manner (à façon de Venise) and dated to the 15th and 16th century [4]. As historic records point to several glassworks operating in Ljubljana in that period, it was certainly challenging to study the glass chemical composition and distinguish the possible domestic production from the Venetian import. As the museum preferred measurements to be done

33

without sampling, the analysis with ion beams was the only choice. The measurements involved about 370 glass fragments. Initially we planned to analyze the glass only PIXE [5]. The statistical treatment of a limited number of chemical elements clearly showed that the glasses of Ljubljana can be divided into two groups; different groups were obtained for the control groups of lateRoman glass and the potassium-based forest glass. However, as the contents of sodium, magnesium and aluminum are quite important for the glass characterization, we have further analyzed glass by PIGE for these light elements [6]. The results confirmed the classification of glasses from Ljubljana into two groups. At the same time we also analyzed a few examples of glass colored plates, undoubtedly imported, which appeared markedly different from the glass of Ljubljana. Further interpretation of our data was done in collaboration with prof. Koen Janssens of the University of Antwerp, who with his group analyzed a large amount of façon de Venise glass from Antwerp, Northern Europe and Italy. Comparing all these analyses together it was possible to show that the glasses of Ljubljana split into two groups due to two different types of the flux [7]. The same types of flux were identified in the original white glass (vitrum blanchum) from Venice and also among the Antwerp glass. The more precious glass, cristallo, was not produced in Ljubljana, as its production was monopolized, but it occurred in Antwerp, possibly due to the weaker political influence of Venice there. We subsequently analyzed a series of glasses from Celje, another Slovenian town, and identified the same types of the flux. This finding indicates that the glassworks in the 16th century Slovenia used the same type of flux as the Venetian glassworks, and that they probably imported it from Venice directly. The glassworks imported either flux and melted the glass themselves, or they imported raw glass as semi product. The answer to this question may be sought from the elements characteristic for the siliceous component of the glass, as the domestic glassworks probably exploited the local sources. The respective analytical method should have sensitivity in the ng/g region, which is inaccessible by ion beam methods. We have therefore performed measurements on a small series of glasses from Ljubljana and Venice, using the laser-ablation mass spectrometry (LA ICP MS) at the University of Warsaw; the glasses had to be sampled. Among the inspected features, the Zr-Hf correlations and the contents of rare-earth elements appeared insignificant for the two groups of glasses. Among the Coryell-type classification, which relies on different chemical properties of rare-earth elements, we found a promising grouping inspecting the ratio of Nd/Dy as a function of Zr [8]. A small group containing no glasses from Venice suggests a local source of silica.

34

We analyzed a limited series of other types of glass, the early industrial glass from Slovenia and Art Nouveau glass from the museum funds. The current investigation involves glass from the archaeological site Gradi$%e above Ba$elj near Kranj; it had two population phases dated to the Late Antiquity and the Carolingian period in the 9th century. A large fraction of the glass fragments can be dated by the archaeological method; they coincide with the transition period when the flux made of plant-ash replaced the soda flux common in Roman glassmaking. The measurements showed only two small beads made with the pant ash; the rest being soda-type glass. This indicates glass of the Roman tradition prevailed in the area well into the 9th century. The analysis of ceramic materials includes the cream-colored earthenware produced in Slovenia in the early 19th century. The current work involves analysis of objects and fragments from the collections of the National Museum of Slovenia; terrain work with identification of the clay sources is also planned. The analysis of ceramic objects imposes two problems: the objects are covered by a glaze, so the bulk material can only be reached in spots where the objects were broken or excessively worn out. The clay material is rather pure, and the characteristic trace elements cannot be measured by the PIXE method. The classification, pointing to particular known workshops was attained according to the relative composition of light elements in the clay; their concentrations were determined by the combined PIXE-PIGE method. 3. Metals Analysis of metal alloys is one of the simplest PIXE applications. The calculation of concentrations can rely on setting the sum of all concentrations to unity, which is feasible if all elements in the object are detected. This condition is fulfilled in archaeological alloys that do not contain light metals, such as aluminum and beryllium. Since our early work [1] we developed a computer code that is based on a fast integration algorithm and incorporates secondary fluorescence effects [9]. Systematic studies of metals started with analysis of the Roman military equipment. The investigated objects included fragments of sword scabbards, daggers, belt plates, strap ends, plaques and medallions. Analysis of a gladius from the river Ljubljanica dated to the 1st c. BC showed brass scabbard fittings [10]. This finding stimulated dr. Janka Isteni% to investigate further the early spread of brass in the Roman world, specifically in the south-eastern Alpine region. Her idea was to analyze a series of well dated brooches and detect the appearance of brass [11]. The brooches were of the types Palmettenfibeln, Schüsselfibeln, Nauheim, Almgren 65, Alesia, Jezerine I and II. Brass started to

35

appear in the Almgren 65 and became prevalent at the Alesia type brooches; bronze was again partly applied for the Jezerine brooches, probably due to the production in local workshops with a limited access to brass. This places beginning of the brass use to around 60 BC [12]. This is earlier than minting of brass coins; the first issues appeared in 46/45 BC, though a common use of brass for dupondii and sestertii was introduced in the reign of Augustus. Brass was an attractive material for the military equipment due to its golden appearance. The Romans produced brass by the cementation technique, melting together finely divided copper, carbon and zinc ore; this procedure was required, as the metal zinc would evaporate before the copper melts. The highest zinc concentrations obtained by this procedure were about 28%; the highest values we detected were about 22%, though most of the objects exhibited zinc concentrations of about 15-20% (Fig. 1). Objects of smaller zinc concentrations were also found, which indicates that the brass objects were recycled, occasionally with bronze. The corrosion processes can remove zinc from the surface; for reliable measurements it was necessary to gently polish the investigated area in a size of a few mm2. 0.20 h=0.5% n=29

0.15 den sity 0.10 0.05 0.00

0

5

10

15

Zn (%)

20

25

Figure 1. Statistical distribution of zinc in the Alesia-type brooches.

Among the other interesting finds from the river Ljubljanica we analyzed silver object from the Hoard of Vrhnika [13] (they were made of high-grade silver gilded in the sunken parts of the relief) and a medallion with the portrait of Augustus [14]. The medallion was made of a cheap lead-tin alloy with a low melting point, but it was silvered at the front side and fixed to the substrate by soldering. The solders were similar as used today, made of the tin-lead alloy. From the iconography, the medallion can be connected with the Augustus’ victory over Parts [14]. Upon requests of the museum conservators we analyzed a few daggers ornamented with silver and brass wire and niello. Identification of these materials was necessary for selection of the cleaning procedure. A heavy clay pot was discovered in the Roman site of Drnovo in 2003. Non-destructive X-ray and neutron radiographic investigations revealed that the pot contained a hidden treasure composed of coins of several objects. After the

36

pot was open, two massive silver brooches (dated to the 3rd c. AD) were investigated by PIXE [15]. The surface composition showed that silver (7580%) was diluted with brass. Two grey spot were observed in the inner part of the bow. The analysis of these points showed a mixture of silver and white bronze, i.e. an alloy of copper, lead and a high amount of tin. This indicated that the core of the brooches was made of bronze which was plated by a thick silver layer. Corrosion layers are absent on the objects made of precious metals, which then represent ideal targets for PIXE analysis. However, gold objects of historical importance are rare and the transportation to the lab imposes certain responsibility on museum curators. We analyzed a Roman gold brooch that was found on the Alpine pass of Kranjski Rak [16]; it is one of the pair, the other being kept in Vienna. The analysis showed that the brooch was made of a rather pure gold, including the needle (now broken), which means that the brooch was made for votive purposes rather than for wear. One may conjecture that the needle made of a mechanically more resistant gold alloy would not break off. The other series of objects was from the Carolingian period – it involved the inventory from the grave 355 at the castle of Ptuj. The grave contained several silver and gold objects, the latter being a pair of earrings and a ring. The composition of the earrings was found completely different from that of the ring; it was tracked that one earring had been repaired in the past, replacing the hook with a new one of a less fine gold. The ring, originally made for a male finger, showed an inhomogeneous mixture of gold and silver. This is characteristic of electrum, and points to manufacture in the local workshops, probably in Moravia according to the stylistic analysis [16]. Non-destructive properties of IBA methods make them an attractive tool for investigations in numismatics, but the analysis involves the surface region of the coins only, about 10 µm thick. We realized this limitation during our early study of Celtic coins [1] where we made a cross scan of the coins cut to half; the coin cores were made of a less noble copper-silver alloy, probably result of the melting and cooling process of the alloy. If cuts of the coins are available, PIXE can be used to determine the bulk composition in combination with other methods, like EDS in electron microscope [17]. In spite of the limitations to the surface analysis, an interesting study was made on the silver coins struck in Slovenian mints in the period of 12th-14th centuries [18]. The coins were made of thin silver sheet, so the concentration gradients are small. The coins were mostly primary strikes and their composition reflects that of the ore. Gold and bismuth appeared as the discriminating elements and we were able to characterize the coins as being gold-type or bismuth-type. A large fraction of bismuth-type coins was struck in the mints of Carinthia and in the south-eastern part of modern

37

Slovenia, which bordered to the Hungarian kingdom at that period. As silver mines were predominantly concentrated in Carinthia, this indicates a commercial flow of the Carinthian silver to the eastern mints and further to the Hungarian kingdom. These relations were stopped by the incursions of Mongolians after 1240, bringing the Slovenian mints do decline. 4. Pigments Proton beam in the air can successfully be applied for analysis of the paintings, since their dimensions usually do not fit to any measuring chamber. The measurements were performed by different standards of accuracy, as sometimes the detection of a few characteristic lines is sufficient for the pigment identification. For the high precision measurements we controlled the size of the air gap between the exit window and painting, as the argon signal from the air was used for normalization. The identification of the pigments was done systematically for the works of the Master HGG (Hans Georg Geiger) on occasion of the exhibition in the National Gallery [19]. The pigments identified were minium, red ochre and vermilion (cinnabar) for red, azurite for blue, massicot and yellow ochre for yellow, different copper-based greens (malachite, verdigris, copper resin), green earth, smalt and umbra black. All pigments were amply mixed with lead white. One painting contained traces of copper in every pigment, probably used as a desiccative. The analysis pointed out that some paintings ascribed to the Master HGG were very likely not his original works. Pigments were also identified in the textiles [20], obtaining additional information on composition of the fibers (vegetal or protein). A pilot study was made about the blue pigment used on the medieval manuscripts. Systematic measurements, extending during several recent years, were performed on the works of Slovenian impressionists. Evaluation of the results is still under way and was published only preliminary [21]. The impressionists used a variety of new pigments, which became available from the rapid development of chemistry at the turn of the 19th-20th centuries. For example, the lead white was gradually replaced by other white pigments, such as for example zinc white. 5. Differential measurements The analyzed target depth depends on the projectile energy. Varying the impact energy we can reach different depths of the target, and by proper numerical procedure we can evaluate the concentration profiles of certain elements. As the ionization cross sections vary rapidly with proton energy, the strongest

38

contribution to the X-ray yields is from the very target surface, making the contributions from the inner parts of the target minute. The numerical procedures are then rather sensitive to small variations of the input data, and special procedures are required to stabilize them. We developed a stable algorithm replacing the matrix inversion by the least squares procedure [22]. The method is based on slicing the target into layers of constant elemental concentrations. The division points are selected according to the mean production depths of X-rays in the target. Plated metal targets represent the simplest application of the method. Some examples are shown in [9, 23], but a systematic study of tinned, silvered and gilded objects from the Roman and Early Medieval period is presented in [24]. The thickness of the plated layers was measured for the tinned and gilded objects, while the plated silver was too thick for penetration of 2.5 MeV protons. The gilding technique was identified as fire gilding according to the detected profiles of mercury. Paint layers can be profiled as well, though the task is more complicated due to the contents of light elements that are invisible by X-rays. The information on light elements has to be supplied independently; normally we specify the chemical compound to which the metal ions are bound. A specific example are frescoes, as their matrix of limestone can be monitored through calcium X-rays. We analyzed a test fresco prepared of known mineral pigments and a series of historic fresco fragments. These measurements also enabled us to study two different normalization procedures, based on setting the sum of all concentrations to unity and by measuring the incident proton numbers. For profiling of the oil paintings, the proton number has to be measured, as the pigments are embedded in an organic matrix. We studied the concentration profiles of the impressionist paintings, as it is known that their techniques applied thick pigment layers, often supported by pigment extenders [21]. The deduced profiles showed abundant mixing of the pigments with lead and zinc white. We were also able to resolve the pigments of the signature of Rihard Jakopi% from the background [25]. References 1. 2.

!. "mit, P. Kos, Elemental analysis of Celtic coins, Nucl. Instr. and Meth. B 3, 416-418 (1984). !. "mit, S. Petru, G. Grime, T. Vidmar, M. Budnar, B. Zorko, M. Ravnikar, Usewear-induced deposition on prehistoric flint tools, Nucl. Instr. and Meth. B 140, 209-216 (1998).

39

3. 4. 5.

6. 7. 8. 9. 10. 11. 12.

13. 14. 15. 16. 17.

!. "mit, G. Grime, S. Petru, I. Rajta, Microdistribution and composition of usewear polish on prehistoric stone tools, Nucl. Instr. and Meth. B 150, 565-570 (1999). M. Kos, M. !vanut, Glass Factories in Ljubljana in the 16th Century and their Products, National Museum, Ljubljana 1994; M. Kos, 15th and 16th Century Glass, National Museum of Slovenia, Ljubljana 2007. !. "mit, P. Pelicon, G. Vidmar, B. Zorko, M. Budnar, G. Demortier, B. Gratuze, S. "turm, M. Ne%emer, P. Kump, M. Kos, Analysis of medieval glass by X-ray spectrometric methods, Nucl. Instr. and Meth. B 161-163, 718-723 (2000). !. "mit, P. Pelicon, M. Holc, M. Kos, PIXE/PIGE characterization of medieval glass, Nucl. Instr. and Meth. B 189, 344-349 (2002). !. "mit, K. Janssens, O. Schalm and M. Kos, Spread of façon-de-Venise glassmaking through central and western Europe, Nucl. Instr. and Meth. B 213, 717-722 (2004). !. "mit, K. Janssens, E. Bulska, B. Wagner, M. Kos, I. Lazar, Trace element fingerprinting of façon-de-Venise glass, Nucl. Instr. and Meth. B 239, 94-99 (2005). !. "mit, P. Pelicon, J. Sim%i%, J. Isteni%, Metal analysis with PIXE : the case Roman military equipment, Nucl. Instr. and Meth. B 239, 27-34 (2005). !. "mit, P. Pelicon, Analysis of copper-alloy fitments on a Roman gladius from the river Ljubljanica. Arheol. vestn. 51, 183-187 (2000). !. "mit, J. Isteni%, V. Gerdun, Z. Mili&, A. Mladenovi%, Archaeometric analysis of Alesia group brooches from sites in Slovenia. Arheol. vestn. 56, 213-233 (2005). J. Isteni%, !. "mit, The beginning of the use of brass in Europe with particular reference to the southeastern Alpine region. In: S. La Niece, D. Hook, P.T. Craddock (Eds.), Metals and mines: studies in archaeometallurgy: selected papers from the conference Metallurgy: A Touchstone for Cross-cultural Interaction held at the British Museum 28 30 April 2005 to celebrate the career of Paul Craddock during his 40 years at the British Museum, British Museum, London 2007, p. 140-147. J. Isteni%, The early Roman "Hoard of Vrhnika": a collection of finds from the river Ljubljanica. Arheol. vestn. 54, 281-298 (2003). J. Isteni%, Kleine Mitteilungen: A uniface medallion with a portrait of Augustus from the River Ljubljanica (Slovenia). Germania (Mainz) 81, 273-276 (2003). !. "mit, Analysis of a pair of silver fibulae. In: A. Mi$kec, M. Pflaum (Eds.), Buried Treasure / The Coin Hoard from Drnovo, National Museum of Slovenia, Ljubljana 2007, p. 76-79 !. "mit, M. Budnar, P. Pelicon, B. Zorko, T. Knific, J. Isteni%, N. Trampu#Orel, G. Demortier, Analyses of gold artifacts from Slovenia, Nucl. Instr. and Meth. B 161-163, 753-757 (2000). N. Civici, S. Gjongecaj, F. Stamati, T. Dilo, E. Pavlidou, E.K. Polychroniades, !. "mit, Compositional study of IIIrd century BC silver

40

18. 19. 20. 21.

22. 23. 24. 25.

coins from Kreshpan hoard (Albania) using EDXRF spectrometry, Nucl. Instr. and Meth. B 258, 414-420 (2007). !. "mit, A. "emrov, Early medieval coinage in the territory of Slovenia, Nucl. Instr. and Meth. B 252, 290-298 (2006). I. Nemec, P. Bohanec, !. "mit, in T. Tr%ek-Pe%ak (Ed.), Conserving and Restoring the Works of Art of Master HGG, National Gallery of Slovenia, Ljubljana 2004, p. 36-40. !. "mit, Analysis of the textile fragments by the PIXE method, Diana 10, 147-148 (2004/2005). K. Kavkler, I. Nemec, A. Smrekar, !. "mit, T. Tr%ek-Pe%ak, Investigation of the Slovenian impressionist paintings by the differential PIXE method, in G. Arun (Ed.), Studies on Historical Heritage 2007, Yildiz Technical University, Antalya 2007, p. 305-312. !. "mit, M. Holc, Differential PIXE measurements of thin metal layers, Nucl. Instr. and Meth. B 219-220, 524-529 (2004). !. "mit, Recent developments of material analysis with PIXE, Nucl. Instr. and Meth. B 240, 258-264 (2005). !. "mit, J. Isteni%, T. Knific, Plating of archaeological metallic objects studies by differential PIXE, Nucl. Instr. and Meth. B 266, 2329-2333 (2008). !. "mit, M. Ur$i%, P. Pelicon, T. Tr%ek-Pe%ak, B. "eme, A. Smrekar, I. Langus, I. Nemec, K. Kavkler, Concentration profiles in paint layers studied by differential PIXE, Nucl. Instr. and Meth. B 266, 2047-2059 (2008).

41

IN SITU CHEMICAL COMPOSITION ANALYSIS OF CULTURAL HERITAGE OBJECTS USING PORTABLE X-RAY FLUORESCENCE SPECTROMETRY D. WEGRZYNEK1, 2, E. CHINEA-CANO1, A. MARKOWICZ1, 2, S. BAMFORD1, G. BUZANICH3, P. WOBRAUSCHEK3, CH. STRELI3, M. GRIESSER4, K. UHLIR4, A. MENDOZA-CUEVAS5 1 International Atomic Energy Agency, Department of Nuclear Sciences and Applications , A-1400 Vienna, Austria, e-mail: [email protected], fax: +43 1 260028222 2 Faculty of Physics and Applied Computer Science, AGH University of Science and Technology, 30-059 Krakow, Poland 3 Vienna University of Technology, Atomic Institute of the Austrian Universities, Stadionallee 2, A-1020 Vienna, Austria; 4 Museum of Fine Arts, Burgring 5, A- 1010 Vienna, Austria 5 Archaeometric Laboratory, Conservation and Restoration Cabinet, Havana's Historian Office, Havana, Cuba X-ray emission techniques play important role in the cultural heritage area. They provide information about chemical composition of an object upon bombardment of its surface with electrons, ions, or electromagnetic radiation. Their useful features include nondestructiveness, multielemental capability, and high sensitivity for inorganic components. Especially widely used is the X-ray fluorescence technique. It utilizes electromagnetic radiation generated by X-ray tubes or radioisotope sources. X-ray fluorescence equipment is relatively simple as compared to the charged particle-based spectrometers which are combined with scanning electron microscope or ion beam accelerator. X-ray fluorescence technique can be easily adapted for in situ measurements. A portable X-ray fluorescence spectrometer has been constructed utilizing commercially available, ready made components. The construction details of the spectrometer and examples of its application are given. The key features of the portable system are the use of polycapillary X-ray optics and a vacuum chamber attachment to enhance detection of low atomic number elements such as Mg, Al, Si, P, S, and Cl. The spectrometer was applied for chemical composition analysis of archaeological artifacts and works of arts from the collections of the Museum of Fine Arts (Kunsthistorisches Museum), Vienna, Austria. The investigated objects included ancient bronzes, coins, samples of pigments, and famous goldsmith work “Saliera” by Benvenuto Cellini (1500-1571). This work highlights also other projects related to the applications of nuclear analytical techniques in support of study and preservation of cultural heritage objects supported by the International Atomic Energy Agency and carried out in the Agency’s Member States.

Keywords: X-ray fluorescence, cultural heritage, non-destructive analysis, in situ.

42

1. INTRODUCTION Elemental analysis of archaeological artifacts and cultural heritage objects can help to reveal ancient material usage, technology of preparation, identification of provenance and in some cases, can be used as an indirect dating tool. A highly valuable result of elemental fingerprinting can help identification of forgery by the detection of anachronistic materials. X-ray fluorescence (XRF) technique is of particular interest for archaeologists because of its multi-elemental capability, generally high level of reliability and most important, being non-destructive and portable [1-3]. Valuable artifacts in many cases have irregular shapes and are fragile making it sometimes a challenge to preserve their integrity during analysis. Over the past years, XRF has been successfully developed to meet the requirements. XRF is now widely accepted and routinely applied by researchers in cultural heritage studies and restorations. In order to support study of cultural heritage objects, a dedicated (trans)portable XRF spectrometer has been designed and manufactured in the IAEA Seibersdorf Laboratories. The spectrometer was applied for chemical composition analysis of bronze samples and pigments as well as for provenancing of works of arts in the Museum of Fine Arts (KHM – Kunsthistorisches Museum) in Vienna [4-5].

2. EXPERIMENTAL A portable X-ray fluorescence spectrometer was used for chemical composition analysis of archaeological artifacts and works of arts. The spectrometer was designed and manufactured in the IAEA Seibersdorf Laboratories in collaboration with the Atomic Institute of the Austrian Universities, Vienna University of Technology. The overall view of the spectrometer, positioned in the front of a modern painting, is shown in Fig. 1. The instrument utilizes ready-made components which include a lowpower, Pd-anode X-ray tube, high voltage power supply, thermoelectrically cooled X-ray detector, signal processing electronics, X-ray collimator, polycapillary focusing optics, two laser pointers, miniature CMOS camera, xyz translation stages, rotational vacuum pump, and a MS Windows XP personal computer running the data acquisition and evaluation software. A customdesigned, detachable measuring head of the instrument hosts the X-ray tube and a small vacuum chamber attachment. All the spectrometer components, except the vacuum pump, can be mounted on a custom-made transportable carriage frame.

43

Fig. 1 Overall view of the portable X-ray fluorescence spectrometer. The spectrometer has been set up to directly characterize pigments on a test painting.

The X-ray fluorescence spectra of samples are excited using direct X-ray tube spectrum or a spectrum filtered through 50 micrometers thick Rh foil. The filtered mode is useful when analyzing polycrystalline type of materials to diminish the presence of interfering diffraction peaks and/or to improve the detection limits for certain elements. The X-ray tube is usually operated at 50 kV and 1 mA (50 W). The tube is powered by a manually controlled high voltage power supply. The accelerating voltage and current can be precisely adjusted within the ranges between 20.00 kV – 50.00 kV and 0.010 – 1.000 mA, respectively. The heat generated during the tube operation is dissipated through the tube housing, which is cooled down by a stream of air induced by build-in ventilator. When operated indoors, at room temperatures around 22 ºC, the tube temperature measured by an internal sensor never exceeds 30 ºC. The X-ray beam is collimated with a cylindrical, brass collimator (inner channel diameter equal to about 0.5 mm) or a polycapillary lens. For the polycapillary lens the effective beam spot diameter on the sample was measured by so-called knife-edge scan. The results of the measurements are shown in Fig. 2. The estimated full width at half maximum (FWHM) of the beam spot was equal to about 160 micrometers. The collimator and lens are mounted in a

44

motorized holder allowing for fast switching between them without the need for any re-adjustment of the collimating/focusing device.

Fig. 2. Knife-edge scan over a beam spot obtained with polycapillary lens.

The detector is a silicon drift type (SDD). Its energy resolution is 140 eV at 5.9 keV and 3 microseconds shaping time. The detector active area is 10 mm2 and effective thickness is 450 micrometers. The principle of the SDD operation allows the detector to maintain very good energy resolution and stable peak position also at high counting rates up to 105 counts per second (cps). The SDD is very useful during in situ operation due to its thermo-electrical cooling (lack of a bulky liquid nitrogen dewar limiting the time of in situ operation) and also due to its high throughput. Very often the analyzed sample contains highly abundant matrix element, e.g. Fe or Ca, resulting in a signal which would saturate or significantly degrade performance of a standard Si(Li) detector preventing the analysis of major/minor and trace elements using the same measuring conditions. In case of routine sample analysis in laboratory environment the material may be processed to eliminate the highly abundant, interfering matrix elements prior the analysis. However, it is not the case during in situ operation or during a noninvasive analysis. Contrary to a typical Si(Li) detector the SDD can operate at high counting rates without noticeable X-ray spectra degradation, the peaks of major/minor and trace components of a sample are registered accurately during the same measurement. It simplifies the instrument operation and reduces the measurement time (all data are collected in a single measurement lasting a few minutes). Additional advantage of using common measuring conditions for all elements of interest is less complicated data evaluation. The amplified signal of the SDD (shaper output) is digitized in a portable analog to digital converter/multichannel analyzer (ADC/MCA) unit. The collected spectra are transferred for inspection, storage and further evaluation to a personal computer running MS Windows XP operating system.

45

Fig. 3. Arrangements of the components in the vacuum chamber attachment.

The laser pointers and the miniature CMOS camera are used for precise selection of the irradiated area. With the aid of the xyz translation stages the instrument is aimed at a region of interest. The distance between the instrument and the sample is adjusted precisely so that the position of the X-ray spot on the analyzed object coincides with the point defined by two crossing laser beams. Only if the irradiated surface is at a proper distance the camera image shows one bright spot formed at the analyzed surface by the two crossing laser beams, otherwise, if the object is too close or too far, two spots are visible in the camera image. The pre-alignment of the laser pointers and the X-ray beam is done with a phosphor screen. Once the pre-alignment is completed it is valid until the collimator/polycapillary lens of the instrument is replaced. The alignment remains valid after switching between collimator and the polycapillary lens by using the motorized collimator/lens holder. The lasers can be switched off and in the camera image one can monitor the immediate vicinity of the analyzed region, otherwise difficult to observe due to relatively short distance (about 1-2 mm) between the instrument measuring head and the examined object. The whole procedure of positioning/selection of the region of interest can be accomplished without moving the sample. It is very useful during investigations of large objects, e.g. sculptures, wall paintings or other objects which can not be easily moved. For the analysis of small samples, e.g. coins, pieces of jewellery, small ceramic fragments, etc., the object can be mounted on a small xy translation table. The table is attached to the instrument measuring head. It allows for precise positioning of small objects without moving the instrument. The spectrometer is equipped with a compact vacuum chamber. The chamber is directly coupled to the X-ray tube. The chamber hosts the X-ray optics, detector, laser pointers, and a miniature CMOS camera. There is also an illumination photodiode installed to provide light for monitoring the vicinity of the analyzed region. A close up view of the chamber, with the front and top wall dismantled, and the arrangement of the instruments inside is shown in Fig. 3. Fully assembled chamber is equipped with a 7.5 micrometer thick Kapton window installed in the front wall. The window is transparent to X-rays and

46

visible light. The air pressure in the chamber is maintained below 1 mbar by a rotation vacuum pump. The length of the path between the sample and the detector, as well as the path between the tube and the sample, still remaining at the atmospheric pressure is reduced to less than 2 mm. It reduces significantly the absorption of low energy X-rays in the X-ray tube-sample-detector path. The reduction of air path is necessary for efficient excitation and detection of the socalled low-Z elements, such as Na, Mg, Al, Si, P, S, and Cl.

Fig. 4. Detection limits of elements based on the measurements of the IAEA Soil-7 RM.

The detection limits of elements have been estimated by measurement of a well characterized sample in a form of pellet made of the IAEA Soil-7 Reference Material. The sample was measured for 1500 s live time at 50 kV/1 mA. The polycapillary lens was used to focus the unfiltered primary beam emitted by the Pd-anode X-ray tube. The obtained detection limits of elements are shown in Fig. 4.

3. APPLICATIONS The portable XRF spectrometer has been used to characterize chemical composition of several objects in the collections of the Kunsthistorisches (KHM) Museum (Museum of Fine Arts), Vienna. 3.1. Qualitative Analysis In Fig. 5 the spectrometer is shown during inspection of pigments in an Egyptian wooden stele, XXVI, Thebes 640B.C. The Egyptian blue pigment was identified by the presence of the peaks of Si, Ca, and Cu in the collected X-ray fluorescence spectrum.

47

a)

b)

Fig. 5. Direct in situ identification of Egyptian blue pigment (CaO·CuO·4SiO) in a wooden stele from XXVI, Thebes 640 B.C (Egyptian and Near Eastern Collection, Kunsthistorisches Museum, Vienna).

In Fig. 6 a richly decorated oriental saddle is shown. The presence of cinnabar, azurite, white, and lead has been confirmed. A qualitative analysis of the inorganic components is quite straightforward. It requires a proper energy calibration of the MCA which is usually achieved with measurement of two samples made of pure substances, e.g. titanium and molybdenum foils. Based on the positions of Ti-K! and Mo-K! peaks a liner relation is established between the channel number (x-abscissa) and the X-ray photon energy of the MCA spectrum display. When the energy calibration is done the positions of all the peaks present in the collected spectra can be expressed in kiloelectronovolts (keV), the characteristics peaks can be identified and associated with chemical elements present in the sample. Based on the peak proportions an experienced analyst may also estimate the concentration ratios of the identified elements. With additional information on the object, it is usually enough to confirm/exclude the presence of inorganic pigments, the use of specific technological process, reveal otherwise “invisible” patches of foreign inorganic substances younger than the original object, etc. In many cases the qualitative analysis will give enough information to answer the curator or restorer question. However, there are cases where precise knowledge about the chemical composition of the objects, its different parts, layers, or specific regions is of prime importance to determine the source of the material, to estimate the object age (indirectly), or to develop an optimum conservation strategy and to decide about the storage/exposition conditions. The determination of chemical composition of the sample (quantitative analysis) is

48

performed with the use of additional software. A few different approaches can be used, depending on the sample type, its degree of homogeneity, availability of standards and reference materials.

Fig. 6. Identification of pigments in an oriental saddle (Arms and Armour Collection, Kunsthistorisches Museum, Vienna).

3.2. Quantitative Analysis In a standard laboratory set up the chemical composition analysis is performed using well homogenized, specially prepared samples, e.g. pressed pellets or fused glass beads. In such a case the measurement geometry is well defined. For a well defined sample and measuring geometry the relation between the intensity of an X-ray peak and the chemical composition of the sample can be described analytically: E ' dI 0 $ )f (E, E , E ) ,( , c, c )dE (1) I = cG* (E ) K (E , E ) f (E , E , ) , ( , c, c K+

K+

!

max

Eabs

" % & dE #

K+

Abs

K+

j =1...n

Enh

K+

j =1...m

j =1...m

Eqn. (1) describes the intensity of an X-ray characteristic peak, IK!, excited in a multielemental sample with the use of polychromatic primary radiation as a function of the element mass fraction, c, and a number of other parameters including mass fractions of all other elements present in the sample, cj=1…n, geometrical constant (G), detection efficiency ("), intensity distribution of the primary radiation (dI0/dE), energy of the primary radiation (E), energy of the Xray peak (EK!), X-ray peak production cross-section (K), and the absorption and enhancement correction coefficients (fAbs and fEnh, respectively). The integral in Eqn. (1) goes over primary photon energies above the photoelectric absorption edge (Eabs) with the initial vacancy leading to the production of the characteristic peak up to maximum energy of photons (Emax) present in the primary excitation spectrum. The primary spectrum includes bremsstrahlung and characteristic radiation emitted by the X-ray tube, eventually modified by a filter or polycapillary optics. For samples similar in matrix composition and characterized by a limited variability of the analyte a simple linear relation between the analyte mass

49

fraction and X-ray peak intensity can be assumed. In such a case, for a given element and its characteristic peak, the absorption and enhancement correction coefficients can be regarded constant and sample independent. Also the integral becomes sample independent. All the sample independent constants can be merged together, including G and "(EK!), in one calibration coefficient, SK! : I K! = cS K! (2) Coefficient SK! is specific for a given characteristic peak and it is common for all samples characterized by similar matrix composition. For each analyte the coefficient SK! is determined by measuring a few calibration samples made of similar or the same material as the “unknowns” and characterized by other analytical techniques. Due to similarity of the calibration samples and the unknowns the interfering effects cancel out or are corrected by the calibration coefficient. It is also seldom that for each type of analyzed objects and analyte we find calibration samples very similar in matrix composition and characterized by other techniques to apply Eqn. (2). We would also like to determine the content of not just one or two elements but as many as possible. In such a case one can start with calibrating the spectrometer by using on Eqn. (1). A set of calibration standards prepared from pure substances, e.g. metal foils, oxides or other pure and stable chemical compounds is measured in well defined conditions (sample distance, incidence and exit angles, etc.). For each peak of interest a combined value of G"(EK!) is calculated, denoted as S’K!: (3) S K( + = c!

Emax

Eabs

I K+ ' dI 0 $ " K (E , E K+ )f Abs (E , E K+ , * , ) , c, c j =1...n )f Enh (E , E K+ , E j =1...m* , ) , c, c j =1...m )dE % & dE #

For other elements/peaks of interest not present in the calibration standards the corresponding values of S’K! can be obtained by interpolation. The concentration of elements in the unknown samples is done by solving a system of equations (at least one equation per each element to be determined): c= S K( + !

Emax

Eabs

...

I K+ ' dI 0 $ " K (E , EK+ ) f Abs (E , EK+ , * , ) , c, c j =1...n )f Enh (E , EK+ , E j=1...m* , ) , c, c j=1...m )dE % & dE #

(4)

The accuracy of calculations is better if the sum of the concentrations of elements is close to 100% or if the chemical composition of the sample matrix not detectable by X-ray fluorescence spectrometry is know and is included in Eqn. (4). The accuracy can be further improved taking into account the in-situ measuring conditions. It can be done by measurement of an additional reference

50

material (RM) sample with similar characteristics to the unknown in terms of composition, surface topology, grain size distribution, etc. The measurement can be used to correlate the known concentrations of elements in the RM, c’, with the ones obtained by solving Eqns. (4): (5) c! = kc The correlation coefficient, k, is then used to correct the concentrations

Fig. 7. Correlation between the determined and certified concentrations of elements in bronze standard samples.

determined in the “unknowns” of the type similar to the RM. Such approach allows for using one calibration for various materials with the aid of a few RM samples. Such approach was used for analysis bronze sculptures. In Fig. 7 a correlation is shown for bronze reference material samples. The method was applied for the determination of a bronze disk sculpture “Madonna and Child” by Donatello (Florence, 1444). The sculpture is shown in Fig. 8.

51

Fig. 8. The portable XRF spectrometer positioned in the front of the Madonna and Child bronze disk sculpture by Donatello (Florence, 1444). The spectrometer has been set up to analyze the spot on the knee of the Child. The analyzed area, the red spot, has been marked with two laser pointers.

An X-ray spectrum of spot located on the forehead of the Child is shown in Fig. 9. Despite a lack of gold plating in this point one can notice the presence of gold. Relatively large peaks of Ca and Cl originate from patina.

Fig. 9. X-ray spectrum of a spot on the forehead of Child, from the bronze tondo “Madonna and Child” by Donatello.

52

Several spots on the bronze sculpture were analyzed. By applying the model described by Eqns. (4) and taking into account the presence of Au and Hg in the gold plating layer the average composition of the bronze and the plating layer were obtained. The results are present in Table 1.

Bronze

Fe 0.1 - 0.2

Cu 65 - 75

As 0.03 - 0.2

Concentration, [wt. %] Sn Sb 15 - 27 4.1 - 9.2

Pb 0.5 - 1.3

Au -

Hg -

Gold 84-86 14-16 plating Table 1. Determined chemical composition of the bronze and gold plating of a bronze tondo Madonna a Child by Donatello (Florence, 1444).

Relatively high concentration of mercury in the gold plating layer confirm that the gold film was applied by using fire-gilding technique. The accuracy of the results obtained during in situ measurements is always less as compared to the results which could be obtained in a laboratory environment after sample preparation. It is in the range of 5% - 20% relatively. Also characterized with the portable X-ray fluorescence spectrometer was a precious goldsmith work, so-called Salliera by Benvenuto Cellini (1500-1571), see Fig. 10. The composition of the gold alloy was determined. As can be seen from the spectra presented in Fig. 11 the gold alloy is relatively pure with only minor presence of copper and silver. One can also notice the importance of selecting the proper measuring conditions for analysis of certain samples. In this case the use of Rh filter in the primary radiation path eliminates the presence of diffraction peaks in the XRF spectra and makes easier the interpretation of XRF data.

Fig. 10. Goldsmith work, so called Saliera, by Benvenuto Cellini (1500-1571). Determination of gold alloy composition.

53

Fig. 11. X-ray spectra of the Saliera gold alloy. Top: direct excitation with not filtered X-ray tube readiation; bottom: direct excitation with 50 µm thick Rh filter in the primary radiation path.

4. CONCLUSIONS A portable X-ray fluorescence spectrometer has been designed and manufactured in the IAEA Seibersdorf Laboratories in collaboration with Atominstitute, Vienna, Austria. The spectrometer has been applied to chemical composition determination of pigments, alloys, and identification of inorganic components in the objects from the collections of Museum of Fine Arts, Vienna, Austria. The portable XRF technique allows for non-invasive identification and determination of inorganic components of analyzed objects. Other X-ray emission techniques, e.g. proton induced X-ray emission (PIXE) and electron probe micro-analysis (EPMA) are also applied for characterization of objects of cultural heritage in a nondestructive way. In recent years the International Atomic Energy Agency supported several regional and national technical cooperation projects and coordinated research programmes related to the utilization of nuclear related analytical techniques for the analysis and preservation of cultural heritage. The management and proper care of the cultural heritage is of importance for both the developed and developing Member States of the IAEA. Especially developing countries should take special care about their cultural heritage, which very often generates significant income due to the tourism and associated commercial activities. There is a need to share the knowledge about the advanced analytical techniques among the

54

archaeologists, conservators, and museum curators. On the other hand the scientists and analytical specialist should get acquainted with problems and methods of the conservation science. The technical cooperation programmes and the general programme activities of the IAEA provide a platform for such collaboration on a regional and worldwide scale. LITERATURE 1.

C.R. Appoloni, M.S. Blonski, P.S. Parreira, L.A.C. Souza, Nucl. Instrum. Meth. in Phys. Res., Section A 580: 710-713 (2007).

2.

R. Cesareo, A. Castellano, G. Buccolieri, S. Quarta, M. Marabelli, P. Santopadre, M. Leole, A. Brunetti, Nucl. Instrum. Meth. in Phys. Res., Section B 213: 703-706 (2004).

3.

K. Castro, N. Proietti, E. Princi, S. Pessanha, M.L. Carvalho, S. Vicini, D. Capitani, J.M. Madariaga, Anal. Chim. Acta 623: 187-194 (2008).

4.

G. Buzanich, P. Wobrauschek, C. Streli, A. Markowicz, D. Wegrzynek, E. Chinea-Cano, S. Bamford, Spectrochim. Acta Part B 62: 1252-1256 (2007).

5.

K. Uhlir , M. Griesser, G. Buzanich, P. Wobrauschek, C. Streli, D. Wegrzynek, A. Markowicz, E. Chinea-Cano, X-Ray Spectrometry 37: 450 – 457 (2008).

55

INTEGRATED GEOPHYSICAL TECHNIQUES FOR THE HIGHRESOLUTION STUDY OF ARCHAEOLOGICAL SITES * MICHELE PIPAN Department of Geological, Environmental and Marine Sciences (DISGAM), University of Trieste, via Weiss, 1 Trieste, 34127, Italy EMANUELE FORTE DISGAM - Exploration Geophysics Group, Near Surface Laboratory, via Weiss 1 Trieste, 34127, Italy We propose a combination of magnetic, electromagnetic (multi-fold ground-penetrating radar) and seismic (tomographic) methods and apply them to the study of three archaeological test sites. The buried archaeological remains in the areas of study are basically characterized by poorly known and variable size, geometry, location and physical properties. The integration of different geophysical techniques helps identifying, imaging and mapping anomalies of potential archaeological interest due to the sensitivity to different properties of the materials. The geophysical results are validated by archaeological excavations that uncover targets in the depth range between 1 to 5 meters from topographic surface.

1. Introduction Geophysical prospecting can provide archaeologists with crucial information about location, depth and characteristics of buried targets of potential archaeological interest. Such information helps archaeological teams in planning and optimizing excavations. Interaction between archaeological and geophysical teams is becoming common practice and successful results are reported by several authors (see e.g. [1], [2], [3], [7], [9], [10], [11], [15]). Nonetheless, the elusive nature of buried cultural heritage and the requisites for detailed site imaging and characterization, imposed by scientific research and engineering applications in areas of archaeological interest, drive geophysical research towards the implementation of advanced methods that can reduce uncertainties *

This work is supported by Halliburton-Landmark Academic Award to EGG-DISGAM, by PRINCOFIN 2006047924_003, by an Italian Ministry of Foreign Affairs’ grant in the framework of bilateral Algerian–Italian cooperation protocol.

56

in high-resolution subsurface archaeo-geophysical studies. This study focuses on the integration of magnetic, electromagnetic (multi-fold ground-penetrating radar) and seismic techniques and on the test of the integrated techniques in selected archaeological sites where low contrast of physical properties and variable depth, size and shape of the targets create conditions of peculiar complexity for geophysical data analysis and interpretation. The primary objective is to perform a joint subsurface characterization based on the integrated interpretation of geophysical images and multiple geophysical information for each point in the volume of study. The task is accomplished through the application of 2-D multi-fold ground-penetrating radar (mgpr) methods, which provide 2-D and 3-D images as well as radar wave velocity fields and instantaneous radar trace attributes. Radar data are further integrated by magnetic measurements and, at one test-site location, by seismic tomography ([7]). The main benefits in the use of ground-penetrating radar (gpr) techniques are the unequalled imaging performances in the shallow subsurface, i.e. in the depth range of interest in archaeological studies (i.e. less than 20 meters at most sites; [5]). Multi-fold data acquisition and processing (see e.g. [2], [3], [4], [5], [6], [10], [12]) provide additional information and benefits, such as radar velocity field, radar wave attenuation as a function of offset, material characterization through radar trace attributes ([8], [14]), accurate localization of targets and image enhancement. Quantitative information, such as electromagnetic wave velocity and attenuation, can be linked to physical properties of the materials (dielectric constant, conductivity). Enhanced quality of the subsurface image, accurate target localization and shape reconstruction in 3-D and physical properties are crucial to reduce the intrinsic uncertainty in archaeo-geophysical data interpretation due to the unknown and highly variable characteristics of the buried targets. Recent improvements in data acquisition equipment and processing techniques resulted in outstanding outcomes of the application of magnetic methods to archaeological prospecting (see e.g. [1], (13]). The high sensitivity of modern magnetometers coupled with focusing of the analysis on small variations of the magnetic field (frequently less than 1 nT) allowed the identification of archaeological targets even in conditions where the contrast between magnetic susceptibility of archaeological targets and surrounding materials is very low. One important element in the successful application of magnetic techniques to archaeological studies is the geometrical patterns associated with buried foundations or remains of buildings, which make the interpretation of results unequivocal. Eventually, seismic methods are basically associated with insufficient vertical/horizontal resolution for archaeological applications. Nonetheless, ultra-high-resolution reflection methods, multi-component techniques and seismic tomography are opening the route towards an extensive use of seismics in archaeological prospecting, particularly where imaging

57

capability is required beyond the depth range attainable by gpr methods (i.e. roughly depths greater than 20 m). We selected three test-sites to test mgpr methods integrated by magnetic and seismic techniques: one in Egypt (Medinet Madi, Fayoum), two in Italy (Aquileia, Udine). All of them are characterized by challenging subsurface conditions, mainly related to low soil-target contrast of physical properties. The results of the geophysical survey give evidence of subsurface anomalies of potential archaeological interest that were successfully validated by archaeological excavation.

2. Methods Multi-fold ground-penetrating radar, magnetic prospecting and seismic tomography are synthetically described to elucidate data acquisition and processing techniques applied in the study of the selected test sites. The interested reader will find more detailed descriptions of the methods in the cited scientific literature. 2.1. Multi-fold ground-penetrating radar (mgpr) Ground-penetrating radar (gpr) is a pulsed electromagnetic technique designed to detect dielectric discontinuities buried beneath the earth’s surface (see e.g. [16]). The basic system is composed of a couple of transmit and receive antennas, which are used to propagate wide-band electromagnetic radiation (in the range between 25 MHz and 2 GHz for most applications) and to detect the backscattering from targets. Arrival time and amplitude of the backscattered radiation are exploited to image dielectric discontinuities. Ursin [17] proposed a unified treatment of elastic and electromagnetic (EM) wave propagation in horizontally layered media and such formal equivalence allowed sharing procedures for analysis and data processing that are used in exploration seismology. Multi-fold gpr (mgpr) refers to the extension of the gpr method to multiple offsets, i.e. multiple transmit-receive couples separated by different distances [6], [10]. Such extension allows the application of techniques employed in multi-channel reflection seismics, with specific reference to image enhancement, focusing of backscattered radiation, target localization and evaluation of material properties from amplitude and arrival time of the backscattered radiation [18]. We used an ultra-wide band (UWB) system (RAMAC, Malå Geoscience) equipped with bow-tie shielded (250, 500 MHz) to test the mgpr method at the selected sites. A distance triggering device based on an electro-mechanical odometer was used to ensure constant 5 cm trace spacing. Average positioning

58

accuracy was below 0.2% . Conventional single-fold methods [i.e. single and constant distance (offset) between transmit-receive antennas] were used in reconnaissance surveys at all test sites. We successively applied multiple common offset profiles, to obtain multi-fold sections with average 1200 % fold (i.e. 12 offsets for each data acquisition point). Data processing was based on the following sequence: velocity analysis, spherical divergence correction, predictive deconvolution, stack and post-stack time migration, band-pass filter, wavelet Transform (WT) based identification of weak reflectors in noisy background (deep contacts). The azimuth-offset analysis was performed on test profiles with different azimuth before mgpr data acquisition to select grid orientation and offset range.

2.2. Magnetic prospecting Magnetometry measures the Earth’s magnetic field and maps its variations in order to identify the location of subsurface sources of anomalies. The rapidly time-variant component of the geomagnetic field, mainly due to electric currents in the ionized layers of the upper atmosphere, imposes the use of at least two measuring devices to remove the time-dependent component. A practical solution is to measure the gradient of the geomagnetic field by keeping the two sensors at fixed and constant distance in space. We used a high-resolution cesium magnetometer with a sensitivity of ±0.01 Nanotesla (nT) to measure the total field and the vertical gradient between two sensors spaced 1 meter apart. Ten measurements per second were taken at walking speed and the use of markers, spaced 5 meters apart, allowed the correct georeferencing of each measured value. Time variation was removed from total field measurements by subtracting the average value calculated every 25 meters. Comparison of gradient and total field data confirmed the validity of total field correction. Total field data were thus mapped in a average range of ±5.0 nT from the mean calculated geomagnetic value. The values exceeding such range were basically associated with metal (iron) scraps or burned materials. A scale based on colour changes corresponding to 0.020 nT magnetic field variations was utilized to map total field measurements.

2.3. Seismic tomography Tomographic inversion aims at reconstructing the kinematical and/or dynamical characteristics of a medium through a detailed mapping of velocity and/or attenuation. We focused on the kinematical analysis and exploited direct arrivals of seismic waves, i.e. transmission through the volume of interest and

59

measurement of traveltimes from a set of source-receiver pairs to calculate the velocity field within a cone-shaped burial mound. The seismic source was a 2.5 kg hammer striking a cylindrical steel plate (12 cm in diameter, 3 cm thick). Data acquisition was performed at ground level to identify velocity anomalies related to the deeper sectors of the mound, where the funeral chamber was most likely located. We achieved a complete 2! angular coverage with dense angular sampling, namely 15° spacing among the 40 Hz geophones and the same angular value among the sources, the two arrays being staggered in space through a 7.5° shift. Traveltimes were picked from the first breaks of 398 raypaths selected out of a total of 576. 178 raypaths were discarded because they were associated with minimum offset source-receiver pairs that sampled only the shallow portion of the mound and provided no useful information for traveltime inversion. The computational grid for traveltime inversion was based on 400 square cells (1.2 m size). The selected cell dimension could be considered adequate to perform the localization of the funeral chamber, which was the primary objective of the study. We applied SIRT (Simultaneous Iterative Reconstruction Technique; [19], [20] , [21]) to reconstruct the velocity field. After picking of the arrival times, we traced curved rays through an estimated velocity model, segmented the raypaths into the pixels of the model, computed the time residual for each ray, and iterative back projected the time estimates to produce model updates ([19]). The inversion process was iterated 7 times, by performing cycles of forward traveltime computation, residuals determination, velocity field upgrading, until we obtained a RMS residual of 1.02% . This value corresponds to a global RMS residual of 0.38 ms, which is comparable with the precision in picking of first breaks. 3. Test site description 3.1. Test site A – Medinet Madi The site is located some 100 km to the south of Cairo (Egypt, Fig.1a), in the desertic area at the southern border of the Fayoum area. It is characterized by variable subsurface conditions, from regular layers of sandy loams produced by eolic deposition to localized build-ups of debris produced by demolished or fallen buildings, limestone blocks of variable size, adobe remains and vegetable fibres (palm leaves) used for shelters (Fig.1b). The mixture of sandy loam and adobe debris with high clay and organic matter content results in locally high conductivity (around 0.015 S/m from laboratory measurement). Conductive and chaotic materials are responsible for attenuation and scattering that locally reduce the penetration of radar waves and the amplitude of primary reflections.

60

Figure 1. (a) Location map of test site A, Medinet Madi (Fayoum, Egypt); (b) Example of shallow stratigraphic conditions from archaeological excavation at the Medinet Madi site.

3.2. Test site B – Aquileia Aquileia was one of the largest towns of the Roman Empire and it is now one of the most important archaeological sites in northern Italy (Fig.2). The area, abandoned for centuries after destruction by barbarian invaders, was covered by a layer of fine-grained sediments (silty loams) of average thickness not less than 100 cm. The site selected for the test is a polygon of approximately 1500 m2 that borders the local cemetery to the south. Two sample cores obtained in a radius of 500 meters from the site show that the layer of archaeological interest is an allochtonous soil made of sediments ranging from sandy gravels to sandy pelites.

61

Figure 2. Location map of test site B (yellow dot, Aquileia, Italy).

3.3. Test site C – Udine The site is a preserved late Bronze Age burial mound located in the alluvial plain of the municipality of Udine (Italy, [22]) and made of lenses of sediments n ranging from pebbles to clay. It has a conical shape with a maximum elevatio of about 4.5m above the surrounding ground level (Fig.3). The average diameter is 26.5m and the base area is about 550 square meters. We performed a complete topographic survey and obtained a Digital Terrain Model (DTM) of the mound, with precision in the range of ± 1 cm.

62

Figure 3. Location map, image and Digital Elevation Model of test site C.

4. Results Multi-fold ground-penetrating radar and magnetic gradiometry give evidence of subsurface features of archaeological interest at test site A. The magnetic anomaly map (Fig.4a), obtained from data filtered in the range between ± 10 nT from mean geomagnetic field of the site, shows orthogonal alignments that are compatible with buried building remains and consistent with azimuth of the buildings exposed by archaeological excavation of neighbouring sectors. Vertical and horizontal slices of the mgpr data volume (Fig.4b,c) confirm the results of the magnetic survey and indicate that the top of the buried remains is located at an approximate maximum depth of 2.0 m from topographic surface, based on the velocity field obtained from mgpr measurements.

63

Figure 4. Test site A: (a) magnetic anomaly map; (b) vertical cross-section of the mgpr reflection data volume (250 MHz); (c) horizontal cross-section of the mgpr reflection data volume (250 MHz, 25 ns two-way-time ! 1m depth)

A magnetic gradiometry reconnaissance survey was first carried out at test site B in order to focus the mgpr data acquisition on the magnetic anomalies of potential interest. The results show a peculiar pattern in the southern part of the explored sector, with a straight NNW-SSE path terminated by an NE-SW trending arch (Fig.5). Mgpr data obtained in this sector give evidence of sharp discontinuities located between 0.6 and 1.4 m from topographic surface (Fig.6a,b). Reflection coefficient analysis indicates that the deeper material is characterized by smaller dielectric constant. Such characteristic is compatible with a sediment-buried foundations/remains contact, where the latter can be made of limestone or brick and exhibit therefore smaller porosity and fluid content compared with the shallow unconsolidated sediments, as demonstrated by several similar cases in the Aquileia area. Eventually, horizontal slices of the mgpr data volume (Fig.6c) show reflection pattern that match the magnetic anomaly and support the hypothesis of a buried structure. The areal shape of

64

magnetic/mgpr anomalies is consistent with the perimeter of a circus, a hypothesis that is confirmed by the available documentary evidences.

Figure 5. Test site B: magnetic anomaly map.

Figure 6. Test site B: example of cross-section of the mgpr reflection data volume (500 MHz), (a),(b) vertical sections; (c) Horizontal section (25 ns two-way-time ! 1m depth).

The peculiar characteristics of test site C (burial mound) encouraged an integrated application of mgpr and seismic tomography to overcome the respective limits of penetration and resolution. Fig.7 is an integrated display of the seismic tomogram and two selected mgpr sections. The tomogram is obtained at present ground level and allows the identification of a remarkable

65

velocity anomaly (D) in the central-northeastern sector of the tumulus. The mgpr sections cross the mound approximately in NS and EW direction and intersect at the top of the tumulus. Clear radar reflector are interpreted in a vertical range (measured from the top of the mound) from 40 cm to nearly 500 cm. The archaeological excavation verified the deep anomaly identified by seismic tomography (D), which is associated with the funeral chamber. The shallow radar reflector are associated with layers and lenses of sediments laid down during construction and characterized by variability of grain size and composition.

Figure 7. Integrated display of mgpr reflection data (250 MHz) and seismic transmission tomogram at ground level.

5. Conclusions The integration of different geophysical techniques can reduce uncertainties in the interpretation of data from archaeological sites and improve both the quality of subsurface models and the amount of quantitative information extracted from the datasets. The geometric coherence of anomaly patterns in magnetic data is often a valid indicator of the presence of buried cultural heritage but even in the best defined cases it can provide only approximate information about burial depth and 3-D subsurface structure. Moreover, a quantitative characterization of

66

the subsurface materials requires additional measurements to achieve a reliable inversion of physical (i.e. electromagnetic, elastic) properties of the materials. The proposed integration of magnetic methods with mgpr and seismic techniques provides an effective solution in terms of imaging capabilities and material characterization. Mgpr can image the subsurface with the highest resolution presently allowed by geophysical methods and it is therefore an ideal complement of magnetic methods in archaeological prospecting. Multi-offset radar data further allow estimates of dielectric constants and evaluation of conductivity of materials. The unequalled performance of mgpr is on the other hand limited where conductive materials (e.g. clay, salt or brackish water) are at the surface or in the top shallow layers. In such conditions, the penetration of radar waves is severely reduced and, in some cases, the application of mgpr is impossible. Seismic methods can overcome the limitations of gpr in terms of penetration, even though a resolution level compatible with archaeological applications requires the use of dedicated wide-band sources and measurement of the slower components of the wavefield (i.e. S-waves). Nonetheless, favourable topographic conditions allow the application of transmission tomography, which allows identification of anomalies (velocity, attenuation) related to elastic properties of the materials. The successful application of the integrated geophysical techniques to different subsurface conditions and buried targets shows the flexibility of the proposed method and the important contribution that geophysics can offer to the non-invasive study of archaeological sites. Acknowledgments This research was supported by a Halliburton-Landmark academic award, by PRIN-COFIN 2006047924_003, by an Italian Ministry of Foreign Affairs’ grant in the framework of bilateral Algerian–Italian cooperation protocol. We are grateful to Elena Barinova, Edda Bresciani, Paola Cassola Guida and Susi Corazza who coordinated the archaeological work at the test sites and helped in data interpretation.

References 1. 2.

Becker, H., and Fassbinder, J.W.E., 2001, Magnetic Prospecting in Archaeological Sites, Monuments and Sites VI, ICOMOS, Lipp GmbH Muenchen, ISBN 3-87490-675-2 Berard, B.A., and Maillol, J.M., 2007, Multi-offset ground penetrating radar data for improved imaging in areas of lateral complexity -

67

3. 4.

5. 6. 7. 8. 9.

10.

11. 12.

13.

14. 15.

application at a Native American site, Journal of Applied Geophysics, 62, 167-177 Berard, B.A., and Maillol, J.M., 2008, Common- and multi-offset ground-penetrating radar study of a Roman villa, Tourega, Portugal, Archaeological Prospection, 15, 1, 32-46, DOI 10.1002/arp.319 Booth, A.D., Linford, N.T., Roger A. Clark, R.A., and Murray, T., 2008, Three-dimensional, multi-offset ground-penetrating radar imaging of archaeological targets, Archaeological Prospection, 15, 2, 93-112, DOI 10.1002/arp.327 Davis, J., and Annan, P., 1989, Ground-penetrating radar for highresolution mapping of soil and rock stratigraphy, Geophysical Prospecting, 37, 531-551 Fisher, E., McMechan, G.A., and Annan, P., 1992, Acquisition and processing of wide-aperture ground-penetrating radar data, Geophysics, 57, 495-504 Forte, E., Pipan, M., 2008, Integrated seismic tomography and Ground Penetrating Radar (GPR) for the high-resolution study of burial mounds (tumuli), Journal of Archaeological Science, 35, 9, 2614-2623 Lemke S. R., 2000, GPR attribute analysis, proceedings of SAGEEP 2000, 263-272 Neubauer ,W., Eder-Hinterleitner, A., Seren, S., and Melichar, P., 2002, Georadar in the Roman civil town of Carnuntum, Austria: an approach for archaeological interpretation of GPR data, Archaeological Prospection, 9, 135-156 Pipan, M., Baradello, L., Forte, E., Prizzon, A., and Finetti, I., 1999, 2D and 3-D processing and interpretation of multi-fold ground penetrating radar data: a case history from an archaeological site, Journal of Applied Geophysics, 41, 271-292 Pipan, M., Baradello, L., Forte, E., and Finetti, I., 2001, Ground penetrating radar study of iron age tombs in southeastern Kazakhstan, Archaeological Prospection, 8, 141-155 Pipan, M., Forte, E., Dal Moro, G., Sugan, M., and Finetti, I., 2003, Multifold ground-penetrating radar and resistivity to study the stratigraphy of shallow unconsolidated sediments, The Leading Edge, 22, 876-881 Schultze, V., Linzen, S., Schüler, T., Chwala, A., Stolz, R., Schulz, M., Meyer, H.G., 2008, Rapid and sensitive magnetometer surveys of large areas using SQUIDs - the measurement system and its application to the Niederzimmern Neolithic double-ring ditch exploration, Archaeological Prospection, 15, 2, 113-131, DOI 10.1002/arp.328 Senechal P., Perroud H. and Senechal G., 2000, Interpretation of reflection attributes in a 3-D GPR survey at Vallee d’Ossau, western Pyrenees, France, Geophysics, vol. 65, 1435 –1445 Witten, A.J., 2006, Handbook of geophysics and archaeology, Equinox Publishing Ltd, ISBN-10: 1904768601

68

16. Daniels, D.J., (Ed.), 2004, Ground Penetrating Radar 2nd edition, The Institution of Electrical Engineers, ISBN 0 86341 360 9 17. Ursin, B., 1983, Review of elastic and electromagnetic wave propagation in horizontally layered, Geophysics, 48, 8, 1063-1081 18. Yilmaz, Ö., 2001, Seismic Data Analysis, Society of Exploration Geophysicists, Tulsa OK, USA, ISBN 1560800941 19. Brzostowski, M.A., and McMechan, G.A., 1992, 3-D tomographic imaging of near-surface seismic velocity and attenuation, Geophysics, 57, 3, 396-403 20. Menke, W., 1984, The resolving power of cross-borehole tomography, Geophys. Res. Lett., 11, 105-108 21. Tien-when, L., and Inderwiesen, P., 1994, Fundamentals of Seismic Tomography, Society of Exploration Geophysicists, ISBN 1 56080 028 3 22. Cassola Guida, P., and Corazza, G., 2002, Udine, S. Osvaldo tumulo protostorico. Scavi 2002, Aquileia Nostra, LXXIII, 754-757 (in Italian)

69

THERMOLUMINESCENCE DATING AND CULTURAL HERITAGE MARCO MARTINI, EMANUELA SIBILIA Dipartimento di Scienza dei Materiali e INFN Università degli Studi di Milano-Bicocca Via Cozzi, 53 - 20125 Milano m,[email protected]

Thermoluminescence Thermoluminescence (TL) is the emission of light observed during the heating of insulating or semiconductor materials, provided that they have been previously exposed to ionising radiation (McKeever 1985; Martini e Meinardi, 1997; Chen and McKeever, 1997) This irradiation may take place in the laboratory or in a radioactive environment. Another possibility, which is exploited in dating applications, is when a naturally occurring material is irradiated by the radiation field of its natural surrounding. The exposure to radiation somewhat perturbs the initial stable configuration of the material and heating allows the release of the accumulated energy. The existence of thermoluminescence is linked to the internal ordered structure of insulators, and to the presence of defects in its lattice. The process can be described, in a simplified way, using the energy band representation of insulators and assuming the presence of two kinds of imperfection in the crystal, as shown in Fig.1. As a consequence of the exposure to ionising radiation, electrons and holes (a hole is a vacancy of an electron) are produced in pairs: they can be captured in specific defects called “traps”, whose energies are within the forbidden gap of the crystal. These traps are metastable, and usually the lifetime of the trapped charges, electrons and holes, is very long at room temperature. The higher the exposure to ionising radiation, the higher the number of trapped electrons and holes. When the temperature of the crystal is increased, the carriers are raised energetically and freed from their traps to the conduction band from which they can recombine one another, thus emitting TL.

70

The curve representing the intensity of emitted light as a function of temperature is called glow-curve (Fig.2), and its shape and intensity depend on the material and on the characteristics of the irradiation field, i.e. the type and energy of radiation and its total amount. The study of the TL properties in a crystal is actually the study of the defects of its lattice and the understanding of the role played by defects and impurities in some physical properties of solids. TL is a very sensitive tool to detect imperfections even in very small quantities This last TL feature mentioned, i.e. its dependence on the amount of energy absorbed during irradiation, called “dose”, is important in the dosimetric applications of TL. (the SI unit for the energy absorbed due to the interaction of ionising radiation with matter, the dose, is the Gray (Gy ), corresponding to 1 Joule/kg). In many cases, in fact, the intensity of the TL is directly proportional to the absorbed radiation dose. Once the dose response is tested using calibrated laboratory irradiations, any unknown dose producing a given TL signal can be easily determined. Several artificial and naturally occurring materials show this favourable property, covering a very wide range of dose (10-2-108 Gy approximately). They are diffusely used in dosimetry and radiation protection practices and can be used to measure the doses due to professional exposure and those accrued as a consequence of nuclear accidents, as well as to monitor the dose inside nuclear plants. New materials have been developed to best fit the characteristics required by the main specific applications which are personnel, environmental, medical, retrospective and high-dose dosimetry (McKeever et al., 1995). Detection of TL signal The definition of TL itself suggests a rather easy way to detect it: what is needed is in fact an apparatus which is able to heat the samples under controlled conditions, and an efficient light detection system. In most cases, the very low level of the emitted signal and the difficulties in controlling and measuring precisely the sample temperature require the use of complex and specifically designed systems. This is particularly true when TL intensity is very low, like in dating applications or when basic studies on defect centres are carried out. In fig. 4 a schematic diagram of a TL measuring instrument is represented. Three main parts can be envisioned: the heating system, the detection system and the signal processing. The most common heating system is composed of a resistive planchet that heats up as a result of the passage of current through it. A common method of measuring the temperature is through the use of a thermocouple welded to the underside of the planchet. A photomultiplier tube (PMT) is normally used to detect the emitted TL. In fact the efficiency of high-

71

gain, very sensitive PMTs allows the detection of very low level signals with a convenient signal-to-noise ratio. The recent development of high sensitivity light detectors, like intensified diode arrays and Charge Coupled Devices (CCD), has allowed the measurement not only of the amount of emitted light, but also of its wavelength (Martini et al.,1996), obtaining information about the centres involved in the recombination processes. An example of such spectra is reported in Fig. 3. Thermoluminescence dating Thermoluminescence dating is the only physical technique for determining the age of pottery presently available. It is an absolute dating method, and does not depend on comparison with similar objects. The application of thermoluminescence in archaeological and geological dating (Aitken 1985, 1990) is based on dosimetry: it stands on the fact that many naturally occurring TL mineral constituents of ceramics, including quartz and most feldspars, are able to act as dosimeters for the amount of ionizing radiation they are exposed to. This radiation mainly comes from the radioactive decay of uranium, thorium and potassium present in the ceramics itself and in its surrounding (typically the burial soil), at concentrations of a few part per million. The radioactive materials having long half lives of 109 years or more, the radiation flux is practically constant. An important point to single out is that, when pottery is fired, it loses all its previously acquired TL. Thus, after cooling, the natural radioactivity causes thermoluminescence to build up again so the older an object is the more light is produced (Fig. 5). The TL level measured in pottery is associated to the dose accumulated since it was fired in kiln , unless there was a subsequent reheating. Any heating at high temperature acts as a clock resetting event. This usually occurs when the items are heated over 400°C. In archaeology, thermoluminescence dating is specific for ceramics bricks, cooking hearths, incidentally or deliberately fired rocks such as flints or cherts. If the radioactivity of the pottery itself, and its surroundings, is measured, the dose rate, or annual increment of absorbed dose, may be computed. The age of the pottery, in principle, may then be determined by the relation Age = Absorbed dose / Annual Dose-rate Typically we are dealing with absorbed doses ranging from a few to a few tens of Gy. The dose-rate is usually within the range 1-10 mGy/year. Even if the principles on which TL Dating is based are rather simple, the practical procedures are not. The precise evaluation of both absorbed dose and dose-rate requires the consideration of various factors affecting the calculations. For example, one of this factors is the way in which the different types of

72

radiation, α, β, or γ, are absorbed by the thermoluminescent minerals contained as small crystals in the pottery. The main dating techniques were developed on the basis of the differences in absorbing radiation by grains of different sizes. The so called "inclusion" technique (Fleming, 1970) considers only quartz crystal grains in the range 100200 µm extracted from the ceramic. The second major TL technique is the "fine grain" (Zimmermann, 1971) which uses all the material that can be extracted, for instance by drilling the sample. A grain size separation is then operated by settling the obtained powder in acetone suspension. It is possible to select a grain size range, typically between 1 and 8 µm. It must be mentioned that some complicating factors can occur, due to the specificity of the materials. In fact, while in dosimetry one can choose the best dosimeter available for a given radiation, in TL dating only the naturally occurring minerals can be used. The clay minerals have e usually low TL; a few of them are hardly thermoluminescent at all; some may not have a straight-line relationship between dose and TL. In addition, some of the accumulated signal may be lost due to thermal and anomalous fading (Wintle, 1973), where part of the TL is lost without thermal excitation, or it may exhibit a spurious, non radiation induced component (Martini et al., 1988). Also, if the sample was poorly fired in antiquity, the TL clock would not have correctly set to zero. The presence of one or more of these effects has great influence on the precision of the final result. If they are absent or small, or can be compensated or corrected for, then the error limits on the dates obtained are typically in the range ±4 to ±8% of the age. Dating applications TL might in principle be used to date any archaeological material containing thermoluminescent mineral and subjected in the past to an heating sufficient to erase any previous signal. Ceramics, due to its widespread diffusion in archaeological excavations, is the more frequent material submitted, toghether with bricks from historical buildings. The clay cores from lost wax metal castings may also be tested. Heated stone material, such as hearths, pot boilers, and burnt flints, can be dated as well, even if some regions are known to present problems for TL, like Indonesia and West Mexico: objects from these areas usually do not successfully yield TL dates, due to the very poor TL characteristics of the raw materials locally used. Possible applications of TL dating beyond man-made artefacts are geological field, where aeolian, fluvial, coastal and, in some cases, marine sediments can be dated. In these case the signal resetting is due to the exposure

73

of sunlight during deposition. Also volcanic formations be dated. A few examples of application of TL Dating techniques in various archaeological fields are reported in detail in the next section. Excavation archaeology An interesting example in excavation archaeology is the case of Bet Gemal (Strus, 2003), an Israeli village inhabited from II century BC to the Islamic period (IX century AD). The site displayed exceptionally well preserved remains: a Roman-Jewish quarter, a Christian Byzantine settlement with several plants for oil and wine production (fig. 6) and an Islamic dwelling place. Each group of remains is relative to different chronological periods. The long occupancy of the site and the cultural and religious changes that took place resulted in a complex, cumbersome stratigraphy posing problems of absolute chronology, in particular regarding the duration of the different occupations. TL dating was performed on several domestic ceramics characteristic of the three periods. Supported by our results, listed in table III, the following absolute chronology of the site could be proposed. The Jewish occupancy had its break at the end of the I century AD, in the historic context of the Jewish-Roman Wars. For the two following centuries, the site should have been almost abandoned until the III century, when the repopulation of the site started and its prosperity grew; the remains of workshops of ceramics, wine and oil presses testify the economical prosperity of this phase. A successive development of the village occurred in the Byzantine period, linked to new constructions like a church and a further oil press that was functioning during the VI century. The last transformation of the village occurred during the VII century, after a destruction on a large scale probably due to the invasion of the Persian army in 614 AD or to the local Muslim victory over Byzantines in 643 AD. The destruction was followed by a general restoration of life, marked by the re-building of several houses and by new industrial and housing projects, until the final abandonment of the village somewhere in the IX century. The impressive stone structure depicted in fig. 6, the bigger of the three oil presses associated to the Byzantine phase, well testifies the economical importance of the site at that time. Another relevant application is the study of the chronology of the Cham civilisation, that developed in central and southern Vietnam from 6th to 16th century. In the frame of an Italian-Vietnamise Programme of Cooperation an extensive TL dating project of the MySon religious complex started in 2005. The site shows the remains of more than 70 buildings of different styles (Fig. 7) built in different periods but always with the same building technique. About

74

300 bricks and ceramics have been sampled and the presently available results show evidence of a chronology much more complex than supposed by former scholars, especially for what concerns the important edification phases occurred during X and XI centuries. Historical buildings Since the stratigraphic techniques initially developed for archaeology, have been extended to architecture, the relative internal sequence of the various building phases of a monument can be usually precisely determined. Their absolute chronology is however sometimes problematic or controversial. In such cases, the contribution of the TL dating techniques could be conclusive (Galli et al., 2002). It must be reminded, however, that care has to be taken when associating the TL age of a brick to that of the structure it belongs to, because the event that is determined is the last firing of the sample. Voluntary human actions (rebuilding, transformation, decay and restoration) can modify the position of a brick in the stratigraphic sequence of a building. Moreover, in case of reuse of materials from pre-existing structures, dates are older than the building; in case of upkeep or mimetic restoration, dates are younger than the building. In case of fire, this event will be dated. The contribution of the archaeometric techniques to the study of ancient buildings is anyway very important. The main advantages of this kind of application are the availability of large quantities of material, the homogeneity of environmental radioactivity and the lesser extent of humidity fluctuation. TL dating in architecture should therefore give precision higher than in excavation archaeology, as confirmed by the statistical analysis preformed on about 1300 ceramic samples submitted to our Laboratory for dating over the last ten years (Martini et al., 2001). It could be appreciated that errors lower than 6% are much more frequent when dating buildings rather than excavated samples. As an example, we report the results recently obtained for the San Lorenzo Church in Milano The cathedral of S. Lorenzo in Milano (fig.8), the more ancient testimony of roman and palaeochristian architecture in Milan, is a complex architectural structure that shows evident traces of several building interventions often lacking of sure chronological attribution.. After performing a detailed stratigraphic analysis on both external and internal surfaces to fix the general building sequences, the different phases were dated with thermoluminescence and radiocarbon. TL was applied only to unbroken bricks and fictile tubes sampled in several wall structures of the complex.. Radiocarbon was used on wooden charcoal

75

scrapes contained in the joint of mortars of walls. In absence of scrapes, calcium carbonates clots found in the mortars themselves were employed. In total , more than two hundred samples were analysed. The very good agreement of all the results relative to the original phase allowed to indicate the narrow period 390-410 A.D. for the foundation of the tetraconc, much storiographically significant than the previous ones, estimated on historical ground. Later medieval reconstruction phases, one of them interesting the dome during X century, have also been uncovered. Burnt flints The possibility of dating burnt flints by TL appeared soon a great challenge to contribute in studying sites whose age is beyond the upper limit of radiocarbon dating (about 40.000 years), and when organic materials are not abundant or not well preserved. Flint is dense siliceous sedimentary rock whose basic component is SiO2, occurring as silica, cristobalite/trydimite and α-quartz. Due to its hardness and conchoidal fracturing properties, it was largely employed in prehistory to manufacture a large number of artefacts (Fig. 9). Some of them were accidentally or deliberately heated and the burning is obviously essential for the erasure of the geological accrued TL. Goksu and co-workers (Goksu et al., 1974) highlighted the possibility of dating burnt flint, presenting at the same time the limits and the specific problems related to such materials: generally low TL sensitivity and sensitivity changes, spurious and regenerated TL and very low concentrations of radioactive elements, circumstance that attaches great importance to a precise evaluation of the ambiental dose-rate. Despite the problems encountered in this application, flint dating is widely used and the results played, for instance, a primary role in the revision of the chronology of the presence of Neanderthal man and of modern human in the Middle-East. We recently studied a group of 20 burnt flints from the prehistoric site of Fumane, in North Italy, Verona province (Martini et al., 2001). It is a huge cave, used as a shelter by ancient men, characterized by paleosurfaces extremely rich in bones and lithic objects. The study of this site is considered very important for the passage from Middle to Upper Palaeolithic and from Mousterian to Aurignatian age in Northern Italy and Europe. Some stratigraphic sequences of the site have been dated with radiocarbon but very few data regarding the human presence are available. The chronology obtained by TL, spanning from 79 + 11 ka to 57 + 12 ka BP, added key information to the archaeological and palaeoenvironmental history of this Pleistocene period, up to now poorly dated.

76

Archaeological glasses The chemical-physical behaviour of silica glasses suggested the use of TL techniques as a suitable method to date these materials. Actually, because of the amorphous nature of glass, numerous factors reduce its thermoluminescence sensitivity. The main problems encountered in glass dating are a generally low TL sensitivity (TL emission per unit of dose) and the emptying of TL traps due to sunlight exposure or to t he low stability of TL traps at room temperature. Both effects result in a loss of TL signal. Moreover, changes in TL sensitivity often occur after repeated heating and irradiation of the same sample. Due to these difficulties, at the present state of the art only a few percent of the samples analysed could be successfully dated.. We focused our attention to a particular class of glass, the vitreous tesserae composing mosaics (Fig. 10). Our study was performed on samples chosen as representative of six sets of differently coloured glass mosaic tesserae. They all belong to wall mosaic decorations and were found in archaeological excavations or taken from mosaics to be restored, all well dated on archaeological grounds. The thermoluminescent emission of these vitreous materials, lacking a long range periodic structure, is due to the impurities present or added to the glass network (Al, Mn, Cr…), the colour centres acting as electron traps and recombination centres. In fact, a good natural TL emission was observed in almost all tesserae, the blue ones being generally characterised by higher sensitivity. Samples were submitted to different protocols for TL measurements, previously described hiavari et al., 2001) and their TL properties were investigated in deep. This allowed selecting eight tesserae characterised by suitable TL behaviour (high sensitivity, trap stability, low optical bleaching and limited changes in sensitivity after heating), that were submitted to dating. They presented a TL sensitivity comparable with that of ceramics materials. For the external annual dose-rates the mean values typical of the different provenance areas have been assumed, with errors taking into account possible wide variations. Under these assumptions, ages with overall errors ranging form 15 to 18% have been obtained. It is however noticeable the general consistency of TL dates with the archaeological ones. It is also remarkable that we could date eight tesserae over the nineteen analysed: the percentage of suitable samples was about 40 % against the 5% reported for glasses up to now (Chiavari et al., 2001).

77

Clay-cores The first application of TL techniques to clay-cores dates back to 1974, when D.W. Zimmermann (Zimmermann et al., 1974) succeeded in testing the authenticity of core materials from a Bronze Horse of the New York Metropolitan Museum of Art. Further attempts devoted to dating, soon enlightened a series of difficulties, complications and limiting factors. First of all, the application is in principle possible only for the objects cast by lost-wax technique, using the remains of thermoluminescent clay-cores heated contemporarily to the casting itself. The possibility of dating such materials depends on its mineralogical composition, and particularly on the abundance of “good” thermoluminescent minerals like quartz and feldspars. A high concentration of carbonates and/or organic material is generally a disadvantage, for the associated spurious, not dose-dependent TL emission. Another phenomenon that is observed with higher frequency than in ceramics is the anomalous fading, a process which empties deep traps at room temperature. The evaluation of the environmental contribution to the annual dose-rate can be problematic, both for the often unknown “archaeological history” of the object to date and for the need to evaluate the shielding effect of the bronze layer on the external irradiation. Due to the sum of these circumstances, the achievable precision in dating bronzes is generally lower than in ceramics, the mean error being generally about +10% of the age. As a further remark, it must pointed out that dating the clay core is not dating the bronze statue itself, except when the correlation between the ages of the two objects is sure, or highly probable. It must be reminded that any TL dating refers to the last heating at high temperature experienced by the item to date: in case of restoration or repair performed by heating, this last event will be dated instead of the original one (Martini and Sibilia, 2003). The possibility of dating clay-cores is furthermore precluded if the object has been intensively radiographed before sampling out the core material. In such a case, unfortunately frequently recurring, the high energy X-ray exposure results in an accrued radiation dose that produces an additional TL emission, superimposed to the archaeological one. The evaluated palaeodose is consequently meaningless. Nevertheless, things are not always so discouraging, and often very satisfactory results can be obtained, like in the case of the Cellini’s Perseo (fig. 11) Benvenuto Cellini (Florence, 1500-1571) wrote that a "great cry of admiration" arose from the throng gathered to watch the unavailing of his Perseo in the Loggia dei Lanzi on April 27th, 1554. After about five hundred years of open air exposure, the state of conservation of the statue was critical,

78

due to the polluted, aggressive urban atmosphere, transforming its historical patina from insoluble to soluble salts. It was therefore fully restored and again its disassembling gave access to its interior, where important fragments of claycore were found. In this case, TL dating was mainly performed to check the reliability of the technique, being the dating itself beyond dispute. The dating result, 1540+35 AD, is in very good agreement with the historical records, confirming the potential of such application, the reliability of the laboratory protocols and the accuracy and precision of instrumental calibrations. Conclusions TL dating of ceramic materials is nowadays a consolidated and powerful technique which supports the archaeological and archaeometric researches. Precisions as good as +5% in the evaluation of the age of various kinds of archaeological findings are often reached, allowing the solution of archaeological or historical problems arising from samples chronologically relatively close. A systematic comparison of TL dating results with those obtained by other absolute dating techniques like radiocarbon and dendrochronology and the dating of samples already well independently dated on archaeological or historical ground is highly recommended, in order to check and improve precision and accuracy of the laboratory experimental procedures.

79

References 1. Aitken, M.J. Thermoluminescecene Dating, Academic Press, New York, 1985 2. Aitken M. J., Science-based Dating in Archaeology, Longman, London, 1990 3. Chen R. and McKeever S.W.S, Theory of Thermoluminescence and Related Phenomena, World Scientific, Singapore, New Jersey, London, Hong Kong, 1997 4. Chiavari C., Martini., Sibilia E. and Vandini M., Quaternary Science Reviews 20 (2001) 967. 5. Fieni L, Galli A., Martini M., Montanari C.,Sibilia E., Proceedings of the 34th Int. Symposium of Archaeometry, Zaragoza, 2004. 6. Fleming S.J., Archaeometry 12 (1970) 133. 7. Galli A., Martini M., Montanari C. and Sibilia E., Proceedings of the 33rd International Symposium on Archaeometry, Amsterdam, 2002. 8. Goksu H.Y., Fremlin J.H., Irwin H.T., Fryxell R., Science, 183 (1974) 651 9. Martini M. and Meinardi F., La Rivista del Nuovo Cimento, 20 (1997) 1. 10. Martini M., Paravisi S.and Liguori C., Radiation Protection Dosimetry, 66, (1996) 47. 11. Martini M., Sibilia E. , Proceedings of the International Conference Archaeometallurgy in Europe, Milano, 2003. 12. Martini M., Sibilia E. and Ferraro L., Proceedings of the 3rd International Conference on Science and Technology for the Safeguard of Cultural Heritage in the Mediterranean Basin, Alcalà de Henares, 2001. 13. Martini M., Sibilia E., Calderon T. and Di Renzo F, Nuclear Tracks, 14 (1988) 339. 14. Martini M., Sibilia E., Croci S. and Cremaschi M., Journal of Cultural Heritage, 2 (2001) 179. 15. McKeever S. W. S., Thermoluminescence of solids, Cambridge University Press, Cambridge 1985. 16. McKeever S.W.S., Moscovitch M. and Townsend P.D., Termoluminescence Dosimetry Materials: properties and Uses, Nuclear Technology Publishing, Ashford, 1995 17. Strus A., Khirbet Fattir-Bet Gemal. Two Ancient Jewish and Christian Sites in Israel, (LAS, Roma) 2003 18. Wintle A.G., Nature 245 (1973) 143. 19. Zimmermann D.W., Archaeometry 13 (1971) 29. 20. Zimmermann D.W., Yuhas S.M.P. and Meyers P., Archaeometry 16 (1974) 19

80

Figures

Fig.1. Traps levels in an insulating crystal

Fig.2. Examples of TL glow-curves

81

Fig.3 Ancient mosaic glass.Wavelength resolved TL spectrum

Fig.4: Diagram of a typical TL measuring system

82

Fig.5: TL growth vs time

Fig. 6 Bet Gemal excavation site

83

Fig. 7: A ruined tower at the MySon religious complex

84

Fig. 8: Back view of S Lorenzo Church in Milano

Fig. 9 Archaeological flints

Fig.10: Byzantine mosaic glass tesserae

Fig. 11: Benvenuto Cellini, Perseo

85

NEW X-RAY DIGITAL RADIOGRAPHY AND COMPUTED TOMOGRAPHY FOR CULTURAL HERITAGE F. CASALI , M. BETTUZZI, R. BRANCACCIO, M.P. MORIGI Physics Department, University of Bologna and INFN, Bologna, Italy X-ray detection systems for high resolution Digital Radiography (DR) and Computed Tomography (CT) have been developed at the Physics Department of the University of Bologna. The target of the research is the development of systems to be applied in cultural heritage conservation and industrial radiology. In the field of cultural heritage, different kind of objects (ancient necklaces, paintings, bronze or wooden sculptures) have to be inspected in order to acquire significant information as the method used to assemble, the manufacturing techniques or the presence of defects. These features could be very useful, for example, for dating works of art or determining appropriate maintenance and restoration procedures. Among the advanced methods available, 3D CT can be successfully used for the investigation of ancient works of art because it preserves their integrity and provides images of inner parts, otherwise not visible. Several high-resolution CT systems, to investigate objects of different sizes (from micro to macro), have been developed at our Department and will be presented in the paper. Some experimental results will be presented too as the micro CT reconstruction of Roman human tooth with carious, the cone beam CT analysis on an Egyptian cat-shaped coffin exhibiting the inner mummy, the CT of an ancient large globe (2 m of diameter) located in Palazzo Vecchio, at Florence, as well as some large painted tables of great artistic interest (i.e. a painting of Raffaello). KEYWORDS: X-ray tomography, micro-tomography, X-ray diagnostics, inner inspection, imaging, 3D reconstruction Contact: [email protected]

1. INTRODUCTION The first, natural, application of tomography was the study of the human body. The impact of this technique in diagnostic medicine has been revolutionary, since it has enabled doctors to view internal organs with unprecedented precision and safety to the patient. Thus, several medical CT systems were developed and today this diagnostic technique is well consolidated. The application of tomography to the Cultural Heritage and industrial fields represents instead an interesting innovation.

86

Medical CT is optimized for the human body (composed mainly by water) and cannot be successfully used on dissimilar objects like those of interest in the field of cultural heritage. In order to perform good, non-destructive evaluations, the most suitable CT system (source, moving equipment, detector and elaboration software) must be carefully chosen to avoid obtaining meaningless results. For that purpose, several X-ray CT systems have been developed at the Physics Department of the University of Bologna designed for specific applications. For small objects (i.e. fossil teeth and ancient jewels) it is necessary to use radiation source having very small focal spot and high spatial resolution detectors. For big or thick objects it is necessary to use highly penetrating radiation sources, very efficient detection systems and very advanced mathematical methods for the reconstruction of the 3D images. Commonly X-ray tubes up to 450 kV are used but a certain interest in high energy CT (i.e. LINAC with more than 1 MV) is now growing up. The detectors developed and used, at our laboratory, are of one dimensional type (linear detectors) as well as two dimensional type (planar detectors). Linear detectors work with a “fan beam geometry” and one-dimensional projections are collected, while planar detectors permit to obtain two-dimensional projections in “cone-beam geometry” [1], [2]. By processing the acquired data with dedicated algorithms, it is possible to reconstruct the two-dimensional inner slice in case of linear projections and the whole 3D volume in case of the two-dimensional ones. Moreover, by means of a dedicated translation axis, the 3D volume can be achieved with linear detectors in scanning mode. Descriptions are given below on the different kinds of equipment developed, considering Cultural Heritage and Industry requirements as well.

2. FAN BEAM SYSTEM Intensified Linear Detector. An interesting innovative detector has been developed at the Department of Physics of the University of Bologna. The main objective was to obtain a wide and intensified linear detector that maintains high spatial resolution. The absence of intensified linear CCD on the market brought us to adopt an innovative solution for the problem: to change the format of a standard intensified CCD camera using a suitably sized coherent fiber optics image guide. By using an intensified camera, we obtained an intensified linear detector [3]. The main components are an intensified digital camera and a coherent linear-to-rectangular fiber optics guide coupled with the photocathode of the camera (see Figure 1). The innovative element of the system is the FO guide, consisting of seven coherent fiber optics bundles arranged in such a way to

87

transform an almost linear area (128.8 mm x 1.45 mm) into a rectangular one. The FO guide output is directly coupled with the EBCCD’s photocathode. A layer of gadolinium oxysulfide (GOS) converts the X-rays into visible light on the input side of the FO guide. In this way an intensified multi-slice linear detector (5600 ! 35 pixels) was obtained. The system is able to make digital radiographies using doses equal to about 1/100 of the standard ones. The detector can be arranged either in Digital Radiography mode (DR) or in Computed Tomography (CT) mode by means of a high-precision translation and rotary mechanical device [3]. Figure 2 shows the CT of an osteoporotic bone [4] an of a tree. The spatial resolution is very high (about 30 microns) for such big samples so it is possible to detect the bone trabecula and the tree rings.

Figure 1. Diagram of the experimental set-up of the linear scanning system.

88

Figure 2. CT of an osteoporotic bone (a) and of a tree trunk (b).

3. CONE BEAM SYSTEMS

When using a cone-beam CT system information can be retrieved as 2D cross-section images or 3D full-volume images allowing the inspection and the classification of the object; moreover, by processing tomographic data, a 3D numerical model of the sample can be obtained for virtual reality applications or digital archives storage. 4. Micro-CT system

A micro-CT system has been set up on the basis of an X-ray detector developed within the framework of a collaboration with the Geosphaera Research Center of Moscow [5]. The detector consists of a Gd2O2S:Tb phosphor layer (30 mg/cm2) deposited on the entrance window of a 2:1 glass fibre-optic taper. The small face of fibre-optic taper is then coupled to a cooled Charged Coupled Device (CCD). The CCD has 1024!512 useful pixels and a 12 bit ADC. Pixel size is 15!15 µm2; therefore the effective area of the

89

detector is about 30!15 mm2 with a detection pixel size of about 30!30 µm2 [6]. The high resolution obtained (20÷30 µm) allows a detailed analysis to be made of the inner structure of fragments of fossils, artefacts and rocks. For example, the micro-CT system has been used in the analysis of paleobiological samples. In collaboration with the Anthropology Section of the National Museum “L.Pigorini” in Rome, a detailed program of micro-CT analysis has been set up for the investigation of paleoprimates [7]. Figure 3a shows the picture of a human molar from the necropolis of Isola Sacra, near Rome [8]. Tomography (Figure 3b) allows a 3D reconstruction of the tooth with a resolution of 30 µm (voxel side). Successively, through virtual cuts, parts of the volume can be removed to provide an inside view of the sample, enabling tooth lesions to be easily located (Figure 3c). A specific diagnosis of caries can be made with certainty based on the results of this analysis.

Figure 3. Picture of a human tooth. (a) 3D reconstruction of the tooth. (b) Virtual cut into the volume in order to locate the lesion (c).

Thanks to the high image resolution and definition of this detector, small rock samples can also be investigated in order to extract geological information. With this detector a rock samples can be analysed in order to study the included minerals as can be seen in Figure 4a and 4b.

90

Figure 4 . Rock sample with minerals enclosed. Tomographic reconstruction (a), segmented minerals (b).

5. Intensified tomographic system for medium-size objects.

Another tomographic system is based on an innovative EBCCD (Electron Bombarded CCD) camera (developed by Geosphaera Research Center), which has a very high sensitivity (about 5!10-5 lux) and allows the detection of low-light images, reducing exposure time and irradiated dose [9]. This EBCCD camera consists of a 24 mm photocathode, a high voltage intensifier tube (ranging from 5 to 10 kV) and a CCD chip with 1024!512 useful pixels of size 13.3!13.1 µm2 each. Electrons extracted from the photocathode hit directly the substrate of the CCD that is sealed inside the tube. In this way an higher conversion efficiency and an higher gain (up to 2000) are obtained with respect to conventional image intensifier. A 12-bit ADC, on the border of the EBCCD electronics, provides the digital output. A diagram of the intensified CT system is shown in Figure 5.

91

Figure 5. Diagram of the intensified CT system.

A micro-focus X-ray source provides a cone-beam source, which irradiates the object, projecting its attenuation profile on a 30x40 cm 2 GOS screen, placed in the entrance window of a large light-tight box. The light image produced on the screen is focused on the EBCCD photocathode by means of a lens. A programmable turntable rotates the sample with a fixed angular step and a set of digital radiographs of the object is collected at different points of view. The radiographic data are then processed and cross-section images of the object are reconstructed by means of specific mathematical algorithms. The set-up is similar to that of micro-CT system, but this intensified system makes it possible the investigation of bigger objects (several tens of cm of size) with a resolution of about 200 m, that is certainly better than that provided by standard medical CT scanners. Thanks to a collaboration with the Archaeological Museum of Bologna, this tomographic system has been used to inspect important archaeological samples, such as bronze objects of the Etruscan section and small mummies from the Egyptian Collection [10]. Among the samples investigated, particularly interesting is a catshaped coffin (Figure 6a). The size of the sarcophagus is 37.7!10!19.5 cm3 and it has a structure composed of different materials. CT data allow very fine distinctions to be made among materials with different densities, thus providing a large amount of information. Figure 6b shows a 3D reconstruction of the sarcophagus, while figure 6c shows how it is possible to extract the cat’s skeleton for a detailed analysis of the mummy.

92

Figure 6. (a) Photograph of a cat-shaped coffin (Archaelogical Museum of Bologna). (b) 3D reconstruction (voxel side ~600µm). (c) Extraction of the cat skeleton, shown in white, from the volume. whole

Transportable CT system for large objects. As valuable works of art can be hardly moved outside the museum in which they are located, there is a strong interest in the development of CT systems specifically designed for Cultural Heritage analysis on-site. In order to fulfill this requirement our research group developed the transportable CT system shown in Figure 7, which has been conceived also for the investigation of large objects.

Figure 7. Scheme of the transportable CT system: the X-ray tube is on the right, the investigated object at the center and the detector on the left.

93

The design of this system is very different from any commercial available CT machine. It consists of a portable 200 kVp X-ray source, a detector and a motorized mechanical structure for moving the detector and the object, in order to collect the required number of radiographic projections necessary for the tomographic reconstruction. The CT detector consists of a 450!450 mm2 scintillator screen (1 mm thick structured cesium iodide), optically coupled to a 2184!1472 cooled CCD camera (Apogee Alta U32). Visible light photons produced by the scintillator are then collected by the CCD camera, equipped with a photographic lens. Thanks to a 45° mirror, the camera is not placed directly on the primary beam for not damaging it and for decreasing the noise, due to the direct interaction of X-rays with the CCD. The CCD camera, the mirror and the scintillator are positioned in a light-tight box, mounted on a horizontal translation axis. The object is fixed over a rotating plate that is placed at a certain distance along the source to detector axis. If the size of the sample is larger than the Field Of View (FOV) of the detector, it is possible to move the detector along the horizontal axis and to translate the object in the vertical direction by means of a lifter. Thanks to a collaboration with the “Opificio delle Pietre Dure”, the most important Italian restoration institute in Florence, it was possible to transport and mount the CT system inside the shielded laboratory of the “Opificio”, where CT scans were performed on several works of art under restoration in the institute. One particularly interesting case is the famous panel painting “The Goldfinch Madonna (Madonna del Cardellino)” by Raffaello [11]. In Figure 8 a digital radiograph and a 3D tomographic reconstruction are shown.

94

Figure 8. (a) Digital X-ray radiograph that shows the presence of: 1) nails used to assemble the pieces of the painting; 2) wooden inserts; 3) breaks of the wood with lacks of the original painted layers. (b) 3D tomographic reconstruction showing the discontinuities in the painted surface.

Another interesting tomographic analysis regards the big globe ended in 1571 by the Dominican monk Egnazio Danti and located in Palazzo Vecchio, at Florence, Italy (Figure 9). Within the restoration project (sponsored by the Municipality of Florence), a CT of the globe was achieved, for exploring the nature and the conditions of the inner structure [1], [12]. The main problem of getting a complete CT was related to the large size of this masterpiece (220 cm in diameter) and to the need of achieving an in situ analysis in a museum with a

95

lot of visitors. For these reasons an ad hoc experimental apparatus was realized and set-up at Palazzo Vecchio. The diagram of the CT system is shown in Figure 10. The 3D CT reconstruction of the globe has clearly shown the entire inner structure that was never seen before (Figure 11), how it was deformed during time, how it could be restored. All the inner structure, made of iron with a total weight of about 350 kg, was estimated from the segmented 3D reconstruction.

Figure 9. The globe in the Room of Maps (Sala delle carte) in Palazzo Vecchio (Florence)

96

Figure 10. Drawing of the used CT system .

Figure 11. Tomographic reconstruction of the globe internal structure.

97

6. Conclusions

Digital radiography and computed tomography are two new and interesting fields of non-destructive evaluations and a tool for scientific investigations. It should however be pointed out that the CT technique is difficult and expensive. In fact, for having good CT images, many hundred of radiographies are necessary with the use of very expensive equipment for moving the objects with a precision of a few microns. The easiness with which CT can be performed in the medical field may be misleading: medical CT was optimized for the human body (composed mainly of water) and cannot be successfully used on bodies with different density. In order to perform good, non-destructive evaluations, the most suitable digital radiography or computed tomography system (source, moving equipment, detector and elaboration software) must be carefully chosen to avoid obtaining meaningless results. Different digital systems have been developed at the Physics Department of Bologna University for tomographic analysis of Cultural Heritage samples or for industrial applications. The developed systems have been tested on various objects. With a field of view of approximately 3 centimeters, a resolution of 30 microns has been obtained with the microtomographic instrument in the study of archeological teeth or with rock samples; while a resolution of 200÷300 microns is achievable by means of the intensified tomographic system investigating objects of medium size and density such as in the case of the Egyptian mummy. As valuable works of art can be hardly moved outside the museum in which they are located, a transportable CT system for large objects has been realized. This instrument has been tested several times: at Opificio delle Pietre Dure (Florence), at Royal Palace of Venaria Reale, at Palazzo Vecchio (Florence) and National Museum of Asmara (Eritrea) where a fossil skull (one million year old) was analysed. Last but not least, we studied and developed an intensified multi-slice linear detector (5600 ! 35 pixels) that can be arranged either in digital radiography mode or in computed tomography mode by means of a high-precision translation and rotary mechanism. The system is able to make digital radiographies using a dose equal to about 1/100 of those usually used, achieving very high resolutions (about of 22 microns with a 20 centimeters field of view [13]). This instrument has been tested on an industrial component and on a trunk of a tree. Here, growth rings are clearly visible in the reconstruction. Our experimentations show the high success of computed tomography applied to Cultural Heritage. Our studies strongly encourage to proceed with the researches.

98

7. REFERENCES [1] – “X-ray and Neutron Digital Radiography and Computed Tomography for Cultural Heritage”, Franco Casali, Physical Techniques in the Study of Art, Archaeology and Cultural Heritage, Chapter 2, Vol. 1 (2006) 41- 123. [2] – “High Resolution Computed Tomography for Industrial Applications based on Coherent Fiber Optics Ribbons”, M.Bettuzzi, R.Brancaccio, A.Berdondini, M.P.Morigi, F.Casali, A.Flisch, A.Miceli, Proceedings of 5th World Congress on Industrial Process Tomography, pp. 958-964, 3rd-6th September 2007, Bergen, Norway [3] – "A new system for Digital Radiography and Computed Tomography using an intensified linear array detector", F.Casali, A.Pasini, M.Bettuzzi, R.Brancaccio, S.Cornacchia, M.Giordano, M.P.Morigi, D.Romani, Proceedings of International Symposium on Computed Tomography and Image Processing for Industrial Radiology, Berlin, June 2003, pp 317–324, BB 84–CD. [4] – “High resolution X–ray analysis of a proximal human femur with synchrotron radiation and an innovative linear detector”, M.Bettuzzi, R.Brancaccio, F.Casali, S.Cornacchia, E.Di Nicola, N.Lanconelli, L.Mancini, M.P.Morigi, A.Pasini, D.Romani, A.Rossi, IEEE Nuclear Science Symposium and Medical Imaging Conference, Roma,16–22 Ottobre 2004, Nuclear Science Symposium Conference Record, 2004 IEEE Volume 5, 16–22 Oct. 2004 Page(s):3312 – 3315 [5] – “An experimental micro-CT system for X-ray NDT”, M.Rossi, F.Casali, M.Bettuzzi, M.P.Morigi, D.Romani, S.Golovkin, V.Govorun, Proceedings of SPIE's 46th Annual Meeting (San Diego, California USA, 29 July- 3 August), SPIE, USA, 2001. [6] – "Development of high resolution X–ray DR and CT systems for non medical applications", F.Casali, M.Bettuzzi, R.Brancaccio, S.Cornacchia, M.Giordano, M.P.Morigi, A.Pasini, D.Romani, F.Talarico, Proceedings of International Symposium on Computed Tomography and Image Processing for Industrial Radiology, Berlin, 23–25 June 2003, pp.329–336, BB 84–CD. [7] – “Analisi di reperti fossili mediante microtomografia computerizzata.”, M.Rossi, F.Casali, D.Romani, L.Bondioli, R.Macchiarelli, L.Rook, II Convegno Nazionale di Archeometria (Bologna, 29 January-1February) Patron Editore, Bologna, Italy, 2001, p. 91-98. [8] – “Advanced methods in human osteodental paleobiology; The 'Isola Sacra Project'”, L.Bondioli, R.Macchiarelli, in International Symposium Humans from the Past: Advancement in Research and Technology, Roma, Italy, 1997. [9] – “Digital radiography using an EBCCD-based imaging device”Rossi M., Casali F., Golovkin S.V., Govorun V.N., Applied Radiation and Isotopes 53 (2000) 699-709.

99

[10] – “Computed Tomography of an egyptian cat-shaped coffin with mummy”, M.Rossi, F.Casali, D.Romani, D.Picchi, Proceedings of the “5th Internationa Topical Meeting on Industrial Radiation and Radioisotope Measurement Applications” (Bologna, 9-14 giugno 2002) J.E.Fernandez e A.Tartari, Editrice Compositori, Bologna, 2002 [11] – Capitolo “Indagine tomografica”,Franco Casali, Maria Pia Morigi, Matteo Bettuzzi,, Rosa Brancaccio, Irene Bernabei, Andrea Berdondini,Vincenzo D’Errico nel Volume "Raffaello: il colore rivelato ..." a cura di Marco Ciatti, Cecilia Frosinini, Antonio Natali, Patrizia Rintano attualmente in corso di stampa. [12] – “X–ray computed tomography of an ancient large globe”, F.Casali, M.Bettuzzi, D.Bianconi, R.Brancaccio, S.Cornacchia, C.Cucchi, E.Di Nicola, A.Fabbri, N.Lanconelli, M.P.Morigi, A.Pasini, D.Romani, A.Rossi, Optical Methods for Arts and Archaeology Conference, edited by Renzo Salimbeni, Luca Pezzati, 13–14 June 2005, Munich, Germany. Journal: Optical Measurement Systems for Industrial Inspection IV. Edited by Osten, Wolfgang; Gorecki, Christophe; Novak, Erik L. Proceedings of the SPIE, Volume 5857OV–1, pp. 253–260 (2005). [13] – “High resolution X–ray analysis of a proximal human femur with synchrotron radiation and an innovative linear detector”, M.Bettuzzi, R.Brancaccio, F.Casali, S.Cornacchia, E.Di Nicola, N.Lanconelli, L.Mancini, M.P.Morigi, A.Pasini, D.Romani, A.Rossi, IEEE Nuclear Science Symposium and Medical Imaging Conference, Roma,16–22 Ottobre 2004, Nuclear Science Symposium Conference Record, 2004 IEEE Volume 5, 16–22 Oct. 2004 Page(s):3312 – 3315

100

COSMIC RAYS FOR ARCHAEOLOGY GIANROSSANO GIANNINI† Physics Department, University of Trieste, Via Valerio 2, Trieste, 34127, Italy Cosmic rays are the natural phenomenon due to elementary particles reaching from galactic space the Earth atmosphere. Cosmic primary particles, mainly protons, produce showers containing many secondary particles of which some, called muons, are able to propagate to ground and even penetrate underground. By using suitable detectors deployed underground a sort of density variations radiography, related to buried structures of archaeological interest, can be obtained. Measurements with instrumentation built on purpose by a collaboration between physicists of the Universities and the INFN Sections of Trieste and Perugia have been performed at the Arcaeological sites of the Aquileia Roman Port and at the Claudus and Trajan Emperors Ports near Fiumicino-Rome, Italy.

1. Cosmic Rays Radiography A group of physicists of the Universities and the INFN Sections of Trieste and Perugia have developed a procedure called Muon Ground Radiography (MGR) which exploits the natural phenomenon of Cosmic Rays to obtain radiographic type images of underground buried structures of archaeological interest. The detectors for measuring muon intensity and angular distribution underground have been designed, built and operated in several archaological sites of which the experiences at the Ancient Roman Ports remains in Aquileia and at the Claudius and Trajan port area in Fiumicino near Rome are here reported. 1.1. Cosmic Rays from Galactic Space to Earth The Earth atmosphere is continuously reached by energetic elementary particles, mostly protons but also other heavier nuclei are present, from the galactic space in which they are accelerated to reach almost the speed of light. Primary particles, with a flux of about 1000 per second per square meter, upon hitting atmospheric atoms nuclei, produce showers, with many secondary

†

Work partially supported by Italian Nuclear Physics National Institute INFN-Trieste, and Regione Friuli Venezia Giulia L.R. n.3/1998, art.16, 2002.

101

particles of which the most penetrating ones are called muons (Figure 1) and are able to attain sea level and even continue underground or in thick materials.

Figure 1. Galactic Cosmic Rays produce in the atmosphere showers with penentrating muons

The muon flux, of about 100 per second per square meter, will be attenuated depending on the amount of mass traversed. Measuring muon flux versus direction allows to determine the amount of material traversed above. 1.2. The Cosmic Rays detector Cosmic ray muons detection was achieved by building a device with high energy physics advanced technology based on scintillating fibers and bars with multianode photomultipliers, readout by compact VLSI electronics (Fig.2,3).

Figure 2. Muon Ground Radiography MGR Detector based on scintillating fibers and bars.

102

The detector is housed in a aluminum waterproof cylinder (Fig.3) to be lowered underground in an excavated hole.

Figure 3. Galactic Cosmic Rays produce in the atmosphere showers with penentrating muons

1.3. Underground Radiography The Muon Ground Radiography concept is represented in Figure 4. By measuring the cosmic ray muon flux for few days there is enough statistics to reconstruct different density features in soil like stone, marble or cavities of down to about 10-20 cm in size .

Figure 4. The detector from ~ 15 m underground can radiograph a conical shaped volume above it.

103

2. Cosmic Rays Archaeological Applications The first use of cosmic ray muons to detect density fluctuations in structures for archaeological purposes dates back ~40 years and is due to Luis Alvarez [1] for the search of new hidden chambers in the Chefren Pharaon Pyramid. This pyramid is somehow peculiar for not having two burial chambers like the pyramid of his father Cheope, and the one of his grandfather Snefru, but only one known chamber. This led to the hypothesis of a hidden chamber; and to find it the directional flux of cosmic rays inside the pyramid was measured using a muon detector based on streamer chamber. Considering the shape of the pyramid and its material density the conclusion was for no evidence of hidden chambers. More recently the technique of measuring the directional flux of cosmic muon underground was used in the determination of the shape of the cavity in Grotta Gigante (near Trieste, Italy) [2]. 2.1. Underground Radiography in Aquileia The first site selected by our Trieste-Perugia group was near the Roman Port in Aquileia (Fig.5), the second largest town in Italy at the time of the Roman Empire, known to have large amounts of unexplored archeological ruins.

Figure 5. Applications of comsic ray radiography in Aquileia, Italy

The precise place where to install the detector underground was near an already excavated area on the unexplored side. The decision followed a large area Laserscan from helicopter, covering ~10 km2, from which clearly appeared topographic features related to underground archaeological structures (Fig. 6)

104

Figure 6. Laserscan topographic color coded rendering of Aquileia with archeological features.

The Muon Ground Radiography selected site near the Roman Port area with the protection hut for the detector and equipment is shown in Fig. 7.

Figure 7. The roman Port Area and the MGR equipment protection hut.

The data taking performed in the summer 2003 allowed to “see”, buried road sidewalks, column bases and walls (Fig. 8) in agreement with other measurements using Georadar prospection techniques in the same area.

105

Figure 8. Cosmic ray muon ground radiography of the Roman Port in Aquileia Underground Radiography in Fiumicino

The town of Fiumicino near Rome is where the Claudius harbour and Trajan port were the largest ones of the Roman Empire in the Mediterranean Sea (Fig.9).

Figure 9. Some historical images of the port structures now mostly buried under Fiumicino.

The old structures are today almost completely buried underground and the most advanced exploration techniques are tried to find archaeological remains. Our group has performed exploration activities in 2004 with Laserscan,

106

Ortophoto (Fig. 10) on a wide area (~15 km2), together with Georadar and cosmic ray muon radiography data taking in selected zones.

Figure 10. Ortophoto of the Fiumicino area showing the Trajan lake once the Trajan Port.

In the selected place a ~15 m deep hole was excavated (Fig.11) for lowering the MGR detector (fig.12)

Figure 11 . The excavation of the hole for installing the MGR detector underground.

The position for installing the detector was chosen because it was still unexplored but, being higher by ~ 5 m with respect to the surroundings (its even known as “Mount” Giulio), is most probably hiding interesting archaeological remains corresponding to commercial activities near the port as goods storage, water tank replenishing reservoirs for ships or cisterns.

107

Figure 12. The MGR detector being lowered in the hole underground.

The large amount of data collected is still being analyzed by the Archaeological Authorities and the cosmic ray muon radiography method

Figure 13. The Cosmic ray muon radiography plot superimposed to georadar measurements.

worked well and allowed to indicate the position of buried walls near a cistern, but much more is expected to be obtained. References 1. 2.

L. Alvarez et al., Science Vol. 167 (1970). E. Caffau et al., Nucl. Instr. and Meth. A 385, 480-488 (1997).

108

SOME EXAMPLES OF EXAMINATION, CHARACTERISATION, ANALYSIS & CONSERVATION TECHNIQUES DEDICATED TO ARCHAEOLOGICAL ARTEFACTS JEAN LOUIS BOUTAINE Centre de Recherche et de Restauration des Musées de France (C2rmf) Paris – France The examination, characterisation and analysis of archaeological artefacts requires various complementary techniques, in order to improve our knowledge concerning their component materials, their elaboration processes, their evolution and/or degradation according time, use and environment and thus to give indications for their restoration and/or conservation. The present paper will give some guidelines relative to the importance and the usefulness of such techniques. This will be illustrated by some examples taken among recent works on artefacts from various geographical origins, various ages and different materials. Moreover, some examples of techniques of consolidation of archaeological artefacts will also be presented. A comprehensive bibliography is attached.

1. Why science & technology for cultural heritage? The problems to be solved can be of one or other of the seven following types: 1. Determination of the nature of the component materials of an artefact 2. Dating 3. Determination of the creative process of a material or of the artefact itself 4. Evaluation of the suffered alteration processes and estimation of their importance 5. Diagnosis of previous modifications or restorations 6. Assistance to the conservator/restorer 7. Forecasting and optimisation of the short and long term destiny in the present conservation conditions (i.e. preventive conservation) To resolve one or several of these issues the conservation scientist and the conservator need a palette of non destructive and non invasive techniques of examination, characterisation or analysis of archaeological artefacts and their conservation, in order to improve our knowledge concerning their component materials, their elaboration processes, their evolution and/or degradation

109

according time, use and environment and thus to give indications for their restoration and/or conservation. 2. Some preliminary remarks Deriving from my professional experience, firstly, in an applied physics laboratory working for cultural heritage institutions, and secondly, as head of the Research Department of the Centre for Research & Restoration of the Museums of France (C2RMF), I want to insist on some points which are not often mentioned in the community of conservation science: 2a Necessity of crossing results from methods based on different physical or chemical processes. Due to the broad diversity of materials, and as the artefacts have often various, complex and undetermined compositions, as their elaboration processes are often unknown or at least uncertain, it is generally useful or necessary to combine various examination, characterisation and analysis methods, in order to get pertinent information (Ciliberto [1], Janssens [2], Creagh [3], van Grieken [4], Pollard [6]). 2b Cutting edge technologies must not conceal classical everyday work techniques: Two examples: Digital photography During a conference [8] in Malacca (Malaysia) in 2004, a very interesting paper was presented, relative to the potential of a commercially available digital camera, modified in order to make also infrared photography. This permitted to extend the range of the useful wavelength up to 1000 / 1100 nm. As the performances (sensors resolution and zoom lens range) are regularly improving in commercially available digital photographic cameras, such a protocol could be largely used. This is an excellent illustration of “low cost technology” adapted to cultural heritage examination (infrared photography is an important tool in this area). Alternatively, at the C2RMF (Paris), recently (since December 2003), a new development relative to a digital multi-spectral photography protocol occurred [9] The equipment and the protocol adopted permit one to realise sequentially, with the same operating conditions: classical photography, infrared photography, UV fluorescence photography and raking light photography. The equipment consists in a Hasselblad H1 type still digital camera, auto-focus with adapted lens (f = 80 mm), Imacon CCD detector 4000 x 5000 pixels, practical

110

equivalent sensitivity up to 200 ISO, useful wavelength ! " 1050 nm (N.B.: for silver halide films, ! " 900 to 1000 nm), used with a video monitor. Such an equipment is used for laboratory or in situ examination, for instance: paintings of the Galerie d’Apollon, Musée du Louvre (Paris) before restoration. In this case the sketch was 12 m in length, and distance from object to camera: 25 m. For UV fluorescence one uses a classical flashlight without protective cache, with 3 to 5 flashes. For IR photography one uses a filter transparent to infrared, the sensor being modified on C2RMF request, with the infrared absorbing filter being dismounted, and set on demand, outside the camera. X-ray fluorescence analysis One must not forget commercially available portable XRF devices and focus on PIXE or synchrotron radiation techniques to realise elemental analysis of cultural heritage components. 2c Nobody is in a position of mastering the more suitable techniques for various sets of cultural heritage artefacts, so there is a necessity of working in networks. Here are some successful examples of European networks active in this area: Progetto Finalizzatto « Beni Culturali » - (CNR Italy) http://www.area.cs.cnr.it/cnr/pf/beni.html CHIMART (CNRS France) http://www.c2rmf.fr/pages/page_id18509_u1|2.htm Red Tematica de Patrimonio Historico y Cultural (CSIC – Spain) http://www.rtphc.csic.es/ COST G7 http://alpha1.infim.ro/cost COST G8 http://www.srs.dl.ac.uk/arch/cost-g8 ENVI-ART - COST D42 http://www.echn.net/enviart EU-ARTECH http://www.eu-artech.org 2d One must be vigilant of not concentrating all the efforts on famous artefacts and, as a consequence, ignore and/or leave to irremediable degradation less prestigious artefacts collections which could eventually be more significant milestones of the mankind history

111

2e One should also always consider the potential use and/or transfer of technology of examination and analysis techniques developed in technical research centres dedicated to various industrial materials (glass, stone, wood, cast metals, ceramics…) 3. Examination, characterisation, analysis of cultural heritage artefacts Some demonstrative examples of study-restoration-conservation of famous cultural heritage artefacts can illustrate these remarks. All of them were subjects of successful multidiscipline projects: 1.

Satiro danzante di Mazzara del Vallo (Sicilia, Italy) [26] The bronze statue of the "Satiro danzante“ was fished out in the Canale di Sicilia in1998, on a bed of 500 m. It represents a mythological figure, a demon, part of the train of Dionysus, god of the wine. The artefact (2.50 m height) could be an original from the Hellenistic era (IV – III Century BC), or a more recent replica from the end of 1st Century AD. The statue is now kept in the Museo del Satiro, a Mazara del Vallo. (Scientific team; Giorgio Accardo – Istituto Superior per la Conzervazione ed il Restauro - Roma)

2.

Nebra sky disc, Halle (Germany) [27] The Himmelsscheibe von Nebra (sky disc) is a bronze disc of around 300 mm diameter, patinated blue-green and inlaid with gold symbols, excavated in 1999, near Nebra, Sachsen-Anhalt (Germany), dated to c. 1600 BC, associated with the Bronze Age Unetice culture. Copper came from Eastern Alps and gold from Carpathian basin. The disc and its accompanying finds are now in the Landesmuseum für Vorgeschichte of Sachsen-Anhalt - Halle (Germany). (Scientific team: E. Pernicka – TU Bergakademie Freiberg)

3.

Leopards weight of Shahi Tump (Balochistan), National Museum, Karachi (Pakistan) [28] & [29] The artefact was discovered in a grave, in the Kech Valley, in Balochistan, southern part of present Pakistan. It belongs to the Shahi Tump – Makran civilisation (end of 4th millennium – beginning of 3rd millennium BC). Height: 200 mm; weight: 13.5 kg. The shell (e = 3 mm) has been manufactured by lost-wax foundry of a copper alloy (12.6 % Pb, 2.6 % As), then it has been filled up through lead (99.5 %) foundry. The shell is engraved with figures of leopards hunting wild goats, made of polished fragments of shellfishes. No identification of the artefact’s use has been

112

given (Fig. 1). (Scientific team: B. Mille, D. Bourgarit (C2RMF), R. Besenval (Musée Guimet –Paris) 4.

Venus Genitrix, or Aphrodite – Roman marble replica (1st Century AD) of a bronze Greek original (5th century BC) - Louvre Museum [30] & [31] The examination (gamma radiography and UV photography) before restoration has been conducted in order to understand previous restorations (Scientific team: B. Bourgeois (C2RMF – Versailles), B. Rattoni (CEA – Saclay) & D. Bagault (C2RMF – Paris))

5.

Canthare with storks – Boscoreale treasury, close to Pompei, Vesuvio eruption 79 AD, Musée du Louvre - Paris [32] The radiographic examination permits to establish conclusions relative to the manufacturing and the shaping of the artefact: different techniques were used in addition to each other, twofold shell, casting, then hammering, machining, stapling and brazing (Fig. 2). (Scientific team: D. Robcis & T. Borel (C2RMF) – Paris).

4. Some examples of conservation techniques ARC-Nucleart in Grenoble (France) is a co-enterprise CEA – Ministry of Culture – City of Grenoble – Rhône-Alpes Region. This Centre, achieving research and conservation / restoration, masters different techniques of consolidation of archaeological artefacts: sterilisation using gamma irradiation, NUCLEART process based on high energy gamma polymerisation, PEG (polyethylene-glycol) impregnation, lyophilisation. So, the Centre is in a position to choose the more appropriate technique for a given kind of artefact (nature of the materials, state of degradation, size…). The techniques are complementary and can be applied to artefacts made of materials like: dry wood, stone, water logged wood, leather, basketwork [125] to [130]. Some examples can illustrate the large scope of archaeological objects which can be treated by one of this palette of conservation techniques (Fig. 3, 4 & 5). 5. Conclusion After this very brief summary of the applications of science & technology to the study, the restoration and the conservation of cultural heritage artefacts, the reader can take benefit of the following wide list of references.

113

References 1. E. Ciliberto, G. Spoto, Modern analytical methods in art and archaeology, Wiley Interscience, New York (2000) 2. K. Janssens, R. Van Grieken, Non-destructive micro analysis of cultural heritage materials, Elsevier, Amsterdam (2004) 3. D. Bradley, D. Creagh, Physical techniques in the study of art, archaeology and cultural heritage, Elsevier, Amsterdam (2006) 4. J.L. Boutaine, The modern museum, chapter 1 in D. Bradley [Ed] [3], 1-39 5. R. van Grieken, K. Janssens, Cultural heritage conservation and environment impact assessment by non-destructive techniques and microanalysis, Taylor & Francis, London (2004) 6. M. Pollard, C. Batt, B. Stern, S.M.M. Young, (2007), Analytical chemistry in archaeology, Cambridge University Press (2007) 7. M.C. Wiescher, Physics methods in art and archaeology, http://www.nd.edu/~nsl/Lectures/phys178/index.htm 8. M. Ashhar Hj Khalid, Proc. First national conference on science and technology in the preservation of national heritage, Malacca (Malaysia), 1619 Aug 2004 9. D. Vigears (C2RMF Paris) private communication (2004) 10. R. Salimbeni, L. Pezzati, Optical methods for art and archaeology, SPIE proceedings series, vol. 5857, Munich (Germany), 13-14 Jun 2005 11. M. Hours, La radiographie des oeuvres d'art, in Idées et Découvertes, Encyclopaedia Universalis, Paris, 240-243 (1978) 12. M. Hours, Les secrets des chefs d'oeuvre, R. Laffont, Paris (1964) 13. A. Gilardoni, L. Ascani Orsini, R. Ascani Orsini, M. Taccani Gilardoni & al., X-rays in art - I raggi X nell'arte (2nd edition), Gilardoni, Mandello Lario (1984) 14. D. Graham, T. Eddie, X-ray techniques in art galleries and museums, A. Hilger, Bristol (1985) 15. J. Lang, Radiography of cultural materials, Butterworth-Heinemann, London (1997) 16. J.L. Boutaine, T. Borel, E. Ravaud, La radiographie au service de l'étude des oeuvres du patrimoine culturel, Proc. 3ème séminaire franco-allemand sur le contrôle non-destructif - COFREND – DGZfP, Forbach (France), 910 Sept 1999 17. B. Bourgeois, J.L. Boutaine, B. Rattoni, Radiographie et restauration de marbres antiques : l'exemple du Musée du Louvre, Proc. ICOM-CC 11th Triennal Meeting Edimburgh, 793-797 (Sept 1996) 18. V. Fell, Q. Mould, R. White, Guidelines on the X-radiography of archaeological metalwork, http://www.englishheritage.org.uk/upload/pdf/X_Radiography.pdf 19. T. Borel, Applications de la radiographie X et à l'étude des objets métalliques d'art et d'archéologie. Emissiographie, in “ A la recherche du métal perdu ” Catalogue de l’exposition de Guiry en Vexin (France), 54-60 (1999)

114

20. O. Sicardy, Application de quelques techniques de contrôle non destructif à la caractérisation des roches calcaires, Thesis, Ecole Centrale Paris (July 1985) 21. P. Vandiver, W. Ellingson, T. Robinson, New applications of Xradiological imaging technologies for archaeological ceramics, Archaeomaterials, 5, 2, 185-207 (1991) 22. A. Ravaglioli, A. Krajewski, Impiego di metodologie tecnico scientifiche nel campo dei reperti ceramici, Museo internazionale delle ceramiche, Faenza (Italy) (1981) 23. M.M. Brooks, S.A. O’Connor, Looking in the past: the potential of Xradiography as an investigative technique for archaeological and ethnographic textiles, in Proc. Recovering the past: the conservation of archaeological and ethnographic textiles, NATCC, Mexico City, 87-97 (2005) 24. P.A. Ruault, Radiologie industrielle (2 vol.), Publications de la Soudure Autogène, Paris (1991) 25. R. Halmshaw, Industrial radiology: theory and practice (2nd edition), Chapman & Hall, London (1995) 26. Anonymous - ICR, Il Satiro danzante da Mazara del Vallo, il restauro e l'immagine, Proc. Convegno « Dal mare al museo », Roma, 3-4 Jun 2003, R. Petriaggi [Ed], Electa, Napoli (2005) 27. E. Pernicka, M. Radtke, H. Riesemeier, C.-H. Wunderlich, European Network of Competence at 1600 BC, Bessy - Highlights 2003 - Scientific highlights http://srs.dl.ac.uk/ARCH/what-the-papers-say/Nebradisk/BESSY_HL01_pernicka.pdf 28. B. Mille, R. Besenval, D. Bourgarit, Early lost-wax casting in Balochistan (Pakistan): the " Leopards Weight " from Shahi-Tump. in Persiens antike Pracht, Bergbau-Handwerk-Archäologie, T. Stöllner, R. Slotta, A. Vatandoust, A. (Ed.), p. 274-280. Bochum: Deutsches Bergbau Museum (2004) 29. B. Mille, D. Bourgarit, R.Besenval, 2005. Metallurgical Study of the "Leopards Weight" from Shahi-Tump (Pakistan). in South Asian Archaeology 2001, C. Jarrige, V. Lefèvre, (Ed.), p. 237-244. Paris: Editions Recherches sur les Civilisations (2005) 30. B. Bourgeois, L’Aphrodite, dite “ Vénus Génitrix ” restaurée, Musée du Louvre, Paris, (1999) 31. B. Rattoni (CEA – Saclay), private communication (1999) 32. D. Robcis, T. Borel (C2RMF Paris), private communication (2007) 33. D.C. Creagh, D.A. Bradley, Radiation in art and archaeometry. Elsevier, Amsterdam (2000) 34. G. Demortier, A. Adriaens, Ion beam study of art and archaeological objects. A contribution by members of the COST G1 action, Report EUR 19218 (2000) 35. V.P. Guinn, Past, present, future of neutron activation analysis, Jl. of radioanalytical and nuclear chemistry, 160, 1, 9-19 (Jul 1992)

115

36. R.J. Speakman, M.D. Glascock, Acknowledging fifty years of neutron activation analysis in archaeology, Archaeometry, 49, 2, 179-183 (2007) 37. Anonymous – University of Missouri, Database on neutron activation analysis in archaeometry, http://archaeometry.missouri.edu/datasets/datasets.html 38. R.L. Bishop, M.J. Blackman, Instrumental neutron activation analysis of archaeological ceramics: scale and interpretation, Acc. Chem. Res., 35, 8, 603-610 (2002) 39. L. Moens, P. Roos, J. de Rudder, J. Hoste & al., White marble from Italy and turkey: an archaeometric study based on minor and trace element analysis and petrography, Jl. of radioanalytical and nuclear chemistry, 123, 1, 333-348, (1988) 40. M.D. Glascock, H. Neff, Neutron activation analysis and provenance research in archaeology, Meas. Sci. Technol., 14, 1516-1526 (2003) 41. R.B. Firestone, H.D. Choi, R.M. Lindstrom, G.L. Molnar & al., Database of prompt gamma rays from slow neutron capture for elemental analysis, Report LBNL-55199 (2004) 42. G. Szakmany, Z. Kasztovsky, Prompt gamma activation analysis, a new method in the archaeological study of polished stone tools and their raw materials, European Jl. of mineralogy, 16, 2, 285-295 (2004) 43. Y. Oura, A. Saito, K. Sueki, & al., Prompt gamma-ray analysis of archaeological bronze, Jl. of radioanalytical and nuclear chemistry, 239, 3 (1999) 44. H. Postma, J.J. Butler, P. Schillebeeckx, C.W.E. van Eijk, Neutron resonance capture applied to some prehistoric bronze axes, Proc. Workshop RICH – Research infrastructures for cultural heritage, Trieste (Italy) 12-13 Dec 2005, Il Nuovo Cimento, 30C, 1, 105-112 (2007) 45. Z. Kasztovszky, M. Mackowiak de Antczak, & al., Provenance study of Amerindian pottery figurines with prompt gamma activation analysis, Nukleonika, 49, 3, 107!113 (2004) 46. Z. Kasztovszky, D. Visser, W. Kockelmann, E. Pantos & al., Combined prompt gamma activation and neutron diffraction analyses of historic metal objects and limestone samples, Proc. Workshop RICH – Research infrastructures for cultural heritage, Trieste (Italy) 12-13 Dec 2005, Il Nuovo Cimento, 30C, 1, 67-78 (2007) 47. L. Bellot-Gurlet, G. Poupeau, J. Salomon, T. Calligaro & al., Obsidian provenance studies in archaeology: a comparison between PIXE, ICP-AES and ICP-MS, Proc. Eighth European conference on accelerators in applied research and technology (ECAART-8), Nucl. Inst. Methods, 240, 1-2,583588, (2005) 48. M.F. Guerra, Analysis of archaeological metals. The place of XRF and PIXE in the determination of technology and provenance, X-ray spectrometry, 27, 2, 73-80 (1998) 49. P. Kuisma-Kursaula, Accuracy, precision and detection limits of SEMWDS, SEM-EDS and PIXE in the multi-elemental analysis of medieval glass, X-ray spectrometry, 29, 1, 111-118 (2000)

116

50. E.S. Lindgren, Special millennium issue on cultural heritage, X-ray Spectrometry, 29, 1 (2000) 51. J. Kunicki-Goldfinger, J. Kierzek, A. Kasprzark, B. Malzewska-Bucko, A study of eighteenth century glass vessels from central Europe by X-ray fluorescence analysis, X-ray spectrometry, 29, 4, 310-318 (2000) 52. H. Bronk, S. Röhrs, A. Bjeoumikhov, N. Langhoff & al., ArtTAX – A new mobile spectrometer for energy-dispersive micro X-ray fluorescence spectrometry on art and archaeological objects, Fresenius Jl. of analytical chemistry, 371, 3, 307-316, (2004) 53. P. Wobrauschek, G. Halmetschlager, S. Zamini & al., Energy-dispersive Xray fluorescence analysis of Celtic glasses, X-ray spectrometry, 29, 1, 25-33 (2000) 54. H. Ida, J. Kawai, Identification of glass and ceramics by X-ray fluorescence analysis with a pyroelectric X-ray generator, Analytical sciences, 20, 8, 1211 (2004) 55. H. Kristmanson, The application of X-ray fluorescence analysis to archaeological samples. A case study from Newfoundland and Cape Breton Island, Canadian Jl. of archaeology, 28, 2, 342-352 (2004) 56. R. Cesareo, G.E. Gigante, A. Castellano & al., Portable systems for energy dispersive X-ray fluorescence analysis of works of art, Jl. of trace and microprobe techniques, 14, 4, 711-725, (1996) 57. T. Pantazis, A. G. Karydas,, Chr. Doumas, A. Vlachopoulos, P. Nomikos, M. Dinsmore, X-ray fluorescence analysis of a gold Ibex and other artefacts from Akrotiri, Proc. 9 th International Aegean Conference – Metron, Measuring the Aegean Bronze Age 58. A. Gianoncelli, P. Walter, E. Dooryhée, L. Ortega, P. Bordet, J. Castaing, Energy dispersive X-ray diffraction in contradistinction to angle dispersive methods, 12th European Conference on X-ray spectrometry – EXRS – Paris; June 2006 59. A. Gianoncelli, J. Castaing, A. Bouquillon, A. Polvorinos, P. Walter, Quantitative elemental analysis of Della Robbia’s glazes with a portable XRF spectrometer and its comparison to PIXE methods.,12th European Conference on X-ray spectrometry – EXRS – Paris; June 2006, X-ray Spectrometry, 35 (2006) 365-369 60. C. Miliani et al., Description of EU-ARTECH ACCESS MOLAB executed projects, http://www.euartech.org/index.php?option=com_content&task=view&id=44&Itemid=95 61. T. Calligaro, J.C. Dran, M. Dubus, B. Moignard & al., Les objets de musée sous le projecteur d’AGLAE, Techné, 13-14, 41-48 (2001) 62. J.C. Dran, T. Calligaro, J. Salomon, Particle-induced X-ray emission, in Modern analytical methods in art and archaeology, Wiley Interscience, New York,135–166 (2000) 63. T. Calligaro, J.C. Dran, B. Moignard, L. Pichon & al., Ion beam analysis with external beams: recent set-up improvements, Proc. ECAART 7, Guildford, (2002), in NIM-B, 188 135-140 (2002)

117

64. T. Calligaro, J.P. Poirot, G. Querré, Trace element fingerprinting of jewellery rubies by external beam PIXE. NIM-B, 150, 628-634 (1999) 65. T. Calligaro, J.C. Dran, J.P. Poirot, J. Salomon, PIXE/PIGE characterisation of emeralds using an external micro-beam. Proc. IBA 14 Dresden (Jul 1999). in NIM-B, 161-163, 769-774 (2000) 66. M.F. Guerra, T. Calligaro, J.C. Dran., C. Moulherat, J. Salomon, Development of a PIXE, PIGE and PIXE-XRF combination for the analysis of the gold from the First Empire of the Steppes, in H Hars & E Burke [Ed], Geoarchaeological and Bioarchaeological Studies. Amsterdam 67. J.C. Dran et al., Description of EU-ARTECH ACCESS AGLAE executed projects, http://www.euartech.org/index.php?option=com_content&task=view&id=43&Itemid=94 68. J. Opitz-Coutureau, A. Denker, C. Couzon & al., Zerstörungsfreie Elementanalyse durch Hochenergie-PIXE an mittelalterlichen „Wiener Pfennigen" und mongolischen Bronze- Fundmünzen aus Karakorum; Proc. Symposium „Numismatics & technology: questions and answers", Kunsthistorisches Museum (25-26 April 2003) Vienna. in: M. Griesser [Ed.] : Numismatik und Technologie: Fragen und Antworten Symposium Proceedings, 49-61 69. E. Ioannidou, D. Bourgarit, T. Calligaro, JC. Dran & al., RBS and NRA with external beams for archaeometric applications. Proc. IBA 14 Dresden (Jul 1999), in NIM-B 161-163, 730-736 (2000) 70. M. Dubus, I. Colson, JC. Dran, E. Ioannidou & al., Elemental and structural analysis of corroded lead seals from the XIIIth. to XVth. centuries: preliminary study. Proc. 6th International Conference on "Non-Destructive Testing and Microanalysis for the Diagnostics and Conservation of the Cultural and Environmental Heritage". Vol. 2. Roma, (17-20 May 1999) 1741-1750 71. M. Dubus, M. Aucouturier, B. Moignard, Atmospheric corrosion monitoring of silver in museums, Proc. Indoor Air quality, Norwich, (27-28 Apr 2003) 72. Anonymous (IAEA), Nuclear analytical techniques in archaeological investigations, Technical reports series N° 416, IAEA, Wien, 2003 73. M.F. Guerra, The study of the characterization and provenance of coins and other metalwork using XRF, PIXE and Activation Analysis. in D.C. Creagh [Ed] [29]. 417-444 74. M.F. Guerra, Ancient gold and modern techniques: identification of gold supplies from Croesus to Mohamed using accelerators and ICP-MS, in "Non-destructive testing and micro-analysis for the diagnostic and conservation of the Cultural and environmental Heritage", Proc. Art 2002, Antwerpen, (Jun 2002) 75. W. Kockelmann, E. Pantos, Synchrotron X-ray and neutron diffraction study of metal artefacts from the Poulton archaeological site, Proc. "The use of ion beams in material sciences, medicine and archaeometry", Namur, Belgium. May 22-24, 2003, http://srs.dl.ac.uk/arch/posters/namur-poster

118

76. E.Pantos, T.Pradell, N.Salvadó, J.Molera & al., Synchrotron radiation and neutrons in archaeological and cultural heritage science, Proc. International School of Physics "Enrico Fermi", Varenna, Italy, (2004), IOS Press, 299307 77. I. De Ryck, A. Adriaens, F. Adams and E. Pantos, A comparison of microbeam techniques for the analysis of corroded ancient bronze objects, Analyst, 128, 8, 1104-1109 (2003) 78. M. Müller, M.Z. Papiz, D.T. Clarke, M.A. Roberts & al., Identification of textiles from the Khirbet Qumran caves using microscopy and synchrotron radiation x-ray fibre diffraction. in Khirbet Qumran et Ain Feshkha, II. Etudes d'anthropologie, de physique et de chimie, Fribourg (Suisse), JB; Humbert & J. Gunneweg (Eds.), Academic Press - Saint-Paul (2003) 79. W. Kockelmann, E.Pantos, and A. Kirfel, Neutron and synchrotron radiation studies of archaeological objects in D.C. Creagh [Ed] [29] 80. I. Reiche, T. Calligaro, M. Menu, J. Salomon, C. Vignaud C. Comparative analysis of odontolite, heated fossil ivory and blue fluorapatite by PIXE/PIGE and TEM. Proc. IBA 14 Dresden (Jul 1999), in NIM-B 161163, 737-742 (2000) 81. I. Reiche, M. Menu, C. Vignaud, Heat induced transformation of fossil mastodon ivory into turquoise "odontolite". Structural and elemental characterisation. Solid States Sciences, 2, 6, 625-636 (2000) 82. S. Siano, W. Kockelmann, L. Bartoli, Archaeometallurgical studies of bronze artefact, ISIS Report 12605 (July 2003) 83. S. Siano, W. Kockelmann, L. Bartoli, M. Zoppi & al., Characterisation of archaeological bronze by time of flight neutron diffraction, proc 33rd Intern. Symposium on Archaeometry Amsterdam (2004) 84. W. Kockelmann, A. Kirkel, E. Jansen, R. Linke & al., Neutron diffraction for non-destructive texture analysis of minted and cast “Taler” coins, Proc. Numismatics & technology questions & answers, Kunsthistorisches Museum, Wien (2003) 113-123 85. C. Andreani, G. Gorini, A. Filabozzi, E. Perelli-Cippo & al., Texture and structure studies on marbles from Villa Adriana, Proc. 1st International workshop on science, technology & cultural heritage, Venice, (June 29-July 1 2004), AIV 86. A. Tsolakidou, V. Kilikoglou, Comparative analysis of ancient ceramics by neutron activation analysis, inductively coupled plasma optical emission spectrometry, inductively coupled plasma mass spectrometry and X-ray fluorescence, Analytical and bioanalytical chemistry, 374, 3, 566572,(2004) 87. S. Shackley, Sources of archaeological obsidian in the greater American southwest: an update and quantitative analysis, American antiquity, 60, (1995) 88. W. Häusler, Firing of clays studied by X-ray diffraction and Mössbauer spectroscopy, Hyperfine interactions, 154, 1-4, 121-141 (2004)

119

89. B. Mille, D. Bourgarit, L’analyse des alliages anciens à base de cuivre : état des connaissances et développement d’un protocole d’analyse par ICP-AES, Revue d’archéométrie 24, 13-26 (2000) 90. D. Bourgarit, B. Mille, The elemental analysis of ancient copper-based artefacts by Inductively–Coupled-Plasma Atomic-Emission-Spectrometry (ICP-AES): an optimized methodology reveals some secrets of the Vix Crater. Measurement Science and Technology, 14, 1538-1555 (2003) 91. B. Mille, D. Bourgarit, J.F. Haquet, R. Besenval, From the 7th to the 2nd millennium BC in Balochistan (Pakistan): the development of copper metallurgy before and during the Indus Civilisation, South Asian Archaeology 2001, C. Jarrige & V. Lefevre [Ed], Editions Recherche sur les Civilisations, Paris (2005) 92. T.E. Jeffries, Laser ablation inductively coupled plasma mass spectrometry, in K. Janssens [Ed]. Non-destructive micro analysis of cultural heritage materials, Elsevier, Amsterdam (2004) 93. V. Lazic, R. Fantoni, F. Colao, A. Santagata, A. Morone, Quantitative laser induced breakdown spectroscopy of ancient marbles and corrections for the variability of plasma parameters and of ablation rate, J. Anal. At. Spectrom. 19, 4,429-436 (2004) 94. V. Lazic, F. Colao, R. Fantoni, A. Palucci & al., Characterisation of lustre and pigment composition in ancient pottery by laser induced fluorescence and breakdown spectroscopy, J. Cult. Heritage 4 (2003) 95. Lazic, F. Colao, R. Fantoni, V. Spizzichino, Recognition of archaeological materials underwater by laser induced breakdown spectroscopy, Spectrochimica Acta part B, 60, 7-8, 1014-1024 (2005) 96. P. Maravelaki-Kalaitzaki, D. Anglos, V. Kilikoglou, V. Zafiropulos, Compositional characterization of encrustation on marble with laser induced breakdown spectroscopy, Spectrochimica Acta B, 56, 887-903 (2001) 97. W. Devos, M. Senn-Luder, C. Moor, C. Salter, Laser ablation coupled plasma mass spectrometry (LA-ICP-MS) for spatially resolved trace analysis of early medieval archaeological iron finds, Fresenius Jl. of analytical chemistry, 366, 8, 873-880, (2004) 98. A. Giakoumaki, K. Melessanaki, D. Anglos, Laser induced breakdown spectroscopy (LIBS) in archaeological science – applications and prospects, Analytical and bioanalytical chemistry, 387, 3, 749-760 (2006) 99. K. Melessanaki, A. Mastrogiannidou, S. Chlouveraki, S.C. Ferrence & al., Analysis of archaeological objects with LMNTI, a new transportable LIBS instrument, Lasers in the conservation of artworks, Springer, Berlin, 443449, (2005), Proc. LACONA V, Osnabrück (Germany) 15-18 Sep 2003 100. R.J.H. Clark, Pigment identification by spectroscopic means: an arts/science interface, Comptes-rendus Chimie, 5, 1, 7-20, (2002) 101. M. Giardino, R. Kuzio, R. Miller, D. Muirhead, Analysis of ceramic color by spectral reflectance, American antiquity, 63 (1998) 102. P. Garside, P. Wyeth, Characterisation of plant fibres by infrared spectroscopy, Polymer Preprints 41 (2000) 2 1792-1793

120

103. P. Garside, P. Wyeth, Using contemporary analytical techniques to characterise natural fibres in historical artefacts, proc. 40th Eastern Analytical Symposium (1-4 Oct 2001) Atlantic City, NJ 104. C. Moulherat, Archéologie des textiles protohistoriques – Exemple de la Gaule celtique, Thesis, University of Paris 1, 2002 105. C. Moulherat, M. Tengberg, J.F. Haquet, B. Mille, First evidence of cotton at Neolithic Mehrgarh, Pakistan. Analysis of mineralised fibres from a copper bead, J. Archaeol. Sci., 29, 12, 1393-1401 (2002) 106. S. Pagès-Camagna, S. Colinart, C. Coupry, Fabrication processes of archaeological Egyptian blue and green pigments enlightened by Raman microscopy and scanning electron microscopy, J. Raman Spectros. 30, 313317 (1999) 107. S. Pagès-Camagna, S. Colinart, The Egyptian green pigment: its manufacturing process and links to Egyptian blue. Archaeometry, 45, 637638 (2003) 108. L. Bussotti, E.Castellucci, M.Picollo, Identification of ancient pigments by Raman microscopy, Proc. Congress: Science and Technology for the Safeguard of the Cultural Heritage in the Mediterranean Basin, Vol. I, 695699 (1995) 109. M. Regert, Productions techniques et résidus alimentaires préhistoriques : les apports de la chimie organique, Techné, 13-14, 71-77 (2001) 110. M. Regert, Investigating the history of prehistoric glues through gas chromatography – mass spectrometry, J. Separation Science, 27, 244-254 (2004) 111. M. Regert M., N. Garnier, O. Decavallas, C. Cren-Olivé, C. Rolando, Structural characterization of lipid constituents from natural substances preserved in archaeological environments. Measurement Science and Technology, 14, 1620-1630 (2003) 112. M. Regert, S. Vacher, C. Moulherat., O. Decavallas, Adhesive production and pottery function during the Iron Age at the site of Grand Aunay (Sarthe, France). Archaeometry, 45, 101-120 (2003) 113. M. Regert, C. Rolando, Identification of archaeological adhesives using direct inlet electron ionization mass spectrometry, Analytical Chemistry, 74, 965-975 (2002) 114. N. Garnier, C. Cren-Olivé, C. Rolando, M. Regert, Characterization of archaeological beeswax by electron ionization and electrospray ionisation mass spectrometry. Analytical Chemistry, 74, 4868-4877 (2002) 115. N. Garnier, Analyse structurale de matériaux organiques conservés dans des céramiques antiques. Thesis, Université de Paris 6. (26 février 2003) 116. M. Regert, S. Colinart, L. Degrand, O. Decavallas, Chemical alteration and use of beeswax through time: accelerated ageing tests and analysis of archaeological samples from various environmental contexts, Archeometry, 43, 4, 549-569 (2001) 117. S. Mirabaud, C. Rolando, M. Regert, Molecular criteria for discriminating adipose fat and milk from different species by nano-ESI-MS and MS/MS of

121

their triacylglycerols: application to archaeological remains, Analytical chemistry, 79, 16, 6182-6192, (2007) 118. M. Regert, Elucidating pottery function using a multi-step analytical methodology combining infrared spectroscopy, chromatographic procedures and mass spectrometry, in: Theory and practice of archaeological residue analysis, [Ed] H. Barnard and J. W. Eerkens. British Archaeological Reports S1650, pp. 61-76 (2007) 119. M. Regert, Produits de la ruche, produits laitiers et matières végétales : quels vestiges pour appréhender les substances naturelles exploitées par l’homme pendant la préhistoire?, in J.P. Poulain [Ed] L'homme, le mangeur et l'animal. Qui nourrit l'autre ?, Les Cahiers de l'OCHA, Paris, 12, 50-64 (2007) 120. F.H. Schweingruber, Trees and wood in dendrochronology, Springer, Berlin (1993) 121. M. Kaennel, F.H. Schweingruber, Multilingual glossary of dendrochronology, Paul Haupt Bern, available at Swiss federal institute for forest, snow and landscape research, Birmensdorf (Switzerland) (1995) 122. J.H. Townsend, K. Eremin, A. Adriaens, Conservation science 2002, proceedings of a COST G8 meeting, Edinburgh (May 2002), Archetype Publishing, London (2003) 123. B. Brunetti, AM. Johansson, Science and technology for the conservation of the European cultural heritage – Research infrastructures, Report EUR 20483 (2003) 124. A. Denker, A. Adriaens, M. Dowsett, A. Giumla-Mair, COST action G8: non-destructive testing and analysis of museum objects, Fraunhofer IRB Verlag, Stuttgart (2006) 125. R. Ramière, Le laboratoire NUCLEART et le Centre d'étude et de traitement des bois gorgés d'eau à Grenoble, in Musées et collections publiques de France, Association générale des conservateurs des collections publiques de France, Paris (1987) 126. Q.K. Tran, R. Ramière, A. Ginier-Gillet, Impregnation with radiationcuring monomers and resins, in Archaeological wood: properties, chemistry, and preservation, 217-233, American Chemical Society, Washington (1990) 127. H. Bernard-Maugiron, A. Ginier-Gillet, X. Hiron, Le traitement des bois humides : bateaux et sarcophages, Les dossiers d’archéologie, 153, 24-31 (1990) 128. Q.K. Tran, X. Hiron, E. Damery, Traitement de la pirogue néolithique P6 de Paris-Bercy : de l’extraction à la conservation muséographique, Proc. 6th triennial meeting of the ICOM-CC working group on waterlogged wood, York (United Kingdom) 9-13 Sep 1996 129. P. Velay, Du chantier au Musée Carnavalet, Archeologia, 370, 18-19 (2000) 130. G. Chaumat, Q.K. Tran, P. Descalle, Densification de répliques en plâtre en vue d'une exposition à l'extérieur, in Le plâtre: l'art et la matière, Créaphis, Paris (2001)

122

Fig 1 - The Leopards weight from Shahi Tump - Photography and 30 MeV accelerator tomodensimetry showing the copper shell and the lead filling

Fig 2 - Canthare with storks – Boscoreale treasury, close to Pompei, Vesuvio eruption 79 AD, Musée du Louvre - Paris – Photography and X-ray radiography showing details of the manufacturing process

123

Fig 3 – Shoe - XIVth Century, leather, Brandes (Isère) lead-silver mine, Musée de l’Alpe d’Huez, consolidation process: lyophilisation – Photography before and after consolidation and restoration

124

Fig 4 – Virgin - XIVth Century, wood, Flavigny sur Ozerain (Côte d’Or), consolidation process: Nucléart

Fig 5 - “Trébuchet” (balance) box - 575-660 AD, yew, boxwood & bronze, 190*70*40 mm, the bronze weights show the effigies of emperor Justinius II (565-578) & empress Sophia, excavated in 1994 from the shipwreck of La Palud I Port-Cros Island, Musée historique de Marseille, consolidation process: lyophilisation

125

PRESENTATION OF DEMGOL: ONLINE ETYMOLOGICAL DICTIONARY OF GREEK MYTHOLOGY E. PELLIZER University of Trieste, Italy

What is known as Western Culture was built over the course of three millennia on the basis of two “Great Codes.” Encoded in the Germanic and Neo-Latin languages that spread over a significant part of the globe, these “Codes” engendered numerous literary and figurative works (as well as philosophies and even theologies) in the traditions of various nations. As shown by Northrop Frye, one of these “Codes” had its origin in the Jewish tradition, in the Torah itself and in the various elaborations of this text, as represented by the writings of the most important and widespread monotheistic religions (the Bible, the Koran, etc.). This “Code” influenced primarily Protestant, German- and English-speaking cultures. The other, however, indubitably goes back to the Greco-Roman tradition, which was preserved and transmitted from antiquity to modern times, endowing Europe with a huge patrimony of fictional tales, both religious and - to use a convenient label “mythical” texts. Today, modern information technology offers students, teachers and scholars unprecedented and unfettered access to this vast store of mythological, literary and historico-religious material in the form of multimedia and multilingual dictionaries and encyclopaedias. Combining ease of use and rapid access, these tools provide an advanced system for exploring the semiotic analysis of texts and other materials as well as for evaluating the impact of ancient myth on the figurative arts in contemporary culture. The Research Group on Myth and Mythography (GRIMM), based out of the University of Trieste’s Department of Sciences of Antiquity "Leonardo Ferrero," has furnished a considerable body of work intended for inclusion in a wide-ranging project known as the Online Etymological Dictionary of Greek Mythology (DEMGOL). Stemming from the doctoral dissertation of Carla Zufferli (University of Trieste), this work is being carried out by Ezio Pellizer (University of Trieste), with contributions from Francesca Marzari (Siena), Luisa Benincampi (Trieste), Alberto Cecon (Trieste) and other GRIMM

126

members; Francesca Marzari and Françoise Létoublon (Grenoble) are working on the French translation (HOMERICA Group, Grenoble); Álvaro Ibáñez (Granada), José Antonio Clúa Serena (Barcelona) and Diana de Paco Serrano (Murcia) are working on the Spanish translation. Other versions are currently being planned, including versions in English and Portuguese. Publication of the Dictionary on the World Wide Web is being undertaken by Giovanni and Nevio Zorzetti on the University of Trieste’s “Hirema” (Historical Resources Management) Laboratory, deploying an application in the Java scripting language that allows contributors to edit and translate the text in a collaborative environment and to bring completed work immediately on-line. The project was conceived as an update of Carnoy’s Dictionnaire etymologique, which is now obsolete and which tends to seek out “Pelasgic” etymologies, often explaining obscura per obscuriora; it also improves upon Room’s Classical Dictionary, which is popular with the English-speaking public but limited in size, addressed to non-specialists, and deficient in both citing the main sources of myths and providing full and reliable etymologies (cf., e.g., the etymon of Antigone). Each entry in DEMGOL includes essential information on the character of the myth and mentions the ancient Greek and Latin sources of major relevance, without attempting to replace the many dictionaries of mythology now widely available. Furthermore, it is generally more detailed in the matter of minor and less well-known figures; in these cases, the main ancient sources are quoted with the aim of providing a historical perspective for the mythological accounts. This is followed by a meticulous study of the most plausible etymologies of the names of the heroes, heroines, gods, animals and monsters of Greek myth. The working methodology of the Dictionary is based on a careful analysis of the etymologies suggested by the scholars in the past, starting from Pape and Benseler’s list or the entries of Roscher’s Lexikon, now more than a century old, and continuing through the most recent research of Frisk, Chantraine, Wathelet, Zamboni, Salvatore, von Krafft and so on. The most plausible and linguistically substantial of these suggestions are chosen, and, wherever possible, Mycenaean attestations of anthroponyms or individual etymological components of those names are mentioned. At present, the work is undergoing continuous scholarly revision and careful editorial control, as the entries are uploaded to the database where they can be consulted immediately on the World Wide Web. In this way, the opportunities for expanding, revising and improving each entry are great, since it is easy to make additions, corrections and new entries or add new images remotely by computer. As of September 2008, there are about 890 entries in

127

Italian, with over 700 translated into Spanish and 400 into French. For the yet un-translated French entries, an experimental system of “Controlled Automatic Translation” is being carried out by computer experts working with a team of scholars at Grenoble’s HOMERICA Centre. If this proves successful, this system will be extended to the on-going Spanish version, as well as the planned English and Portuguese versions. Currently, each entry contains dozens of images and links to other relevant websites; soon, however, hundreds of mythological images will be available on a global scale, mainly for modern and contemporary art and graphic art. Finally, testing is being carried out by an international committee of scholars and referees (A list is given on the first page of the GRIMM website, http://www.units. it/~grmito/ - GRUPPO). Because of the quantity of images and links that connect certain entries to other websites - pointing up the broad influence that Greek mythology has had on the culture of modern Europe and still has on contemporary society -, DEMGOL has great promise as a tool both of educational development and scholarly outreach, even at great distances. Even scholars who tend to favour material culture in the study of ancient civilizations will understand the importance of such a versatile analytical tool: it can be extremely useful in reconstructing the narrative or thematic structures of a huge deposit of cultural materials stratified through the development of Greco-Roman, medieval, Renaissance and modern culture, in terms both of the literary and folk forms of imaginative production, and of the iconic forms of European figurative art (above all in a Warburghian perspective that is open to the contribution of ethno-anthropological research). The structure of the work itself allows the reader to consult partial sections, such as the “Vocabulary of Symbolic, Hybrid and Monstrous Animals,” or a category of myths related to astronomy (“catasterisms”) recording the numerous heroes, heroines or animals transformed into constellations. To this end, the iconographic apparatus in DEMGOL should prove of great use (IMAGES section). Thanks to the structure and the capabilities of the electronic database, these images can be multiplied to such an extent that it will be possible to offer a very wide survey of the iconic forms that myth has assumed diachronically in figurative art in Europe, from the Middle Ages and the Renaissance to today. Unlike other paper or computer tools now in existence, DEMGOL pays particular attention to the presence of this cultural heritage in contemporary graphic art and art, on a global and worldwide scale. Needless to say, the entries available in DEMGOL today (which is already useful and operative, even if incomplete: over 1000 items have been planned)

128

are linked to one another internally, allowing the reader to consult the Dictionary quickly and effectively, both for research and instruction, in three major languages spoken in Europe as well as in South America, Frenchspeaking Canada and the Philippines. DEMGOL has also been recommended by INTUTE (http://www.intute.ac.uk/), a website which gives information on “high-quality Internet resources for education and research.” For example: if one wished to expand and extend a study of the “mythical” models of some famous figures of ancient statuary - such as the famous Apoxyòmenos of Lo!inj -, DEMGOL provides an easy and quick way to research the typology of the young man of ephebic age, like Hyacinthus and Narcissus (fig. 1). Similarly, if one wished to see what the ancient Greeks meant by “Satyr,” DEMGOL makes it easy to go back to the images of the famous Satyr of Mazara del Vallo, with an meticulously researched etymology. The effectiveness of this tool is even greater in the study of mythical themes, however, which, while mostly Hellenic, entered into the astrological (and partly astronomical) traditions of the Western and Arab worlds. Of course, everyone knows that Jupiter, Saturn, Mars and Venus are late-Latin translations of planets that for centuries were called Ares and Aphrodite, Zeus and Cronos, and that nearly all the constellations have Greek names. By selecting DEMGOL’s category “catasterisms,” visitors can explore the numerous myths that can be “read” in the stars, viewing Canis Minor (Maira or Mera, fig. 2), Hydra or Cetus (Kétos) or sea monster (fig. 3): Villa Farnese, Caprarola. There is no need to linger long on the importance of this tool for the scholar who would like to deepen his or her knowledge of the vague, unclear and imprecise notion of “myth”—a concept that covers a series of cultural realities being studied by modern anthropologists, theologists and historians of religion, making use (at least beginning with Lévi-Strauss) of theoretical and methodological tools similar to those used by students of the “exact” sciences, such as linguistics and cultural semiotics. When studying the two (or three) millennia that have witnessed the development of the cultures of medieval and modern Europe in the Mediterranean, DEMGOL can be used as a kind of compass for navigating the (hopefully calm!) seas of culture and science in the contemporary world, as a consideration of the fundamental heritage of themes and structures, tales and images in the study of its origins, in religion, philosophy, literature, art and science, is absolutely indispensable. So far, GRIMM has received funding from the following sources: a grant from the Rhône-Alpes region of Grenoble (2006, EU 25.000=), along with funding from the Italian Ministry of Scientific Research (2005-06, EU 13.600=)

129

for the entries related to the vocabulary of symbolic, hybrid and monstrous animals, as part of a research work of national interest (PRIN) in alliance with Siena, Turin and Palermo. In the past, GRIMM relieved upon a smaller funding stream, allocated by Fondazione CRTrieste (2004, EU 2.000=). Currently, GRIMM is seeking funding for postdoctoral grants that will allow us to finish the translation of DEMGOL into English and Portuguese, to enrich the iconographic apparatus and to complete the editing of numerous other entries on minor mythological characters, which will soon number over 1000. GRIMM is a not-for-profit endeavour, and is intended to involve young scholars and students in collaborative, interdisciplinary research projects promoting the popularization of European culture and distance learning.

Images:

Figure 1. The ephebe of Losinj (Apoxyòmenos)

130

Figure 2. Canis Minor

Figure 3. Cetus: fresco of Villa Farnese in Caprarola (Viterbo)

131

BUILDING UP AN ARCHAEOLOGICAL RESTORATION & CONSERVATION DEPARTMENT IN FRIULI-VENEZIA GIULIA FULVIA LO SCHIAVO *

1. Foreword The application of “hard sciences” to archaeology and in general to cultural heritage is certainly not a new topic, since the National Council of Research (CNR), due to the initiative of Luigi Donato, dedicated in the Fifties an Institute to Le Scienze Sussidiarie dell’Archeologia (“Sciences subsidiaries to Archaeology”). From this pioneering period onwards, the CNR Institute enhanced his activity not only following all fields of Cultural Heritage (the present name is: Istituto per le Tecnologie Applicate ai Beni Culturali, “Institute for Technologies applied to Cultural Heritage””) but also interlacing and optimizing the research, according with the new discoveries and orientation of Cultural Heritage and its conservation, improvement and opening to the public. From this point of view, to interconnect the “hard science” to the peculiar situation of Friuli-Venezia Giulia Cultural Heritage – and particularly to archaeology, is a basic necessity. Many of the original themes identified of this International Conference are, right in this moment, under the highest attention and immediate application of

*

At the time of the Conference, the present author was acting Superintendent of the Archaeological Heritage in Friuli-Venezia Giulia, at the same time having the main responsibility in Tuscany. In April 2008 this interim assignment was suspended. Nevertheless, since this paper was officially delivered at Velj Losinj and is based on experiences matured both in Tuscany and in Friuli-Venezia Giulia, it is correct to keep the original text. I am greatly indebted to the restorers of Friuli-Venezia Giulia Superintendence: Antonella Crisma and Luisa Zubelli, Daniele Pasini and Gianni Gallet, and to Franca Maselli Scotti, former acting Superintendent, Director of Aquileia Museum and territory, responsible for Archaeological Heritage in Trieste and Gorizia provinces, excellent archaeologist and great friend.

132

the Soprintendenza per i Beni Archeologici (Superintendence for Archaeological Heritage) of Friuli-Venezia Giulia, among which: 1. MARINE ARCHAEOLOGY, including the problem of waterlogged wood and the restoration of any handmade object recovered from under water; 2. ANALYSIS & RESTORATION OF ANCIENT BRONZES AND METAL MADE OBJECTS; not to forget mosaics, stucco works, glass and fayence objects, pottery, and so on; 3. ENVIRONMENTAL & CLIMATIC IMPACT ON CULTURAL HERITAGE, including the problem of conservation in the sites chosen for the exhibitions; 4. SCIENCE FOR GREEK AND ROMAN ARCHAEOLOGY IN THE EASTERN ADRIATIC. 5. It is also necessary to add SCIENCE FOR PREHISTORIC AND PROTOSTORIC ARCHAEOLOGY IN THE EASTERN ADRIATIC, highly important because less self-protected and in worst conditions of conservation for the stratification of centuries of life and use of the site. 6. MOBILE LABORATORIES FOR CULTURAL HERITAGE ANALYSES, and many other matters. The natural conclusion from these premises, taking advantage by all possible interconnections and acquired experience, is to trace the outline of an Archaeological Restoration and Conservation Department for the Friuli-Venezia Giulia region and, here, to present a project of high and mutual interest, from the points of view of the human sciences, of the hard sciences and also of the enjoyability by the people and by the tourists.

It is evident that there are too many things to say to such a stimulating audience. Therefore, after many positive discussion with my dear friend and colleague Manuela Montagnari, I selected one single aspect that summarizes all the others. Now I’ll explain my point and then I’ll illustrate it through many examples. 2. The point Science and archaeology = equal = conservation, restoration and management

133

Archaeology is the science investigating the past. = equal = Archaeology is a Science-based discipline About the difference between “hard” and “soft” science, there is not much to say: it is out-of-date, based on “ignorance” and must be totally revised. “Human” science is a far better term than “soft”, but “human” do not include ancient history and archaeology, that are true sciences, in the investigation and in the reconstruction of the past: the best comparison is medical science methodology: working on the basis of diagnosis, prognosis, therapy and deriving prophylaxis. On the contrary, without science, we are in the domain of invention, romance, legend, that we all appreciate, while reading a relaxing book, but that we are not discussing now and that cannot be accepted in the field of conservation of our precious Cultural Heritage. There are two “chains” of actions: the “Discovery Chain” on the field, that is mostly archaeology and that is not going to be discussed here, with the exception of its consequence: Survey/Discovery, Excavation, Documentation, Diagnostic, Restoration and Conservation, leads to: the Culture of Management. The second chain is the “Exhibition Chain” in the Museum: Documentation, Diagnostic, Restoration, leads to: the Culture of Management. Both “chains” lead to the Culture of Management. “Management” is a stronger word than the Italian manutenzione (“maintenance, upkeep, care”) and an unfortunately highly despised idea – and practice –, that implies cleaning, checking and monitoring, in the broader sense, the state of health and/or disease of the cultural heritage: using a medical work, this is prophylaxis. It is evident that a negligence in the field of prophilaxis is cause of illness and ultimately of death, for human beings as well as for monuments and objects. Paola Pelagatti (Pelagatti 1994), former responsible of Southern Etruria Superintendence invented this concept, and Giorgio Bonsanti (Bonsanti 2004),

134

former director of the Opificio delle Pietre Dure in Florence wrote recently a short important paper on the subject. Archaeology as a Science-based discipline do not allow to select at random the object of an action, of a study, of a research: the statistical base and the population behavior are essential before any selection and all along any intervention. On the contrary, the criteria of selection of a work of art for restoration/ conservation is mainly based on its age, value, importance of the author, its meaning in the history of art and also in the history of a place and of a social group. It is looking after the excellence that the selection is made and not on the basis of a context. The archaeological method of restoration/conservation, aiming to understand the whole and operating only in consequence of a full analysis, as well as a doctor in front of a sick person, follows in some way an opposite method with respect to the uniqueness of a work of art. 3. Publication. Archaeology and restoration/conservation, as any other Science-based discipline, need careful studies, notes, experiments registered, written and published. Publication are due both to scientists and to the broad public, on the principle that everybody needs to know and that no discovery is such if nothing is not thoroughly explained. This wide argument is not going to be discussed here (Lo Schiavo 2006). 3.1. Examples in Tuscany: 3.1.1. Archaeological Research nowadays: the example of the Etruscan Sanctuary of Poggio Colla-Vicchio (Firenze). Archaeological research on the site of Poggio Colla is going on successfully since many years, lead by P. Gregory Warden, of the Southern Methodist University of Dallas, author of many studies, papers and volumes (Warden 2008 with previous bibliography). The exciting results of the research concern a rich Etruscan sanctuary on the acropolis, where a temple was built near a natural crevice and hundreds of offerings made of gold, iron, bronze, stone and pottery were buried and scattered all around.

135

As an example of field survey in extension the research through a new magnetometer was used on the nearby site of Podere Funghi, where in 2000 an Etruscan furnace had been located, but the discovery of many pottery sherds on the ground suggested the idea that other similar structures should be present. The magnetometer, needing only regular rows of small superficial holes to insert the bar of the apparatus, enables to check quickly and cheaply wide extension of land (Warden 2007). At Poggio Colla, the traditional methods of excavation are coupled with the highest sophisticated technologies of documentation. A field laboratory is run in parallel with the excavations, assisted by the specialists of the Conservation/Restoration Centre of the Superintendence for Tuscany Archaeological Heritage in Florence, through the close collaboration of the Superintendence archaeologist responsible for the territory of Mugello, Luca Fedeli, allowing the participation and training of the students. 3.1.2. Il Cantiere della Navi di Pisa e Centro di Restauro del Legno Bagnato (“Pisa Ancient Ships Yard and Restoration Centre of Waterlogged Wood”). The archaeological research at S. Rossore-Pisa, where a heap of wrecks and other remains (up to now about 30) dating from VIth BC to VIIth AD were discovered, are by now a well known subject: an archaeological non-stop enterprise from the excavation to the restoration/conservation of all finds (Camilli 2006 with previous bibliography). To discuss at length the pioneering techniques of conservation of waterlogged objects, not only wood, but also basketwork, metal objects, pottery, and so on, would be a perfect topic for a second International Conference in the framework of Archaeology = equal = Science, but exceedingly wide; in many occasions there were anticipations on these subjects (Camilli a cura di 2004). It is important to stress that the archaeological enterprise works from the very beginning in connection with at least 20 different scientific institutions and universities in Italy and in Europe: Istituto Centrale del Restauro (ICR-MiBAC) Roma; Scuola Normale Superiore, Pisa; ARCO, Laboratorio di archeobiologia, Museo Civico di Como; Dendrodata s.a.s., Verona; Istituto per la Conservazione e Valorizzazione dei Beni Culturali (ICVBC) CNR, Firenze; Istituto per la Valorizzazione del Legno e delle Specie Arboree (IVALSA) CNR, Firenze; Dipartimento di Archeologia, Università di Pisa; Dipartimento di Agronomia, Università di Pisa; Dipartimento di Chimica e Chimica Industriale, Università di Pisa; Dipartimento di Scienze e Tecnologie Forestali e Ambientali, Università di Firenze; Dipartimento di Agronomia, Università di Genova; Dipartimento di

136

Scienze della Terra, Università di Firenze; Istituto per le tecnologia applicate ai Beni Culturali (ITABC) CNR, Roma; Insegnamento di Archeologia Marittima, Università di Roma III; LENS, Firenze; Escuela Española de Arquelogìa, Roma; Museum für Antike Schiffart, Mainz; NucleArt, Grenoble (Camilli …; the updating on the activity of the Restoration Centre for the Waterlogged Wood can be found in the site http://www.cantierenavipisa.it and on the on-line review Gradus). The Restoration Centre for the Waterlogged Wood is originated by the Conservation/Restoration Centre of the Superintendence for Tuscany Archaeological Heritage in Florence, working particularly since 1966 on an international level and responsible for exceptional exploits, such as the restoration of the Riace bronzes, the Cartoceto bronzes (Rastrelli 2006) and, recently, the Amazons Sarcophagus (Bottini, Setari a cura di 2008) and the Minerva from Arezzo (Cygielman, a cura di, 2008). Another important department is the Archaeoanthropology and Archaeozoology Laboratory (Pacciani 2008). 3.2. The Restoration Laboratory at Aquileia. Aquileia is a world by itself. A great Roman town, a true capital in the north of Italy, rich and important for cultural level, trade connections, works of art, roads, aqueducts, splendid public and private buildings decorated with mosaics, stucco and sea-shells decorations and moreover with statues, fountains, inscriptions. Since the very beginning of the archaeological research, restoration/conservation was a non-stop enterprise. The site where the analyses and conservation techniques take place is an elegant two-storied house, in the same complex and within the same enclosure of the Archaeological Museum, following the two cloisters where the mosaics and epigraphs are exhibited, in front of the manager’s and administrative office of the Museum and of the Archaeological Park. It is large enough and displays appropriate equipment to allow a lively activity. In time, the mass of objects duly restored exceeded the dimension of the Museum cases and also exponentially grew what we use to call “the second choice Museum” (Museo di seconda scelta), a wide deposit where the objects are set side by side, at the students’ and scholars’ disposal. Actually, in the Restoration house of Aquileia there is material enough to fill up not less than three Museums. The main problem is that, considering that findings go on without interruption, mostly because of rescue excavations happening day by day, since the modern town of Aquileia overlays the Roman colony, not to forget the Forum restoration, which means at first the complete excavation of the area and

137

of the foundations of the structures, the few restorers – though excellent – are unequal to the huge task. We are speaking of four persons, one of which is mostly attending a local political appointment, and only one of a high specialization level. In consequence, they are working on various materials: pottery, glass, metals, stones, but their best accomplishment, about which they acquired a notable renown, are the mosaics. 3.3. The Restoration Laboratory at Trieste. Three restorers are active in Aquileia and only one in Trieste, highly specialized, mostly dedicated to pottery but also working with good results on mosaics, and attending to the conservation and management of the many archaeological sites existing in the town. The Laboratory is housed in the rich and well equipped kitchen of the Palazzo Economo, an elegant middle class residence, where from the beginning all the Superintendences of Friuli-Venezia Giulia and later on also the Regional Direction for Cultural and Landscape Heritage is located. A second smaller Laboratory is dedicated to the works of art and a second bigger one is in Udine. The kitchen of Palazzo Economo is an interesting and well preserved XVIIIth Century example with porcelain stove and decorated walls, and with a copper and iron oven, more a museum room than a modern laboratory for conservation. 3.4. The Restoration System. It is evident that four restorers, in a building packed up with archaeological materials and an historical kitchen are not enough to ensure the restoration/conservation of the Archaeological Heritage of Friuli-Venezia Giulia, mostly because this heritage is growing early through the planned excavations under the direction of the Superintendence archaeologists and through the excavation “in concession” (in concessione) under the direction of the different Italian and foreign Universities archaeologists (Lo Schiavo, a cura di, 2008). On the other hand, in Villa Manin at Passariano (Udine) there is a Regional Restoration and Cataloguing Centre. At the moment, the more developed specialization is dedicated to paper, books and bindings restoration, through a High Training School, while to archaeology is dedicated mainly a cataloguing activity.

138

The ideal project would be to join the two institutions in order to build up a new strengthened system, a “Restoration System”: 1. 2. 3.

4. 5. 6.

new wide deposits to house the archaeological materials as it is discovered, where to carry out the diagnostic activity; in the same site, also first aid and planning of specialized restoration activity can be made; distribution of the materials to different specialized laboratories in the Region, in Italy or abroad if necessary, according to the kind of material and state of preservation; parallel training of students from different Universities and parallel archaeological cataloguing; temporary or permanent exhibition of the results in local museums; parallel popular and scientific paper- and digital- publications.

The Restoration System should be accessible to various Superintendences and to various Universities, according to different projects, followed and shared, step by step by the scientist, taking advantage by the possible applications to experiments to various materials. Financing should be participated by the Ministry for Cultural Heritage and Activities, the Region Friuli-Venezia Giulia, the Provinces, the local municipalities interested in the conservation and exhibition of their heritage, bank or banking association and a system of sponsorships, local industries and firms. Sponsorship and administration should be autonomous, guaranteed by banks or banking associations acting as treasury for the investors. If necessary, the Restoration System can become a Foundation. Even if I do not minimize the difficulties, I am convinced of the practicability and feasibility of this ideal: let us hope that somewhere and somehow, at least as an experiment, can develop and grow up, to the advantage of the Archaeological Heritage.

139

RELATIVE SEA LEVEL CHANGES BY USING ARCHAEOLOGICAL MARKERS: THE INTERREG ITALIASLOVENIA PROJECT "ALTO ADRIATICO” * STEFANO FURLANI DiSGAM, Università degli Studi di Trieste,via Weiss 2 34127 Trieste, Italy FABRIZIO ANTONIOLI ENEA, Special Project Global Change, via Anguillarese 301 00660 S. Maria di Galeria, Rome, Italy RITA AURIEMMA Dipartimento Beni Culturali, Università degli Studi di Lecce, via D. Birago 64 73100 Lecce, Italy Six submerged archaeological sites located along the NE Adriatic coast (Italy, Slovenia and Croatia) and dated ~2.0 ka BP were studied. In particular, we provide new precise measures measured with respect to the present sea level of submerged archaeological and geomorphological markers (notches), that are considered good sea level indicators. The interpretation of their functional heights, related with sea level at the time of their construction, allows to obtain data on the relative changes between land and sea. These data have been compared with the predicted sea level rise curves, using new mathematical models for the glacio-hydro-isostatic contributions associated with the last deglaciation. The northeast Adriatic (Italy, Slovenia and Croatia) is an area of subsidence and we use the calibrated model results to isolate the isostatic from the tectonic contributions. This indicates that the Adriatic coast, from the Gulf of Trieste to the southern Istria, has been tectonically downlifted by no less then ~ 1.5 m since Roman times.

1. Introduction Sea-level change is the sum of eustatic, glacio-hydro-isostatic and tectonic factors. The first is global and time-dependent, while the other two vary according to location. The glacio-hydro-isostatic part along the Italian coast was recently predicted and compared with field data, at sites not affected by significant tectonic processes [48]. The aim of this paper is to provide new data on geoarchaeological and geomorphological markers to study the relative sea level rise during the late *

This research has been partly funded by the EU Project Interreg IIIA, Phare CBC Italia–Slovenia.

140

Holocene along the coastlines of North-eastern Adriatic (Slovenia, Croatia and Italy) surveyed in the Interreg IIIA Italy-Slovenia Project. Archaeological and geomorphological indicators represent a powerful source of information from which the relative movements between land and the sea can be evaluated. Archaeological evidence in micro-tidal areas, such as the Mediterranean Sea, can provide significant information for the study of relative sea level changes during historical times. Ancient coastal structures require a precisely defined range of functionality related to the sea level at the time of construction. Slipways, fish tanks, piers and harbour constructions, generally built before ~2 ka BP, provide a valuable insight of the regional variation in sea level in the last 2000 years [48, 49 and references therein]. Quarries carved along the coastlines and located near fish tanks and harbours or villas of the same age can provide additional data, both on the past water level and on their own functional elevation above sea level, although the quarries are not very precise indicators [24]. In this paper, we examine archaeological evidence from the North-eastern Adriatic coasts (Italy), where the development of maritime constructions reached its greatest concentration during the Roman times and where many well preserved remains are still present today. The best preserved sites were examined providing new information on their relations to the mean sea level in the I century A.D. Isostatic and tectonic contributions to this change are estimated from observational and model considerations to establish the eustatic change over this period. In addition we present new data on late Holocene sea level and on the vertical rate of tectonic movements in the Gulf of Trieste (Italy), in Slovenia and Croatia (Fig.1). These provide a key for the understanding of the geodynamic evolution of the Mediterranean basin. Unpublished archaeological markers such as docks, piers and pavements (all presently submerged), and geomorphological markers, such as core stratigraphy, as well as tidal notch data, were used as benchmarks recording the relative vertical motion between land and sea since their construction or formation. The heights of the selected archaeological markers were measured with respect to the local sea level. The interpretation of their functional heights provided new evidence on the changes. These data, together with their relative error estimation (elevation and age), were compared with predicted sea level rise curves using a new prediction model for the Mediterranenan coast. This model consists of a new esl function (the ice-volume equivalent sea level change, [43]) that assumes a small continuous melting of the Antarctic ice sheet until recent times. The accuracy of these predicted values is a function of the model parameter’s uncertainties defining the earth response function and the ice load

141

history (esl). This new model is more accurate if compared to the previous one by [48], especially in northern Italy, because of the inclusion of an Alpine deglaciation model [50] and because of improved Scandinavian and North American ice sheet models [51]. The results provide new data on the sea level rise and tectonic rates in the North-eastern Adriatic coast during late Holocene. 2. Geodynamic setting The Alpine Mediterranean region marks the broad transition zone between the African and the Eurasian plates and its tectonics are a result of the evolution of the related collisional plate boundary system [52, 33, 21]. Thus, the geodynamics of this region are driven by lithospheric blocks showing different structural and kinematic features including subduction, back-arc spreading, rifting, thrusting, normal and strike slip faulting [53, 59, 60]. The recent dynamics of the region are shown by the distribution of seismicity that outlines the plate boundaries and the quasi-aseismic domains such as the Adriatic and Tyrrhenian areas. These areas have been interpreted as rigid blocks or microplates or as undeformed sedimentary basins, limited by lithospheric-scale structures such as subduction fronts and large strike slip fault systems [20, 70]. Instrumental and tectonic data show a complex deformation pattern related to the kinematics of the Adriatic region, which has been interpreted as a block (Adriatic block) that is independent -or partially independent from the African plate [1, 80, 79, 63, 62]. Although this region displays an active deformation and its kinematics are still debated, interpretations of recent GPS observations considered this area as a unique crustal block rotating counter-clockwise [72]. This block moves independently from the African plate and displays a NorthSouth shortening in the central and eastern southern Alps at 1-2 mm/a and a northeast-southwest shortening between 1.6 and 5 mm/a along the Dinarides and Albanides. A comparison between the motion predicted by the rigid-rotation of Adria and the shortening observed across the area of the largest known earthquake that struck this region (the 1976 Friuli earthquake) suggests that the 2.0 ± 0.2 mm/a motion of Adria is absorbed in the southern Alps through thrusting and crustal thickening, with very little or no motion transferred to the north, and a northward-dipping creeping dislocation whose edge is located within a 50 km wide area beneath the southern Alps [17]. The geological features of the region is characterized by a thick carbonatic succession dating from Upper Jurassic in the Central part of the Istrian peninsula to Lower Eocene, which continued during the Lower-Mid Eocene with turbiditic flysch deposits [16, 77, 31].

142

3. Recent movements Along the North-Eastern Adriatic coast the MIS 5.5 geomorphological marker does not outcrop [22]. These deposits, which have been observed in boreholes between 85 and 117 m below sea level in the northern Emilia-Romagna region, provide evidence that a significant tectonic subsidence occurred during the last 125 ka. The amount of the subsidence rates, however, is not straightforward, since large uncertainties exist both in terms of age and position of the paleoshorelines of the sampled deposits. Given all the above uncertainties, subsidence at a rate of ~1.0 mm/a can be estimated for this area. Two further sites located in the northern Adriatic (Veneto and Friuli), display lower subsidence values (-0.7 and -0.2 mm/a [22]) with respect to those markers located in Emilia Romagna, being located close to the Po Plain, thus witnessing a crustal flexure due to the Southern Alpine and Dinaric contraction. Pirazzoli [60] surveyed some sites in southern Istria and northern Croatia, which display a well developed notch at -0.5/0.6 m, while Fouache et al. [25] extended the investigations to Northern Istria, finding archaeological and geomorphological markers at around the same depths and related to some Roman age remains as well as to submerged notches. Lambeck et al. [48] summarized late Holocene data for the Emilia, Veneto and Friuli coastal plains using lagoonal markers sampled and dated in cores at different depths. The results show tectonic subsidence with lowering values from west to east at 1.1, 0.45, 0.37 and 0.28 mm/a. Benac et al. [5] have provided a detailed description of marine notches between -0.5 and -1.0 m in the Gulf of Rijeka, possibly downward displaced by the co-seismic deformation occurred during an earthquake around 1000 years BP. 4. Materials and Methods 6 archaeological markers and several tidal notches were surveyed along the North-eastern Adriatic coast (Table 1). The surveying involved four steps: measurement, correction, error bars and comparison. Measurements of the elevation of the submerged archaeological markers with respect to the local sea level at the time of survey. Values reported in Table 1 are the mean values of multiple measurements collected in correspondence of the best preserved parts of the investigated structures. Correction of surveyed data via the Trieste tide gauge data collected at the time of surveys. Data are reported in Table 1. Error bars for the elevations and age values of the archaeological markers have been provided. Their functional heights have been evaluated on the basis of accurate archaeological interpretations provided by the staff of archaeologists. Age errors have been estimated from the architectural features; elevation errors have been

143

derived from the measurements, corrections and estimation of the functional heights (Fig. 2). The comparison of predicted and observed sea levels sea level elevation predicted by the new Lambeck model with respect of the current elevations of the markers (i.e. the relative sea level change at each location). At the sites where the elevations of the markers are in agreement with the predicted sea level curve, we hypothesized tectonic stability of the locality. On the contrary, when the elevations of the markers differ from the predicted sea level curve, we hypothesized that the area is subjected to tectonic subsidence.

A Site name

B Coordinates

C Survey date (yyyy/mm/dd, h)

1 Stramare

45°36'07'' 13°47'24'

2005/07/16 h 13:40 GMT

2 Punta Sottile

45°36'08''' 13°43'10'' 45°35'34'' 13°42'53'' 45°31'57'' 13°38'41'' 45°29'59'' 13°30'13'' 45°29'59'' 13°30'13'' 44°54'40'' 13°46'29'' 44°54'39'' 13°46'35''

2005/05/25 h 19:55 GMT 2005/11/10 h 15:10 GMT 2004/10/26 h 10:30 GMT 2005/10/17 h 13:00 GMT 2005/10/17 h 13:30 GMT 2004/10/27 h 12:30 GMT 2004/07/05 h 14:20 GMT

3 Jernejeva draga San Bartolomeo 4 Sv. Simon San Simone 5a Savudrija/Salvore 5b Savudrija/Salvore 6a Briunj 6b Briunj

D E Type and Archaeolo measured gical age height (m) (yr BP) Walking 1900±100 surface, -1.66 Pier, 1950±50 -1.65 Vivaria 1900±100 dock, -0.70 Pier, 1950±50 -1.40 Pavement, 1950±50 -1.18 Pier, 1950±50 -0.10 Pavement, 1950±50 -1.20 Dock/Pier, 1950±50 -1.10

F Tide (m)

G Corrected Height (m)

+0.06

-1.60

H Functional height (m) 0.0 a.m.s.l.

I s.l. change (m)

References

+0.25

-1.00

-0.10

-0.80

+0.40

-1.00

-0.32

-1.50

-0.40

J

-1.60 ± 0.60

This paper

0.60 a.m.s.l.

-1.60± 0.60

0.60 a.m.s.l.

-1.40 ± 0.60

Auriemma et al. (2007, in press) This paper

0.60 a.m.s.l.

-1.60± 0.60

Degrassi (1957)

0.0 a.m.s.l.

-1.50 ± 0.60

This paper

-0.50

1.00 a.m.s.l.

-1.50 ± 0.60

0.00

-1.20

0.60 a.m.s.l.

-1.80 ± 0.60

+0.10

-1.00

0.60 a.m.s.l.

-1.60 ± 0.60

Fouache (2000) and this paper Degrassi (1957) Fouache (2000) Degrassi (1957) and this paper

Table 1. Measurement data and inferred sea levels for archaeological sites in the NE Adriatic region. A: Site names and numbers and the latter are also shown in Fig. 1. B: WGS84 coordinates of the surveyed sites. C: Year, month, day and hour of measurement. D: Field measurements (before correction). E: Age of the archaeological sites. F: tidal correction applied for tide amplitude at the time of surveys. Tide values at each location are computed with respect to the Mean sea Level of Genova, using data from the local reference tide gauge data of Trieste, which are the archaeological sites nearest to the permanent stations, and including tide time delays at each site. G: Corrected elevation of archaeological structure surveyed as derived from data in columns D, F and G. H: Functional height of the marker used with respect to mean sea level. I: Estimated relative sea level change. Errors are within the tide amplitudes of ± 0.60 m for the Adriatic sea. J: References. For more information about the tide gauge of Trieste, also see http://www.univ.trieste.it/~dst/OM/OM_mar.html

Elevation measurements -with respect to the current sea level at the time of the surveys- were performed through the use of optical and mechanical methods (Salmoiraghi Ertel automatic level or invar rod). All the measurements of the archaeological features’s depths were made in times of low wave action and they were related to the sea level position for that particular moment. Since the investigated archaeological structures (fish tanks, harbours, piers) were used year-round, we assumed that the defining levels correspond to the annual mean conditions at the time of construction. The measurements are therefore reduced to the mean sea level applying tidal corrections at the surveyed sites, using the data of the nearby tide gauge at Trieste. Elevation measurements are given [76]

144

with respect to the Italian reference plane network of the Istituto Geografico Militare (Genova Mean Sea Level 1942, [27]). The estimation of the tide amplitudes and their correction is a crucial element for the northern Adriatic sea; the measurement of the markers’ elevation must be properly corrected for tides, as they here show the largest values in the whole Mediterranean basin (up to ~1.8 m and mainly produced by meteorological variability in a closed basin, as opposed to the normal values of max ~0.45 m , from the Tidal Data Base of the Italian Istituto Idrografico della Marina). For the above mentioned reasons, local tide amplitudes were also estimated using data from the nearby permanent tide gauge located in Trieste (recording since 1890). In order to estimate the sea level change in each location, and to compare the observed results in different locations, we defined the functional heights of the archaeological benchmarks. This parameter is defined as the elevation of specific architectural parts of an archaeological structure with respect to an estimated mean sea level (tidal sea level) at the time of their construction. It depends on the type of structure, on its use and on the local tide amplitudes. Subsequently, functional heights also define the minimum elevation of the structure above the local highest tides. To improve the interpretations, we also measured the functional heights at some modern harbour structures (piers and docks) located along the coasts of the Gulf of Trieste, comparing them with those measured at the archaeological sites located in the nearby areas. For example, we assumed that the pavements at the top of the piers were in the range 0.5/1.0 m above sea level. Subsequently, as the tide amplitude is up to ± 0.9 m in the Gulf of Trieste, during particular meteorological events, the top surfaces of some small piers or docks can be nearly submerged during maximum tides. On the other hand, the seafloor in some basins can become dry during the lowest tides. It is worth noting that the architectural features and functional heights of modern piers and docks are in agreement with those of Roman age. This information can also be deduced from previous publications [23, 24, 32], from historical documents (Vitruvius, [32]), from the remnants of Roman age shipwrecks (which provided data on the size of the ships or boats and their draughts [69, 75, 14] and through rigorous estimation of the functional heights of the piers, by using and interpreting different type of markers on the same location [49]. As far as we know, navigation during Roman times was mainly seasonal (mare clausum from October to March) and the Roman ships that used these coastal structures had draughts of ~0.5 m, which fit the features of the observed archaeological markers. The use of these structures, their age and

145

conservation, the accuracy of the survey and the estimation of the functional heights were all used in considering the observational uncertainties at each site. 5. Data The archaeological sites we surveyed were already known by the scientific community, although most of them have never been used before for geomorphological studies and they had not been surveyed using direct underwater observations. Since this is a critical point for a rigorous estimation of the relative sea level changes based on archaeological markers, surveys were performed to provide new and affordable data for our geophysical and geomorphological goal. 5.1. Geomorphological markers The North-eastern Adriatic coast is the result of the Holocene submersion that was largely completed about 7 ka calibrated (cal) BP. Afterwards, the sea level rose only slowly up to the current elevation. With the exception of storm or tsunami deposits found nearby Pula [2] at an elevation of about +0.7 m, along the North-eastern Adriatic coasts, no marine notches or fossils have ever been found at elevations higher than the current sea level. Tidal marine notches are considered to be good markers of coastal tectonic movement. Pirazzoli [67]observed submerged marine notches in Croatia at ~-0.6 m and Fouache et al. [25] studied and measured some submerged notches along the Istrian coast at the same altitude. These notches have been attributed to Roman age. Benac et al. [5] measured the submerged notch on the Gulf of Rijeka at a depth between -0.5 and -0.6m and in Bakar Bay between -1.03 and 1.15m (Fig. 3). These Authors measured notch depth with respect to the local Biological Mean Sea Level and they ascribed the recent position of the notches to rapid coseismic subsidence following an earthquake in AD 361. In view of these observations and with the aim of providing new measurements on the whole NE Adriatic area, we surveyed the Northern limestone coast of Istria and the Gulf of Trieste (Italy), providing high-density measurements. South of this area, further data were collected in the Kornati islands of southern Croatia as well as in Montenegro (Fig. 3). If we take a close look (east to west) at the the Gulf of Trieste (Italy), we observe that at Miramare there is a well carved notch at an elevation of -0.6 to 0.8 m (tide corrected). Only ~6 km west from Miramare, its elevation increases to -0.9 m. Between Sistiana and Duino (Italy), toward North-west, the depth of the notch continues to increase from -1.3 m, going down to -2.5 m, as measured at six different locations (Fig. 3). In accordance with the local tide

146

amplitudes (the highest in the whole Mediterranean sea) the width and amplitude of the notch are ~1 m (with a well defined tidal notch shape (Fig. 4). Unfortunately, biological organisms have not been preserved, preventing dating of the notch. A submerged tidal notch [67, 25, 5] runs south of the previously described one along the coastlines of Istria and Croatia at an average elevation of -0.6 m. South of the Gulf of Rjeka, towards Montenegro, the present day tidal notch was not observed. A submerged notch was instead found at an altitude of about -0.5 m. Fig. 3 illustrates this notch’s distribution and elevation. Our observations show that the present day notch is absent along the limestone coasts of the Northeastern Adriatic, between Duino (Italy) and Kotor (Montenegro), while a submerged notch was observed at about -0.6 m below the present day sea level. 5.2. Archaeological markers In addition to the above mentioned gemorphological markers, descriptions and data were also provided for seven coastal Roman age archaeological sites (see Fig. 1 and Table 1) that were well related with sea level. 5.2.1. Stramare (Muggia, Trieste) At Stramare, near the Ospo Stream mouth, the terrace behind the narrow beach is characterized by many traces of protohistoric and Roman habitation, found despite the damage caused by the modern industrial district [9, 10, 11, 12, 54, 55, 64, 65, 66, 81]. The lower terrace continues below the current sea level. In ancient times, the terrace was a land extension protecting the left side of the Ospo Stream mouth. Probably, the pars rustica or the pars dominica of a maritime “villa” once faced this open area. On the west side, this terrace was contained by a wall that was very similar to the emerged ones. The upper side of this wall is currently 1.6 m above the present day sea level (Table 1). The wall was built with thick stone slabs laid facing the ground with its foundation 0.5 – 0.6 m under its actual upper surface. At the time of its construction, the wall’s foundation level, now at 1.66 m below the sea level (-1.60 corrected for tide, pressure and wind), would have emerged at least for part of the day. On the northern and eastern sides, the terrace slopes down to 3.0 m: this elevation difference most likely marks the old seashore, and it is sheltered by large stone blocks, some close together, some scattered. Shards of amphorae and common ware of Roman imperial age occur in this submerged terrace but it’s difficult to establish this building’s chronological range and its use (Fig. 5 site 1, Table 1).

147

5.2.2. The pier at Punta Sottile SW (Muggia, Trieste) The Punta Sottile Pier was discovered in the 1980s [28, 29, 81, 61, 13] and it was recently excavated. The structure lies 40-50 m off the coastline. The first portion of the pier is made up of blocks belonging to the shore platform that, in this area is very regular such that the break lines could be wrongly interpreted as an artificial structure, and, partially, it is composed of cut blocks arranged and flanked in the areas where there is no shore platform (Fig. 5 site 2). The pier is 12 m long from its foot and 2.5 – 2.6 m wide. It was built with the so called “a cassone” technique, typical of the landing structures of the Eastern Adriatic Sea. Its façade is made of opus quadratum, with large 3 m long parallelepiped sandstone blocks containing a nucleus made of rubble and joined (in places) by transverse blocks. There are two overlapping layers of blocks: the first one is placed on a foundation that follows the marly shore platform. The foundation is made of a small heap of stones, pebbles, ceramic shards; the last of which allowed a safe dating of the time of pier construction, i.e. back to the central decades of the Ist century A.D. The sea-bottom is 1.1 m and 2.2 m deep at the pier foot and at its head, respectively, slowly sloping westwards, whereas the actual upper pier surface lies on a sub-horizontal plan between 1.15 m and 1.4 m. We hypothesize the former existence of a third layer which would have resulted in a near horizontal surface of the pier that joined the shore platform behind it (Fig. 5 site 2). This leads us to the assumption that the original pier depth was about 1 m, while the walking surface was possibly between at 1.1-1.0 m. If we therefore add the initial depth of -1.65 m (corrected to -1.00 m) to the functional height, the data corresponding to the relative sea level rise equals to 1.60±0.60 m (Fig. 5 site 2, 9A). 5.2.3. Jernejeva Draga/San Bartolomeo (Ankaran, Slovenia) A large fishery was discovered and excavated (this paper) in the S. Bartolomeo bay, situated very close to the Italian-Slovenian border [37]. The structure is composed of two large docks and most probably a pier. Its total length is 135 m, with a width of 50 m, while the west side is 80 m long. The docks are today contained in an embankment made of disconnected stones, but in Roman times the embankment probably had façades, at least on the inner side. Its eastern side is the main sea level indicator: it is an embankment for the eastern dock, a pier and a quay at the same time. Its shape is arched, but its foot is straight, 30 m long and 2.6 m wide. The pier has two (external) façades built near the close by stone blocks and the rubble of heap of stones. The actual pier surface seems to have lost one or two rows of blocks since its construction and if two large fallen blocks on the north side are placed one over the other are indicative of the

148

elevation of the ancient walking surface at -0.7 m which corrected (-0.80 m) and added to the functional height (0.60 m) indicates a sea level rise of 1.40 m (Fig. 5 site 3, Table 1). The suggested age of the S. Bartolomeo fishery is the beginning of the Imperial Age because of its analogy with other similar structures and of the amphora shards found between the stones of the embankment (Fig. 5 site 3). 5.2.4. Sv. Simon/San Simone (Izola, Slovenia) The splendid structures of the S. Simone bay “villa” and its harbour (the largest one on the Istrian coast and measuring over 8000 square metres) have been well known since the 16th century AD. Unfortunately, these structures were filled with concrete in the last decades [18, 19, 73, 74, 75, 7, 41, 6, 36, 57, 37]. The building includes a quay, a pier, a breakwater and other working areas. The pier starts from the South-west quay corner and is today only visible in the foundations of the modern wharf. The pier is 55 m long and 2.5 m wide and, in different stretches, it shows three layers of large (~2 m long) yet differently sized stone blocks. The lower layer is larger, in accordance with the Vitruvian construction rules, and on its upper layer, large mooring rings were probably placed, as recalled by the 19th and early 20th century observers. Today, the pier surface lies at a corrected height of -1.0 m and if the functional height was at least ~0.6 m, a sea level change at an average value of ~1.60 m can be estimated. The archaeological findings from the excavations at the “villa” allow us to date the most important habitation phase as being the 1st and 2nd centuries AD (Fig. 5 site 4). 5.2.5. Savudrija/Salvore (Croatia) The bay is sheltered by two large piers stretching out from opposite seashores. The 1990s excavations allowed us to conclude that the piers were built with large local stone blocks to protect the quay, which was 70 m long. Two inscriptions - one of which was dated back to the first half of the 1st century AD - were retrieved from the harbour area. These inscriptions suggested the presence of many buildings, both residential and commercial [19, 39, 40, 34, 56]. We collected measurements from two different areas: the first is an underwater terrace, in front of a quite well-preserved building standing on the beach (the so called cistern); the terrace is contained by quite large blocks lying on the shore platform at a corrected height of -1.50 m. Neither the function nor the date of this terrace are known. We assume that it was a shipyard or other functional working area connected with the buildings behind it and emerging above sea level only sometimes during the day. The second measurement is

149

from the southern pier which was built “a sacco”, i.e. with walls of large overlapped blocks in several layers (up to three conserved in the inner side) and stone rubble within it; this pier is higher than the others and it is located at a corrected height of -0.50 m below the present day sea level. For this 51 indicator, we estimated a functional height of at least 1 m above sea level, because it was probably a breakwater built to protect the inner basin of the harbour. Thus, a functional height of 1 m can be estimated as a minimum value (Fig. 5 site 5). 5.2.6. Brijuni/Brioni (Croatia) On Brijuni island in Verige Bay (Val Catena) lies the archaeological area of a splendid Roman “villa” with its harbour. The latter was active up to the late Roman period and its break-waters, quays and piers are all presently below sea level. Recent archaeological excavations performed during the ’90s documented the shapes of the underwater structures and specified the period of use [19, 78, 35, 71, 56, 57]. We performed measurements on the surface of the fishery foundation which was built using large stone blocks. Its shape is rectangular and it is 12.5 m long and 5 m wide. In the middle of its eastern side, some kind of steps following the natural slope of the sea floor reach the lower layer. Nowadays, the pavement surface of the fishery is at -1.20 m (tide, wind and pressure corrected) and indicates a relative sea level change of -1.80 m. Because of the depth (which is the same as the quay behind the piers) or because of the architectural typology and building technique, we cannot exclude that this may have been a thermal area (no longer active) built along the coastline, with steps at its entrance. Additional measurements were made at one of the two piers that close the Bay of Verige. This pier’s upper surface is currently located at -1m (Fig. 5 site 6) and for this site, a sea level change of 1.60 m can be estimated. 6. Data The theory used for describing the glacio-hydro isostatic process has been previously discussed [47] and its applications to the Mediterranean region has been most recently discussed in Lambeck et al. [48, 49] and Lambeck and Purcell [50]. The input parameters into these models are the ice models from the time of the last interglacial to the present and the earth rheology parameters. These are established by calibrating the model against sea level data from tectonically stable regions and from regions that are sensitive to particular subsets of the sought parameters: data from Scandinavia to constrain the northern European and Eurasian ice models [42, 51], a re-evaluation of the North American data for improved Laurentide ice models (Lambeck et al., unpublished) and data from far-field sites to improve the ice-volume equivalent sea-level function [46]. Iterative procedures are used in which far-field data is used to establish the global changes in ice volume and mantle rheology and near-field data is used to constrain the local ice sheets and mantle rheology. The

150

procedure is then iterated again, using the near-field derived ice models to improve the isostatic corrections for the far-field analysis. The Mediterranean data, being from the intermediate field, has been previously included in this analysis mainly to establish constraints on regional mantle parameters and the eustatic sea level function (loc. cit.) and on rates of tectonic vertical movements [45, 3]. In this paper we have used the most recent iteration results for the ice models [51] which includes improved ice models for the three major ice sheets of Europe, North America, Antarctica and Greenland back to the penultimate interglacial, as well as mountain glaciation models including the Alps [50]. This last addition impacts primarily on the sea-level predictions for northern Italy and Slovenia. The time-integrated ice volumes are consistent with the ice-volume equivalent sea-level function previously established [46, 44]. The Italian data discussed in Lambeck et al. [48] has not been used in arriving at the new model parameters. The adopted earth model parameters are those that have provided a consistent description of the sea-level data for the Mediterranean region. The Mediterranean data alone has so far not yet yielded solutions in which a complete separation of earth-model parameters has been possible, nor in which these parameters can be separated fully from eustatic- or ice-model unknowns but the combination used here provides a set of very effective interpolation parameters that describe well the observational data and that allow for an effective separation of tectonic and isostatic-eustatic contributions to sea level. Also, the eustatic parameters determined from the Mediterranean region are consistent with those obtained from other regions of the world [44]. The solutions indicate that three-layer rheological models largely suffice for the region: an effective elastic lithosphere with thickness ~ 65 km, an upper mantle from the base of this lithosphere to the 670 km seismic discontinuity with an effective viscosity of 3x1020 Pa s and a lower mantle with an average effective viscosity of ~ 1022 Pa s (earth model m3) (see also [48]) viscosity of 2x1020 Pa s and m-3 denoting 3x1020 Pa s. For the sites within the Gulf of Trieste (Slovenia) the predictions are also very similar for the individual sites and the observations can be combined into a single sea-level function with the Gulf. At these sites the hydro-isostatic signal is greater than e.g. it is in Sardinia [4] because of the coastal geometry and the alpine 53 glaciation signal [50] and as a consequence the predicted sea levels for recent millennia lie significantly closer to present sea level than do the Sardinia levels at comparable times. Beyond the Gulf of Trieste, geographic variability in sea level becomes more significant and observations from Brijuni lie up to 2 m lower than the first group because of the coastal geometry and alpine glaciation effects. This is further illustrated in Figures 6 in which the predicted shoreline elevations and gradients are shown for three coastal sections: along the western and inner coasts of Istria and along the along the Kornati Islands (see Figure 3 for locations). The predicted gradients for the two earth models m-2 and m-3 are similar over these distances and the major rheological dependence is shown

151

through the elevations. Between the southern side of the Gulf of Trieste to the southern end of Istria, a shoreline that formed at 2000 years BP would slope from north to south at about 0.3m/100 km and one along the Kornica islands would be predicted to slope at ~ 0.2m/100 km. 7. Discussion As discussed above the sea level response to the last glacial cycle is not expected to follow a eustatic function but will vary geographically across the Mediterranean and this is seen also in the examined area: the NE Adriatic Coast (Fig. 6). Any tectonic responses will accentuate this spatial variability. Thus whether the observational evidence for sea-level change is used for establishing a reference surface for estimating quantitative rates of vertical motion, for estimating eustatic change, or for evaluating the glacio-hydro-isostatic parameters, consideration must be given to all contributions. The NE Adriatic coast is a subsiding environment although for the Istria and southern Croatia coast the elevation of the MIS 5.5 shoreline is still unknown and long-term vertical tectonic rates have not yet been established. But this is an area with both historically and instrumentally recorded seismicity [30, 15] and one of horizontal deformation as measured by space geodetic methods [72, 17]. Figure 6 illustrate the comparisons of observations and predictions for the evidence from the Gulf of Trieste and from Brijunj. At both locations the predictions lie above the observed values, irrespective of whether earth-model m-2 or m-3 is used and this is consistent with a regional subsidence. The Gulf of Trieste data points are self-consistent suggesting that the entire southern side of the gulf has subsided by the same amount, between 1.4 and 1.6 m over 2000 years, depending on the choice of earth model. Likewise, the two data points from Brijunj are selfconsistent and point to a comparable subsidence, of 1-4 to 1.7 m during the past 2000 years. The average sea-level estimates for the two localities are –1.53 ±0.08 and –1.70±0.10 for the Gulf of Trieste and Brijunj respectively and the difference, while statistically not significant, is consistent with the predicted gradient along the coast of Istria. As previously noted, tectonic subsidence along the NE Adriatic coast can be anticipated from the absence of deposits or morphological expressions of the MIS 5.5 level above present sea level. The Late Holocene data points alone do not permit a distinction to be made between coseismic displacement and uniform subsidence. The model predictions indicate that in the absence of tectonics sea level has been close to its present level, and possibly marginally higher, for a prolonged period (Fig. 6) and the absence of the present tidal notch, as noted in areas of falling relative sea level (relative uplift) [3] here indicates that the recent relative change has been one of rising sea level which lends support to the model m-3. West of the Gulf of Trieste from Venice, Tagliamento and Grado plains, earlier estimates indicate that here the subsidence rates have been greater at between 0.7 and 0.3mm/year [48]. The submerged notch, widely reported from the Gulf of Trieste as far south as Montenegro, occur at a depth of about 0.6 m in both the eastern portion Gulf

152

of Trieste and along the Istria coast reaching 0.85 m at Briunji. For both earth models, there is a prolonged period when sea level is predicted to have been close or slightly above present sea level (Fig. 6) and in which notches can have been carved into the limestone coast only to be subsequently displaced by a coseismic event(s) of sufficient amplitude to displace the notch below the tidal range. Thus the notch itself is postulated to be the result of the eustatic-isostatic balance in sea level while its current position is an indication of coseimic activity having occurred after notch development and after the formation of the deeper sea-level markers at 2000 years BP. If model m-2 is appropriate then the notch formation would have started as early as 4000 years ago in the Gulf of Trieste and the absence of a notch below the 2000 year marker lends support to the model m-3 in which sea level did not reach its present level until much later years ago (Fig. 6). The absence of any trace of a modern notch suggests that the coseismic event was relatively recent and that sea level has continued to rise into recent time unless notch formation is influenced by surface water conditions (salinity, temperature, pH) in which case it would mean that these conditions have changed over the past 2000 years. It has been postulated that the displacement occurred as a 4th –6th century paroxysmic seismic event [68, 74, 5] but this cannot be validated by the present data as the historical catalogues [8] do not extend into this region. Recent measurements of limestone erosion-dissolution rates in the intertidal zone have shown that along the Northern Adriatic coast they are approximately 0.2 mm/yr compared with 0.02 mm/yr at measurement sites in the Trieste Classical Karst (Inner Karst) [26]. Preliminary measurements by one of the authors (Furlani) indicate that the limestone lowering values along the Tyrrhenian Sea coast are greater than those observed along the Northern Adriatic coast and this difference could be crucial to explain why the present day tidal notch is lacking in the Adriatic sea. The new data from the Adriatic region provide further evidence for the complexity of sea level change and contribute to the understanding of this change by making it possible to separate out the various causes. The area is one of rising sea level over and above any anthropological changes that may be occurring and the new data together with the model interpolations provide elements for evaluating flooding hazard scenarios where minor relative sea level rise can produce extensive coastline flooding, as in the NE Adriatic coastal region. The Adriatic coasts of Croatia, and Italy have been downlifted at 1.5 -1.6 metres, since roman times. The combined effect produced by the action of the Global Isostatic Adjustment and tectonics, both still active in the Adria block, which suffers from the complex geodynamic setting of the Mediterranean, can be responsible of a mean (regional subsidence plus cosismic displacements) of this area of 0.7 mm/y during the last 2000 years. In particular, the Adriatic coasts of Croatia and Italy have subsided by ~1.5–1.6m since Roman times at an average rate of ~0.75 mm/a (Fig. 7).

153

8. Conclusion Our data provide new estimate of the relative sea-level change and vertical land movements in the Northeastern Adriatic Sea, based on archaeological, geomorphological data as well as geophysical data and model. In the studied area, the difference between the data and the used model can be attributed to active tectonics occurred during the last 2000 yr. Results show that during the past ~2000 yr, a relative sea-level change has occurred at up to -2.08±0.60m since 1900±100 yr BP in northern Adriatic. The observed changes include a vertical tectonic signal at a rate of ~0.75 mm/a occurring in the last two millennia, which produced a significant downward displacement of the coastline of ~1.5–1.6m. Acknowledgments We are thankful to: Carla Braitenberg and Franco Stravisi for the helpful discussion on tide gauge data, Stavros Frenopoulos for assistance during scuba field survey in the Adriatic coastal sites. This research has been partly funded by the Australian Research Council (K. Lambeck) and EU Project Interreg IIIA, Phare CBC Italia–Slovenia: F. Antonioli, R. Auriemma, D. Gaddi, A. Gaspari, S. Karinja, V. Kovacic´. References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

H.M. Anderson and J.A. Jackson, Geophys. J. R. Astron. Soc. 91, 937-983 (1987). F. Antonioli and S. Furlani., International geological Congress on the Adriatic area, Urbino, 19-20 June, 2006 (2006). F. Antonioli, S. Kershaw and L. Ferranti, Quaternary International. 145146, 19–29. (2006). , F. Antonioli, M. Anzidei, K. Lambeck,R. Auriemma, D. Gaddi, S. Furlani P. Orrù, E. Solinas, A. Gaspari, S. Karinja, V. Kova!i" and L. Surace, Quaternary Science Revue. 26, 2463-2486 (2007). 33 #. Benac, M. Jura!i" and T. Bakran-Petricoli, Marine Geology, 212, 21(2004). E. Boltin Tome. Annales 1, 51-58 (1991). E. Boltin Tome and V. Kova!i!, ArhPregl, 29, 233-234 (1998). E. Boschi, E. Guidoboni, G. Ferrari, G. Valensise and P. Gasperini,, Catalogo dei forti terremoti in Italia dal 461 a.C. al 1980, ING SGA Bologna. (1995). D. Cannarella, Adriatico, 3-4, 21-24 (1962). D. Cannarella, Adriatico, 9-10, 8-41 (1965). D. Cannarella, BA LI, 195-197 (1966). D. Cannarella, Il Carso, Trieste (1968). Carta Archeologica del Friuli-Venezia Giulia. – - Dipartimento di Scienze dell’Antichità Regione Friuli Venezia Giulia unpublished.

154

14. G. Charlin, J.M. Gassend and R. Lequèment, Archeonautica, 2 (1978). 15. C. Chiarabba, L. Jovane and R. Di Stefano, Tectonophysics, 395, 251–268 (2005). 16. F. Cucchi, C. Pirini Radrizzani and N. Pugliese, Mem. Soc. Geol. Ital., 40, 1987, 35-44 (1989). 17. N. D’Agostino, D. Cheloni, S. Mantenuto, G. Selvaggi, A. Michelini and D. Zulianim, Geophysical Research Letters,. 32, L19306, doi:10.1029/2005GL024266 (2005). 18. A. Degrassi, AT III ser., X, 329-341 (1923). 19. A. Degrassi, AMSI, LVII, 24-81 (1957). 20. J.F. Dewey, M.L. Helman, E. Turco, D.H.W. Hutton, S.D. Knott, Alpine Tectonics, M.P. Coward, D. Dietrich and R.G. Park (eds), Geological Society Special Publication, 45, 265-283 (1989). 21. C. Faccenna, W. Thorsten, L. Becker, F. Lucente, L. Jolivet and F. Rossetti, Geophys. J. Int., 145, 809-820 (2001). 22. L. Ferranti, F. Antonioli, A. Amorosi, G. Dai Prà, G. Mastronuzzi, B. Mauz, C. Monaco, P. Orrù, M. Pappalardo, U. Radtke, P. Renda, P. Romano, P. Sansò and V. Verrubbi, Quat. Int., 145–146, 30–54 (2006). 23. N.C. Flemming, Spec. Pap.-Geol. Soc. Am. 109, 1– 125 (1969). 24. N.C. Flemming, C.O. Webb, Z. Geomorphol. N.F. 62, 1 –29 (1986). 25. E. Fouache, S. Faivre, J.J. Dufaure, V. Kova!i! and F. Tassaux, Zeitschrift fur Geomorphologie, 122, 33-46 (2000). 26. S. Furlani, F. Cucchi, F. Forti and A. Rossi, Geomorphology in press. 27. E. Gamboni, Boll. Geodesia Sc. Aff. 1, 155-165 (1965). 28. A. Gobet, Borgolauro 4, 14-16 (1983). 29. A. Gobet, Borgolauro 7-9, 113 (1986). 30. E. Guidoboni, A. Comastri and G. Traina, Eds. Istituto Nazionale di Geofisica Rome, Societa Geofisica e Ambiente, Bologna, 504 (1994). 31. M. Herak, Acta Geol. 16, 1-42 (1986). 32. A. Hesnard, Le strutture dei porti e degli approdi antichi, ANSER, 175-204 (2004). 33. L. Jolivet and C. Faccenna, Tectonics 19, 1095–1106 (2000). 34. M. Juri$i", Izdanja HAD-a 18, 163-168 (1998a). 35. M. Juri$i", Obavijesti HAD – a 1, 30, 81-90 (1998b). 36. S. Karinja, Arheolo!ka istra"ivanja u Istri 18, 177-193 (1997). 37. S. Karinja, Kultura, 259-276 (2002). 38. K.A. Kastens, J. Mascle, Proc. Ocean Drill. Progr. Sci. Results, 107, 3-26 (1990). 39. V. Kova!i", Buj!tina 98, 13-16 (1988). 40. V. Kova!i", MG (#asopis muzealaca i galerista Istre) 2, 2, 10-11 (1996). 41. G. Labud, AMSI LXXXIX, 7-17 (1989). 42. K. Lambeck, C. Smither and P. Johnston, Geophys. J. Int., 134, 102-144 (1998). 43. K. Lambeck, and J. Chappel, Science, 29, 679-686 (2001).

155

44. K. Lambeck, Glacial Isostatic Adjustment and the Earth System, (J.X. Mitrovica and B.Vermeersen, Eds.) American Geophysical Union, Washington DC, 33-50 (2002). 45. K. Lambeck, Geophys. J. Int., 122, 1022-1044 (1995). 46. K. Lambeck, Y. Yokoyama and A. Purcell, Quaternary Science Reviews, 21, 343-360 (2002). 47. K. Lambeck, A. Purcell, P. Johnston, M. Nakada and Y. Yokoyama, Quaternary Science Reviews 22, 309-318 (2003). 48. K. Lambeck, F. Antonioli, A. Purcell and S. Silenzi, Quaternary Science Revue, 23, 1567-1598 (2004a). 49. K. Lambeck, M. Anzidei, F. Antonioli, A. Benini and E. Esposito, Earth and Planetary Science Letter, 224, 563-575 (2004b). 50. K. Lambeck and A. Purcell, Quaternary Science Reviews 24,1969–1988 (2005). 51. K. Lambeck, A. Purcell, S. Funder, K. Kjær, E. Larsen and P. Möller, Boreas, 35(3), 539-575 (2006). 52. E. Mantovani, D. Albarello, C. Tamburelli and D. Babbucci, J. Geodyn., 21, 35-72 (1996). 53. E: Mantovani, N. Cenni, D. Albarello, M. Viti, D. Babbucci, C. Tamburelli and F. D’Onza, J. Geodyn., 31(5), 519-556 (2001). 54. F. Maselli Scotti, AMSI LXXVII, 340-341 (1977). 55. F. Maselli Scotti, AAAd XV I, 345-381 (1979). 56. R. Matija$i", AAAd XLVI, 161-174 (2001a). 57. R. Matija$i", AAAd XLIX, II, 693-711 (2001b). 58. S. Medas, Boats, Ships and Shipyards. Proceedings of the Ninth International Symposium on Boat and Ship Archaeology, Venice 2000, C. Beltrame eds, Oxford, 42-48 (2003). 59. C. Meletti, E. Patacca and P. Scandone, Pure Appl. Geophys. 157, 11-35 (2000). 60. P. Montone, A. Amato and S. Pondrelli, J. Geophys. Res. 104, 25 595-25 610 (1999). 61. Museo Muggia 1997. F. Maselli Scotti (ed.), Il Civico Museo Archeologico di Muggia, Trieste (1997). 62. J.M. Nocquet and E. Calais, Geophys. J. Int. 154, 72–88 (2003). 63. J.S. Oldow, L. Ferranti, D.S. Lewis, J.K. Campbell, B. D’Argenio, R. Catalano, G. Pappone, L. Carmignani, P. Conti and C.L.V. Aiken, Geology, 30(9), 779-782 (2002). 64. P. Paronuzzi, Atti e Memorie della Società Istriana di archeologia e storia patria XXXVI, 215-232 (1988). 65. M. Peracca, Mostra protostorica e romana di Muggia, Trieste (1968). 66. P. Piani, AVen IV, 115-132 (1981). 67. P.A. Pirazzoli, Annales de l’Institut Oceanographique 56, 101-111 (1980). 68. P.A. Pirazzoli, Wiley, Chichester, 211 (1996). 69. P. Pomey, Boats, Ships and Shipyards. Proceedings of the Ninth International Symposium on Boat and Ship Archaeology, Venice 2000, C. Beltrame eds, Oxford, 57-65 (2003).

156

70. C.D. Reuther, Z. Ben-Avrham and M. Grasso, Terra Nova 5, 234-257 (1993). 71. I. Schrunck and V. Begovi", JRA 13, 253-276 (2000). 72. E. Serpelloni, M. Anzidei, P. Baldi, G. Casula and A. Galvani, Geophys. J. Int., 161, 861 – 880, doi:10.1111/j.1365-246X 2006.06218 (2005). 73. V. %ribar, AVes 9-10, 275-276 (1958-59). 74. V. %ribar, Razprave in poro$ila 10, 271-274 (1967). 75. J.R. Steffy, Israel Antiquities Authority, Jerusalem, 29-47. (1990). 74. S. Stiros, Journal of Structural Geology 23, 545-562 (2001). 75. M. Stokin, ArhPregl 26, 93-94 (1986). 76. F. Stravisi and N. Purga, Rapp. Oceanogr. Meteor. 112, 10 (2005). 77. I. Veli", J. Ti$ljar, D. Mati!ec and I. Vlahovi", 80° Riunione estiva Trieste della Società Geologica Italiana, 6-8 settembre 2000, 237-245 (2000). 78. D. Vrsalovi!, Arheolo!ka istra"ivanja u podmorju isto$nog Jadrana, prilog poznavanju trgova$kih plovnih putova i privrednih prilika na Jadranu u antici, Zagreb (1979). 79. S.N. Ward, Geophys. J. Int. 117, 441-452 (1994). 80. R. Westaway, Earth Plan. Sci. Lett. 96, 393-406 (1990). 81. M. &upan!i!, Atti XX, 381-392 (1989-1990).

157

Fig 1. Map of the Northeastern Adriatic Sea showing the location of the archaeological and geomorphological markers sites investigated in this paper.

158

Fig 2. Sketch of the method used for the archaeological measurements and the concept of functionality.

Fig 3. Map of the Eastern Adriatic coast. The legend contains the locations where the submerged tidal notches were measured by the authors.

159

Fig 4. A: The submerged tidal notch at -2.2 m, Duino (Trieste, Italy). The notch amplitude is larger than 1.0 m in accordance with the local tide amplitude. B: The submerged tidal notch has been surveyed at -0.8 m at Rovinj (Croatia).

160

Fig 5a. Cross sections of the archaeological sites in the studied area and their relationships with the current and past sea level. 1, Stramare; 2, Punta Sottile; 3, San Bartolomeo.

161

Fig 5b. Cross sections of the archaeological sites in the studied area and their relationships with the current and past sea level. 4; San Simone; 5 Salvore; 6, Briunj.

162

Figure 6A. Comparison of predicted model results with observational evidence in the NE Adriatic coast.

Figure 6B. Same as figure 6A but on expanded scale.

163

Figure 7. Photos of the sites described in this paper. A: Brijuni, site xa of Table 1. B: Brijuni, site xb of Table 1. C: Measuring the Sv. Simon pier, site x of Table 1 D: Measuring the Salvore pier, site xb of Table 1

164

DIGITIZATION AND MULTISPECTRAL ANALYSIS OF ARTISTIC OBJECTS : EXEMPLARY CASES AND WEB DOCUMENTATION* GIUSEPPE MAINO AND SILVIA MASSARI † ENEA, 5, via Martiri di Monte Sole, 40129 Bologna, Italy, and University of Bologna, Faculty of Preservation of the Cultural Heritage, 5, via Mariani, 48100 Ravenna, Italy Results of multispectral analyses are presented, performed on paintings and ancient books, thus elucidating the execution techniques and the conservation status. Moreover, a multimedia database is described, used for documentation on and off-line of these and analogous studies.

1. Introduction A multispectral digital system, recently developed at the ENEA laboratories in Bologna and applied to the investigation of many artistic and archaeological works, is presented, ranging from infrared radiation to visible light and ultraviolet fluorescence, in order to perform suitable analyses of paintings, frescoes, illuminated codes, parchments, books and documents in historical archives, etc., preliminary to any restoration or cleaning. Relevant software for multispectral image analysis and digital restoration has been developed associated with and complementary to this hardware system and applied to a cases of main historical interest, namely the incunaboli and cinquecentine in the library of Minori Osservanti in Bologna and the XVI century books in the Library of Padri Minimi of Paola in Calabria (Italy) and paintings by Marco Zoppo, Lianori, Vasari, Lorenzetti, Gandolfi and Raphael. We also describe the implementation and validation of a multimedia database for archiving information about diagnostics, conservation and restoration of historical and artistic objects. The software architecture is based on a Content Management System (CMS) and allows the development of a dynamic website. This information system allows us to provide a update and easily searching engine for documentation of the work so far performed and for comparison among different analyses carried out, for instance, on paintings by the same author or of the same period of time.

* †

This work is supported by NEREA project, funded by Regione Emilia-Romagna, PRRIITT. e-mail: [email protected] ; [email protected].

165

2. The GIANO project for cultural heritage Once completed the ENEA international project, GIANO (Innovative Graphics for National Cultural Heritage and young people’s Occupation) the necessity has risen to let public know its many results, creating a dedicate website. GIANO’s main purpose was to plan, set up and validate a software application (concerning innovative graphics, virtual reality simulations, linked databases) for keeping records and documentation about relevant historical and artistic cultural assets, using dedicate hyper-textual and multimedia methods. Three main applications have been implemented – for demonstrative purposes – whilst developing the project. All of them are characteristics of a wide range of cultural assets: • Libraries and historical archives, especially in Calabria and Sicily, mainly important for the existence of inedited documents related with Bisanzio presence in the Southern and Insular Italy; • Diagnostic imaging and restoration reports (including written and photographic reports) of historical and artistic assets; • Mediterranean wall mosaics (IV-XIV A.D.). All these results, useful for scientists and conservators, but also remarkable for interested people and tourists, must be available in a simple and effective way. At the moment only two ways are available to organise big size websites: • Collection of documents (hundreds or thousands) consultable by the public; • On-demand database related applications that use a dynamic way to show multimedia documents and data. The first solution requires complex, long and expensive maintenance; moreover, it is not suitable in a dynamic situation where information changes frequently. In fact, it could be very difficult to maintain the data consistency, while their updates should be done by dedicated people, with a relevant consequent expenditure of time and human resources. On the other hand, the second scenario matches more with the above situation. It consents to produce a large number of web-pages, re-using graphic components, maintaining distinct interface layouts and developing the code to recover web-pages data. Moreover, it allows a proficient management of all human resources involved in the project. The main aim of this project was therefore to create databases to hold and integrate all GIANO results: For this reason is important to implement applications that let users know the data sources, producing on-demand documents. 3. Multispectral non-destructive analyses An interesting example of multispectral investigation performed within the GIANO project is represented by the diagnostic analyses on the panel of St.

166

Giovanni Evangelista by Pietro Lorenzetti, kept at the Museum ‘Amedeo Lia’ in La Spezia (see figure 1), and attributed to this important artist by Federico Zeri in nineteen- sixty eight.

Figure 1. The panel representing St. Giovanni Evangelista in the ‘Amedeo Lia’ museum by Pietro Lorenzetti.

3.1. History of the painting This panel, that was part of a famous split polyptych, had to be opposed, according to the traditional tipology of similar paintings, to a St. Giovanni Battista that probably was at the other side of the panel. Two parts of this panel represent St. Caterina d’Alessandria at the Metropolitan Museum in New York and St. Margherita at the Mason Perkins museum in Assisi. Moreover, other splitted parts are the two sides of the Madonna with the child coming from the Loser heritage at Palazzo Vecchio in Florence. A probable reconstruction of the polyptich sees the Madonna in the middle, the two saints at her sides (St. Margherita is on the left, and St. Caterina d’Alessandria is on the right) and at the far sides the Saints Giovanni Evangelista and Giovanni Battista, whose panel is missing. Federico Zeri noted that on the upper side of the panel of St. Caterina d’Alessandria in New York there was a notice saying “S. AGNES” and that on the St. Giovanni panel in La Spezia there was an abrasion, so we can presume that there was the same writing or something similar, reminding a second upper order. We have a few more elements to prove that there was a second order above the first one: A Saint Bishop, that is in the collection De Noailesse at Fontainebleau, and two spires with a Saint Martyr and St. Antonio Abate, both in the National Gallery in Prague.

167

3.2. Results of the multispectral analysis Figure 2 shows the ENEA team at work in ‘Amedeo Lia’ museum during the experimental multispectral investigation performed with the instrumentation described in ref.1.

Figure 2. The multispectral apparatus operating in the ‘Amedeo Lia’ museum on the Lorenzetti painting.

Figure 3.The head of St. Giovanni Evangelista in visible light as result of mosaic of many partial frames.

Many images of small parts of the painting have been grabbed at different wavelength and then assembled by means of a suitable computer program to obtain multispectral images of the whole painting. Analyses by infrared radiation, ultraviolet fluorescence and visible light were made using the multispectral digital system MUSIS two thousand and seven. Figures 3 and 4 show a detail of the panel in visible and infrared radiation, respectively. It is possible to recognize in figure 3 the preliminary drawing that, compared with other works by Pietro Lorenzetti, confirms the attribution by Zeri to this artist. Moreover, under the abrasion there are marks of a writing of which we can read only the initial S.

168

Figure 4.The head of St. Giovanni Evangelista in infrared reflectography as result of mosaic of many partial frames.

Finally, figure 5 shows the panel in ultraviolet fluorescence, where the repaintings made by preliminary restorations are identified by darker regions.

Figure 5.The head of St. Giovanni Evangelista in ultraviolet fluorescence as result of mosaic of many partial frames.

4. The multimedia database system In order to carry out our website we followed the standards of quality suggested by the Italian Ministry of the Cultural Activities and assets within Minerva Project. Experts in cultural and computer areas are involved with the important task of codifying common rules for spreading digitalized cultural assets all over Europe. The relevant guidelines can be found on the site www.minervaeurope.org: Its high standard level is accessible to everyone and what is more important, is available to everybody, special people included.

169

Three main applications have been implemented – for demonstrative purposes – whilst developing the project. All of them are characteristics of a wide range of cultural assets: a) Libraries and historical archives, especially in Calabria and Sicily, mainly important for the existence of inedited documents related with Bisanzio presence in the Southern and Insular Italy; b) Diagnostic imaging and restoration reports (including written and photographic reports) of historical and artistic assets; c) Mediterranean wall mosaics (IV-XIV A.D.). Contents are so organized on the base of common categories (metadata), using a data system called “CMS” (that is Content Management System). This name suggests that CMS are: • Software; • Skills, knowledge and techniques necessary to build and manage this kind of software in a hyper-textual way (all documents are available on internet), with precise communication standards (priority, visibility, etc.). A Content Management System builds and updates a website, managing all phases: Setting, editing, publishing texts, images and sounds. Moreover, if it would be useful in a portal, it should classify and organize all the information to easily find, implement, modify and link them or to re-use them in a different part of the website: The bigger is the website, the more important is that the CMS is flexible and efficient. The mainly characteristics of a CMS are therefore: • User-friendly interface; • Fast possibility of inputting, modifying and finding information; • Capability to adapt to frame and graphic website need; • Safe and flexible usability; • Interface uploading via browser; • Use of graphic template for showing contents; • Manage of different customer roles and workflow; • Database for images, texts and graphs; • Find and integrate information from other sources; • Manage mailing lists and mail boxes; • Manage and order links, news, FAQ, events; • Searching usability; • Customise graphic contents. Some of more interesting characteristics of our website will be then shortly shown and discussed. The on-line system we used is called “Museo and Web”: it can be dowloaded with an open source license. We modified it adding new modules and adapting the database to our contents. The CMS has a “back-end” or administrative section, that permits the site organization and management, as well as the contents upload, change and/or cancellation; and a “front-end” the real site, that allows users to look up for contents. By this way the contents can be created, edited, translated and filed by

170

many more users at the same time and independently, allowing a cooperative management of all the information. After installing the software, you have to login, in order to enter the administrative section. There are three groups of users who can enter this section: 1. The administrator who have a complete control on the section management, 2. The supervisor, who have less control; 3. The editors, who can’t publish contents but only write proofs. Now, entering the administrative section, there are some modules to manage contents: managing the site, the projects, the publications, the printing review, etc.. Each module consists of sub-modules. The site structure is the frame of our web programme, a sort of general index of contents and its management is carried out trough an index. Next to each diagram node there are some buttons that permit, when pushed, to create, modify or cancel any structure elements. Moving to the front-end in the site home page, one finds the project-logo that introduces the real site, the meta-surfing (top right side), offering functions like “search”, “map” and “multilingual choices”, as well as the main surfing repeated on every page, where the users can find acronyms of the main projects, the works, the places or monuments studied. Moreover, there is a part, in the home page, dedicated to the interactive functions on the site (the login for the access to special pages, to the newsletters registration, forum). The main events that can stimulate the visitor interest are on the left vertical bar. Inside the GIANO MENU you can look at : 1. The presentation of the project (intervention areas), objectives, collaborations, different types of works and materials) 2. Lectures: the ones organized by GIANO project or those in which it took part; it’s possible to do researches according to titles, years, places, categories. 3. GIANO publications: It is possible to perform researches according to titles, topics, descriptions. Pushing on the research result, it’s possible to visualize the complete description and download files. 4. The second term of the menu is the most important: It allows to enter the various activities, to see the results, etc.. For example, entering the research projects area, one can see a short presentation of the three macro-areas, of which the GIANO Project is composed: ! CONSERVATIO (specific databases creation); ! ZIKKARON (a Jewish term meaning memory, used for the creation of a prototype system, useful to transfer ancient texts in digital format with a graphic-textual methodology; used for historical libraries in Calabria (such as the library of Padri Minimi in Paola)and to take a census of inedited, file sourced about Bizantine influence in Southern Italy and Isles; ! TECHNE’ (innovative diagnostic technologies for a deep knowledge of the works of art).

171

For example, entering TECHNE’, one can find a short list of all the activities (Lianori, Vasari, Marco Zoppo..); pushing one of them (for example the panel Crocifissione by Orazio di Jacopo) there will be access to the page containing all the details of the activities, the textual description of the researches and relative images, and it will be possible to see the complete work , thanks to the categories linked together in the database. Moreover, there is the possibility to read whatever one wants about the artist (biography, works of art, contemporary artists, etc..) and to overview the file dedicated to the Museum, owner of the work of art (in this case, at the Museum of Osservanza in Bologna, with touristic, historical information and with reference to all the works contained in our database and belonging to the museum. From the place or the monument, one has access to the itinerary where the monument is located (for example for the Museum of Osservanza in the itinerary Sacred Art) and see the other monuments studied during the project and included in the same itinerary. The site is still a work in progress but it will be possible to visit it soon at www.bologna.enea.it/giano. References 1. G.Maino, S.Bruni, S.Ferriani, A.Musumeci and D.Visparelli, Multispectral analysis of paintings and wooden sculptures, in Proceedings of II Congresso Nazionale AIAr Scienza e Beni Culturali, Patron Editore, Bologna (2002) 203.

172

ACTUOPALAEONTOLOGY: A POLYFUNCTIONAL TOOL FOR ARCHAEOLOGY BRESSAN G1., FONDA G.2, KALEB S.1, MELIS R.3, MONTENEGRO M.E.4, MOURGUIART P.5, PUGLIESE N.3,4, RICCAMBONI R.3, RUSSO A.6, SODINI N.7, TROMBA G.7 1 Dipartimento di Scienze della Vita dell’Università degli Studi di Trieste, via L. Giorgieri 10, 34127 Trieste, Italia. 2 Via dei Leo 10, 34100 Trieste, Italia. 3 Dipartimento di Scienze Geologiche, Ambientali e Marine dell’Università degli Studi di Trieste, via Weiss 2, 34127 Trieste, Italia. 4 Museo Nazionale dell’Antartide, Sezione di Trieste, via Weiss 2, 34127 Trieste, Italia. 5 IRD, 213 rue La Fayette, 75480 Paris cedex 10, France. 6 Dipartimento di Paleobiologia e dell’Orto Botanico dell’Università degli studi di Modena e Reggio Emilia, via dell’Università 4, 41100 Modena, Italia. 7 Sincrotrone Trieste, S.S.14, km 163.5 34012 Basovizza, Italia. Actuopalaeontology is the logical synthesis between palaeontology and biology. The goals of actuopalaeontology are: i) to emphasize the role of the palaeontological disciplines within the archaeological research; ii) to realize a palaeontological guide for both, archaeologists and researchers from other disciplines, in other words, nonspecialists; iii) to reconstruct natural and/or anthropized scenarios: palaeoenvironments and palaeoclimates. In particular, this work plans to discuss actuopalaeontology’s role in geoarchaeological research, following the points of view of both archaeologist and actuopalaeontologist. The archaeologist has to obtain by him-self preliminary taxonomic and environmental observations to subsequently involve the right specialist. He should be able to recognize the organic remains found in archaeological excavations and boreholes through a first simple identification and, subsequently, obtain a preliminary palaeoenvironmental interpretation, using a table reporting the life-environment of the most common organisms. The actuopalaentologist has to better define the preliminary interpretations of the non-specialist evidencing the precise taxonomic aspects and the right palaeocological and palaeoclimatic results. The actuopalaeontologist has to interact with other disciplines; moreover, he should be able to perform specific analyses like microtomography, SEM and isotopic geochemistry to refine the palaeoenvironmenalt and the palaeoclimatic interpretations.

1. Introduction Actuopalaeontology is the logical synthesis between palaeontology and biology. It concerns the study of the modern organisms which present structures that will fossilize along the time, focusing the attention on some of their morphological features showing similar adaptations to a given environment. Thus,

173

actuopalaeontology represents the obvious product of the taxonomic uniformitarianism (sensu DODD & STANTON, 1991). Earth sciences, including actuopalaeontology, show frequent and new applications towards the knowledge and exploitation of the cultural heritage. In particular, actuopalaeontology may represent a precious tool in the archaeological research to define the scenarios where men lived, evidencing the natural and anthropic processes that have determined and controlled their evolution. Thus, actuopalaeontology deals with fossils of organisms that are recent dwellers of environments located near human settlements (lakes, swamps, coastal settings, lagoons, etc.). This research intends to evidence the potentiality of this discipline towards a very important scientific sector interesting the cultural heritage represented by geoarchaeology. Premises of this work are: - the actuopalaeontologist should be able to recognise the fossils recorded in the archaeological excavations or boreholes: thus, he should demonstrate a good taxonomic knowledge; - the actuopalaeontologist should be able to link these fossils to well defined environments: thus, he should be a palaeoecologist focusing his attention on adaptive life-strategies of organisms; - the actuopalaeontologist should correlate fossils to well defined palaeoclimatic conditions; thus, he should be able to highlight climatic aspects, also considering the (palaeo)biogeography, geochemistry and other geological disciplines; - the actuopalaeontogist must respect the cultural goods: thus, if possible, he should perform analyses without damaging the materials.

-

-

This work plans to discuss these premises following two points of view: a) the point of view of the non-specialist, who should be able to perform the preliminary observations and interpretations in order to subsequently involve the specialist; b) the point of view of the specialist (actuopalaeontogist), who will better define and refine the preliminary interpretations of the non-specialist. Thus, these points of view a) and b) will be discuss in each section below.

2. Taxonomic definition The taxonomic definition of the organisms found in the archaeological sites is essential to construct all the interpretations of the geoarchaeological research.

174

a) Taxonomy is the first and main problem for the non-actuopalaeontologist. Actually, fossils are very frequently discovered by the non-specialists, who need to make a preliminary description of the same, at least. Today, the organisms that are potential fossils are very numerous: without counting, one must think to include in this fossil stock all the vegetal and animal species provided of hard (inner or outer) skeletal parts consisting of calcite, aragonite, silica, phosphate, etc. Thus, to become a taxonomist is the first problem. A methodology based on the shapes of the organisms and on a simple dichotomous key is proposed, in order to reach this preliminary taxonomic definition. Fig. 1 represents a method to obtain the taxonomic definition, based on the shapes. The identification-key proposes a series of consequential questions to the non-specialist who has found an “unknown” form (Fig. 2) in his sample. The reader has to follow the scheme from left to right, up to the final pictures which show the solution of identification. He has to answer the following questions: FIRST QUESTION: is the organism constituted by a skeleton consisting of one or more than one element? ANSWER: only one! SECOND QUESTION: kind of shape? ANSWER: conical! THIRD QUESTION: kind of growth? ANSWER: coiled! FOURTH QUESTION: kind of coiling? ANSWER; helicoidal! FINAL RESULT: GASTROPOD. Most organisms can be preliminarily recognised. Moreover, a series of index-cards will represent a useful actuopalaeontological guide, which shall also includes boxes to identify the significant morphological characteristics of the discovered organisms. Once reached a preliminary and precise autonomous taxonomic definition, the non-specialist has to contact the right specialist for detailed analyses. b) The omniscient palaeontologist, actuopalaeontologist or biologist does not exist, we hope! He normally is specialised in the taxonomic definition of few animal/vegetal groups of organisms. The most important tool is his experience. The specialist will be able to use a suitable scientific literature to correctly define species and assemblages of species. Thus, he may confirm or correct the preliminary taxonomic definitions given by the non-specialist. However, some specialists are also proposing simplified but complete taxonomic keys to widely divulgate their speciality. This is the case of the key of the red calcareous algae (Fig. 3) proposed by BRESSAN & BABBINI (2003).

175

With the help of this key, the algologist may guide the non-specialist towards a good taxonomic definition using well defined simple features (thallus architecture). FIRST QUESTION: what is the external (macroscopic) aspect of this organism between these five chances? (see “home page” of the identification key). ANSWER: C one! Free-living thalli: cylindrical, formed by small branches, or subspherical with lumpy protuberances; (see table 8 in Fig. 3) SECOND QUESTION: what is the macroscopic and microscopic aspects of this organism between these four chances? ANSWER: A one! Cylindrical thalli, formed by free-living branchlets, simple or variously ramified; and tetrasporangial conceptacles multiporate, sunken; (see table 9 in Fig. 3.) THIRD QUESTION: what is the macroscopic (i) and microscopic (ii) and submicroscopic (iii) aspects of this organism between these two antinomic chances? ANSWER: A one! i - branches diameter: up to 1.5-2mm; ii - epithallial cells: flared; iii - cells of branches: medulla: rectangular; (15) 20-25 (30)µm long, 7-12µm in diameter; cortex: rectangular-tapered; 20-25µm long, 5-8µm in diameter. …..the morphometric data shall be compared with dispersion graph too. FINAL RESULT: Lithothamnion corallioides (see the synthesis of the characterization of this species: vegetative features, reproductive features, ecology; bathymetric and geographical distribution; photo-gallery). 3. Palaeoenvironmental interpretations Most archaeologists need to know the environmental evolution of the studied site in order to have a complete panorama of the events which took place in and around the archaeological site. The beds crossed by the excavations or boreholes are good archives of natural or man-related events. The actuopalaeontologist has to highlight these events through the analysis of the fossils: in this way, he can follow the evolution of wetlands, the changes of the shore-lines, the dramatic end of human settlements and so on. a) Once defined the examined material from the taxonomic point of view, the non-specialist may deal with a related preliminary environmental interpretation. An example is represented by a washed material coming from a hypothetical excavation (Fig. 4). Fig. 5 represents a scheme showing the typical life-environments of the main organism groups. The researcher can highlight the stripes corresponding to the identified organisms . In this

176

case, he recognised gastropods, ostracods, characean algae and thecamoebians. This way the identified assemblage suggests a fresh water/slightly brackish water environment. b) Although probably correct, this environmental interpretation must be controlled by the specialist. Actually, the non-specialist does not always possess the sensitivity and the experience to define the autochthony or the allochthony of the studied specimens. The actuopalaeontologist is able to evidence the link which relies the organisms to a given environment highlighting their life-strategies (modes of life, feeding, etc.). He distinguishes the autochthonous and the displaced forms in relation to their preservation state, population composition and ontogenetic cycle. Well preserved shells might be probably autochthonous; vice versa badly preserved shells are very probably displaced. Some other organisms (for example, ostracods) show a growth by moulting, abandoning the previous and smaller carapaces; other forms (for example, mollusc bivalves and brachiopods) present a continuous growth as demonstrated by shells with concentric growth lines. In the former case, the finding in the sample of differently-sized carapaces might demonstrate that the form has accomplished its life in that environment: it has reached the adult stage abandoning its former and younger carapaces in that environment. In the latter case, it is presumable that a bivalve species is in situ if it is represented by a complete shell (right and left valves), even if disarticulated. On the contrary, the exclusive finding of right or left valves may suggest a transport due to currents. Additional palaeoenvironmental observations may derive from the analysis of the morphological features of the organisms. In this case, the taxonomic uniformitarianism is essential. Shape, ornamentation, size and thickness of outer skeletal parts (shells or carapaces) represent precise adaptations to a given environment. For example, strictly benthic ostracods present subrectangular carapace, both in frontal and dorsal view; their thickness depends on the degree of hydrodynamism of the bottom (fragile carapace=calm water; thick carapace=high energy water); the ornamentations (wings, spines) are prominent on bottoms of calm waters to avoid the burial in anoxic deposits, and absent in high energy settings to avoid mechanic breaking; the size is usually very reduced in calm waters and large in high energy bottoms; etc. Therefore, the specialist is able to correctly define the composition of the assemblages and to give an accurate environmental interpretation.

177

4. How to become a palaeoclimatologist? Archaeologists sometimes need to know the climatologic evolution of a geographic area. Historic sources may be very important as well as the palaeontological data. Cold or warm episodes are usually evidenced by well defined organisms. For example, penguins recorded in Quaternary Mediterranean deposits indicate cold phases. On the contrary, hippopotamus bones of some Quaternary breccias reveal warm episodes. Pollen and spores analyses are some further essential tools to reveal these climatic variations. For example, samples containing conifer pollens may testify cold episodes; vice versa, tropical pollens may obviously attest warm climate. Effectively, this analysis is not easy for a non-specialist, since it requires a detailed taxonomic knowledge of plants and animals. However, the actuopalaeontologist has to sensitise the non-specialist since the palaeoclimatological background is very useful to better explain sea-level changes, evolution of alluvial plain and deepen the economic development of a given human settlement. For example, a small microcrustacean (ostracod) was recorded in some recent Aquileia stratigraphic units. Nowadays, this species is absent at this latitude, but it lives in north-central aqueous environments. It indicates a cold episode during the 4th-5th century A.D. Another very important tool to highlight the climatic changes is the isotope geochemistry. As demonstrated in several papers, the ratio of the isotopes of some elements in the organism shells may evidence the climatic fluctuation (O18/O16) , together with very detailed environmental data. For example, the geochemical data obtained from the ostracod valves are widely used. TURPEN & ANGELL (1971) demonstrated that the ions required for the construction of the valves are extracted from the water. CHAVE (1954) and CADOT & KAESLER (1977) have shown that the chemical composition of valve’s carbonate is in relation to the temperature of the water. BODEGART & ANDREANI (1981) have proved that this composition depends also on water chemistry. CHIVAS et al. (1983) proved that the concentrations of Mg of the carbonates of the valves depend simultaneously on the temperature and on the Mg/Ca ratio of the water, thus, the Mg/Ca ratio recorded in the ostracods valves can be used as an indicator of the variations of the palaeotemperature and the palaeosalinity of the water at the time of the deposition of the valves. They have also confirmed that the absorption of Sr++ is actually independent from temperature, but linked to the Sr/Ca ratio of the water. In consequence, the relation Sr/Ca of the valves reflects the salinity of the water. Moreover, the stock of the ∑ CO2 of the water depends on two factors: a) the exchange with the atmospheric CO2. b) the atmospheric production/consumption ratio of the organic matter in situ.

178

Therefore, the concentrations of the isotopes of oxygen and carbon in the water are in relation to the atmospheric concentration of these isotopes. Consequently, the potentiality of these data is very high, since they allow us to refine the reconstruction of the ancient scenarios, as already demonstrated in the palaeoenvironmental research performed by one of us (M.E. MONTENEGRO) in several sites of archaeological interest (Titicaca region, etc.) 5. Without damaging the material… Cultural heritage, including fossils, deserves to be respected and protected. Actuopalaeontological research is usually performed with non dramatically destructive methods. However, sediment samples coming from excavations or boreholes must be treated and analysed in laboratory. Simplifying, this phase requires the elimination of the finest grain-size fraction (washing through specific sieves) and the destruction of the organic matter (hydrogen peroxide treatment). The final product is represented by the micro or macroscopic material that will be analysed by the actuopalaeontologist. New methodologies allow us to analyse the material without previous and dangerous treatment. Recently BRESSAN et al. (2007) have introduced a modern non-destructive approach using 3D X-ray microtomography before SEM analysis. The goal of this study is to easily identify the important features which are not evident without breaking the thallus of the plant. Actually, red algae are now increasingly identified within the archaeological research in Mediterranean areas. Thus, the algae may acquire progressively more importance in defining the ancient marine scenarios of archaeological interest. Since they are often encrusting organisms on manufacts (for example, necks of Roman amphoras), it is very important to avoid breakings and damages. Computed microtomography (µ-CT) satisfies this requirement: it is one of the most advanced techniques in the field of non-destructive evaluation tests. It allows imaging of the internal microstructure of the rock and soil pore space, by measuring the three-dimensional X-ray attenuation coefficient map of the sample. Thus, red algae may give environmental and climatic data that become very useful in these sites. 6. Final considerations The actuopalaeontologist has to guide the non-specialist towards preliminary interpretations. Basic tool is a right identification of the organisms, which is propaedeutic for further environmental and climatic interpretations. Therefore,

179

next step shall concern the planning of an actuopalaeontological guide to mainly address the archaeologists towards these simple interpretations for geoarchaeological research. In this way, this purpose should meet the numerous archaeologists’ requirements. Providing preliminary data, this guide should address the archaeologists towards the collaboration with specialists of palaeontological and geological disciplines in general. Therefore, the reconstructions of the ancient environments, where men lived, may become more refined, also with the integration of the results of the geochemical analyses, which provide a climatic framework of the area. With this method, actuopalaeontology and in general geological disciplines might present an important role, not only for the palaeoenvironmental and palaeoclimatic interpretations, but also by the proposition of non-invasive methodologies for the better preservation and exploitation of the cultural heritage. References BODERGAT A.M. & ANDREANI A. M. (1981). Mise en évidence de la réponse adaptative d'une espèce euryhaline Cyprideis torosa (JONES, 1850) à des conditions écologiques difficiles par l'analyse multi-élémentaire en spectrometrie de masse à étincelle; International Symposium on Concept and Method in Paleontology : contributed papers, Barcelona, 5–8 May 1981 / edited by Jordi Martinell. Universidad de Barcelona, Departamento de Paleontologia, 135–140. BRESSAN G. & BABBINI L. (2003). Corallinales del Mar Mediterraneo: guida alla determinazione. Biol. Mar. Medit. 10 (2) :1–240. BRESSAN G., FAVRETTO S., KALEB S., TROMBA G. & VITA F. (2007). Applicazione della microtomografia computerizzata a raggi X allo studio predittivo della struttura di alghe rosse calcaree, XXXVIII Congresso Società Italiana di Biologia Marina, 28 maggio – 2 giugno 2007, S. Margherita Ligure (GE). BRESSAN G., FAVRETTO S., KALEB S., TROMBA G. & VITA F. (2007). X–Ray microtomography application to a predictive evaluation of coralline algae structure. Elettra Synchrotrone Research Highlights (Bioscience and soft matter): (science update series): 28–29.

180

CADOT H.M. & KAESLER R.L. (1977). Magnesium content of calcite in carapaces of benthic marine Ostracoda. University of Kansas Paleontological Contributions, Paper 87, 1– 23. CHAVE K.E. (1954). Aspects of the biogeochemistry of magnesium. 1. Calcareus marine organisms; Journal of Goeology, 62: 266–283. CHIVAS A., DE DECKKER P. & SHELLEY J.M.G. (1983). Magnesium, strontium end bariumpartitioning in nonmarine ostracode shells and their use in paleoenvironmental reconstructions - A preliminary study. In: Maddocks, R. F. (ed.) Applications of ostracoda. 8th International Symposium on Ostracoda. University of Houston Geosciences, Houston: 238–249. DODD J.R. & STANTON R.J. (1990). Paleoecology, Concepts and Applications, 2nd Edition, J. Wiley Ed.: 502 pp. TURPEN J. & ANGELL R. (1971). Aspects of molting and calcification in the ostracode Heterocypris. Biological Bulletin, 140: 331–338.

Acknowledgments The authors thank DR. L. BABBINI, DR. S. FAVRETTO AND DR. F. VITA for their useful advice in preparing this work.

181

FIGURES:

Fig. 1: the identification-key: a method to obtain the taxonomic definition, based on the shapes.

Fig. 2: the “unknown” form.

182

Fig. 3: the key of the red calcareous algae proposed by BRESSAN & BABBINI (2003).

183

Fig. 4: washed material coming from a hypothetical excavation. 1-ostracod; 2-gastropod; 3thecamoebian; 4-characean alga (girogonite).

184

Fig. 5: scheme showing the typical life-environments of the main organisms groups. Red highlight indicate the organisms found in the sample and their related life-environment. This way, the deposit of the excavation (see Fig. 4) belongs to fresh water or slightly brackish water environments.

185

Fig. 6a: a slice performed at the SYRMEP beamline of ELETTRA: the high resolution of this image highlights a peripherical zonation produced through a periodical calcification (see the bands).

186

Fig. 6b: preliminary non-invasive microtomography slice to choose our target: this predictive method produce a lot of slices that allow us to orientate the very small micro biopsy to the subsequent not destructive ultra structural analyses at the Scanning Electron Microscope (SEM).

Fig. 7: a SEM digital photo of an epibiosis phenomenon highlighted by means a microbiopsy (see Fig. 6b - circled area). You can see easily four layers of organisms as: diatoms, serpulides, coralline algae, cyanobacteries.

187

ROBOTICS TOOLS FOR UNDERWATER ARCHAEOLOGY G. CONTE, S. ZANOLI, D. SCARADOZZI, L. GAMBELLA Dipartimento di Ingegneria Informatica, Gestionale e dell’Automazione Università Politecnica delle Marche via Brecce Bianche - 60131 Ancona - Italy The use of remotely operated vehicles and of automatic data gathering and processing techniques can provide new tools and methods for the investigation of underwater archaeological sites. This paper describes part of the work done in this direction in the framework of the European research project VENUS!, focusing on the development of efficient procedures for using Remotely Operated Vehicles in exploring and mapping underwater archaeological sites.

1.

Introduction and motivation

Underwater archaeology can provide valuable information about practically all aspects of life and organization of the societies that developed maritime activities or interacted in some way with the marine environment. Unfortunately, the study of submerged archaeological sites, mainly wreckage sites, is made difficult by the harsh characteristics of the environment. In traditional marine archaeological surveys, on-site data collection is performed by divers and it implies manual recording of a large number of measures and pictures. The investigators have to return repeatedly to the same location and the whole process is expensive, demanding, cumbersome and time consuming. A possible way to overcome these difficulties consists in developing tools and methods that allow to employ Remotely Operated Vehicles (ROV), originally conceived for the off-shore industry, in the exploration of submerged archaeological sites [5, 6, 7]. ROVs can substitute divers in taking pictures and videos of underwater sites, with little or no modifications of current practices and procedures, increasing productivity and reducing, at the same time, risk and labour for human operators. Since ROV’s can reach depths which are beyond divers’ capability, can work for long periods of time in unfavourable conditions !

Work partially supported by the European Community under project VENUS (Contract IST034924) IST Programme 6th FP for RTD". The authors are solely responsible for the content of this paper. It does not represent the opinion of the European Community, and the European Community is not responsible for any use that might be made of data appearing therein.

188

and can easily carry heavy and sophisticated equipment, the advantages of their use, also at this elementary level, are obvious. However, ROV’s and the related technology can be better exploited for developing and implementing new techniques and procedures, which can greatly improve the practice of underwater archaeological exploration. The work described here goes in this direction, with the aim of defining new, efficient methodologies for applying ROV’s to that field. In exploring submerged archaeological sites, the ROV can be used to carry the sensors of an automatic system which, collecting and processing data, produces on-line - that is during the acquisition phase - enriched maps of the sites. The construction of such maps represents an efficient way of structuring the data, namely the images collected by the ROV, by exploiting inherent space and time correlations, which generates valuable information. In addition, the availability of the maps during the acquisition phase allows one to guide and control the work in a logic feedback fashion, with the effect of increasing performances and accuracy. Further processing, in a second time, generates 3D representations of the site in virtual reality, that may be used for deepening the archaeological study, for documentation and monitoring of the actual condition of the site and for dissemination of the information to a large audience. In the sequel, we summarize part of the work done in the framework of the EU research project VENUS [1, 2] for defining and realizing a novel procedure that integrates hardware and software components in a practical, operational scheme for the semi-automatic exploration and survey of underwater archaeological sites. Besides avoiding the use of divers, novelty is given by the availability on-line of the survey’s outcomes and by the consequent possibility to implement a logic feedback strategy for governing the whole activity. The operational scheme realized for employing the ROV will be described in Section 2, together with the ROV’s sensory apparatus, the procedure for data acquisition and the basic lines of the data processing. The processing exploits both filtering and data fusion techniques, as well as photogrammetric techniques, that produce photographic, enriched maps by mosaicing pictures. Only the conceptual aspects of the work will be described, while, for technical details, reference is made to [1, 2,]. To conclude, Section 3 mentions the experimental activity and it describes the lines of future work. 2.

Development of an operational scheme

One of the main activities in the survey of underwater archaeological sites consists in mapping the site on the basis of manual sketches and of photos. A

189

large amount of photos and videos can be collected in a semi-automatic way by using Remotely Operated Vehicles (ROV), which are guided from the surface and can carry photo/video-cameras and lighting apparatus. ROVs can also bring sonar close to the site, taking high resolution acoustic images of it. Generation of a map by patching together photos and its enrichment by the 3D information contained in the acoustic images is, to some extent, possible if the camera’s position and orientation are known for every shot and if acoustic and optical images can, in some way, be correlated. This can be viewed as a data association problem, in which auxiliary data need to be correctly associated to the principal ones, namely to photos. Photos are then patched together by specific mosaicing and photogrammetric techniques that, in particular, exploit the information generated by data association. The solution of the data association problem mentioned above, however, presents a number of difficulties, due to the nature of the involved data and to the characteristic of the acquisition methods. In order to enter into the details, it is convenient to refer to a specific experimental equipment, as the one described in the next section.

2.1.

ROV and sensory apparatus

The ROV that has been used in developing and validating the work described in this paper is a small work-class DOE Phantom S2 [4]. The sensory apparatus of the vehicle used for navigation consists of a monocular CCD PAL camcorder, an Inertial Measurent Unit (IMU), a compass and a depth meter. Additional sensors have been installed on board in order to perform the archaeological survey. These are a high definition photo-camera (Nikon D300, 14mm Sigma™ lens, 2 flashguns Nikon™ SB800), a DV video-camera (Sony HDR-HC7E) and an imaging sonar (675KHz, fan beam Kongsberg Simrad MS1000). The system comprises also an Ultra Short Base Line (USBL) acoustic positioning system, consisting of a measuring unit on the ROV’s supply vessel and a transponder on the ROV (35–55KHz Sonardyne Scout). Although different configurations are possible, the one chosen shows a good balance between efficacy, versatility and cost. The ROV’s Navigation, Guidance and Control (NGC) system is implemented on a PXI/FPGA/PC station. Its structure allows manual guidance through a console or automatic guidance, by means of a virtual reconfigurable Man/Machine interface, using data coming from the navigation sensors [4]. At low level, an onboard real-time microcontroller (Freescale 68K/Coldfire RISC MicroController) takes care of interfacing the additional optical and acoustic

190

sensors (photo/video-camera and sonar) with the NGC system. The NGC system processes all the data coming from the sensors and from the USBL positioning system and it assists the operator in guiding the ROV, by implementing autodepth and auto-heading procedures that keep constant the depth and the heading of the vehicle during the survey. The sensors have very heterogeneous characteristics and they provide data with different modalities and formats, as described in the following Table. Table 1. Characteristics of sensors and data. Sensor

Characteristics of data

Acquisition frequency

Photo-camera

JPG image

0,3Hz

Video-camera

Video Stream - DV - Full HD

25fps

3axial accelerations and angular velocity, pitch

250Hz

IMU

and roll angles (attitude = roll, pitch, yaw angle) Compass

Yaw angle (Magnetic North reference)

20Hz

Depth meter

Depth

200Hz

Sonar

Acoustic return

10Hz

USBL

Geographic coordinates x,y,z (DGPS reference)

1Hz

2.2. Data acquisition Optical and acoustic images acquisition takes place while the ROV performs a sequence of parallel, linear transects above the area of interest. During each transect, the ROV’s speed and average distance from the seabed are chosen according to the shooting frequency and to the characteristics of the photo/video-cameras, in such a way that subsequent frames overlap, assuring a complete coverage of the surveyed area. The photo-camera can be operated in an automatic way by setting the shooting frequency or manually. Since natural light is scarce, the use of flash is mandatory and, due to the recharging time of the flashguns, this limits the frequency of acquisition of the photos (about 1 photo every 3s with the experimental equipment we consider). While the video stream is recorded, low quality images are obtained by sampling it at 10Hz. Heading, depth and speed are automatically kept constant by the NGC system of the ROV. Acoustic images of the sea bottom are taken automatically at the frequency indicated in Table 1, as well as navigation (accelerations, angular velocities, pitch, roll and yaw angles, depth) and position data [8, 9]. The USBL positioning system works by evaluating the coordinates of the measuring unit by Differential GPS. The transponder located on the ROV allows then to evaluate the relative position of the vehicle with respect to the measuring

191

unit and, from this, the position of the vehicle in the DGPS reference is computed. The use of acoustic measurements in computing the position induces a delay of about 1s in the acquisition. The effect of measurement errors is attenuated by the use of Kalman filtering in the USBL positioning system.

2.3. Data association and processing Data association aims at enriching the information contained in each photo and, therefore, it uses the sequence of high definition images and that of sampled low quality images as main data. Basically, data association is made using the timestamp associated to every single sensory datum, coping with the differences in the acquisition frequencies. At a first level, JPEG/EXIF data are created by associating to each (low and high quality) image the last datum available from each navigation sensor, from the sonar and, taking into account the inherent delay, from the positioning system. In principle, this allows to get with reasonable accuracy and almost in real-time information about position and attitude of the photo/video-camera at every shot. The associated acoustic image, in addition, provides information about the distance between the camera and the pictured area. Actually, in this way raw measurement errors are not filtered and they may disturb any subsequent processing or use of the JPEG/EXIF data. Since navigation data are acquired at a relatively high frequency and since the delay in acquiring the position forces, in any case, to delay also the data association, navigation data (attitude, depth) can be filtered over a suitable time interval (up to 3s, for association to high quality images, or up to 1s, for association to low quality images). The position evaluated by the positioning system is not a raw data, since, as mentioned above, Kalman filtering is used in its computation, and hence the last available datum can be directly associated to each image. Similar considerations apply to acoustic images, which are obtained by filtering the returns of single pings. On-line processing, at this point, consists in mosaicing the acquired photo. This operation is performed using specific image processing techniques based on SIFT [3], which look for corresponding features in groups of images. Information about camera position, attitude and distance from the seabed contained in the EXIF area is used to guide the process of searching for corresponding features (in particular by selecting overlapping images and by orienting and scaling images) and this helps in increasing performances by reducing uncertainty and speeding up the operation.

192

2.4. Operational scheme in archaeological survey The result of the on-line processing described above is a 2D map of the site that is made directly available to the scientific investigators on the ROV supply vessel. The survey and exploration of an archaeological site can therefore be organized according to the operative scheme of Figure 1, where the logic sequence of phases which characterizes the work is illustrated. Starting from preliminary information on the site, collected by various means from the surface or underwater, archaeologists determine the areas and the waypoints of the survey. The mission is designed in details by engineers and, during the acquisition, information is displayed, in form of 2D map, to archaeologists. On the basis of the outcome, they can therefore modify the waypoints and the objective of the missions during its execution in a logic feedback fashion, so to guarantee area’s coverage in spite of occasional malfunctioning of the photo/video apparatus, to correct possible errors in the previous planning and to focus on interesting features. It is clear that this way to operate improves efficiency and efficacy with respect to the traditional one, in which the survey’s outcome is available only at the end of the acquisition phase. The definition of this operational scheme and its validation represent a substantial achievement in the development of new techniques for the exploration of archaeological underwater sites and they are among the basic contribution of VENUS [12, 13]. The JPEG/EXIFF data can further be processed using photogrammetric techniques and fused with acoustic images of the sea bottom in order to generate 3D maps of the site [10, 14]. In this off-line phase, the construction of the 2D mosaic is revisited and, with the aid of sonar data, images are enriched with relief. In some cases, geo-referenced bathymetric maps of the area, constructed by means of side-scan sonar from the surface, are available. Resolution of such maps can be increased by means of the sonar data collected, at close distance, by the ROV and optical images can be represented on them. In addition, tags containing notes or links to data bases of archaeological interest can be added in order to augment the content of information [11]. 3D augmented maps can be imported into virtual reality environments, which reproduce the submarine world, and be navigated in order to simulate a tour of the site. The possibility to explore the site in virtual reality represents potentially a powerful tool for making easier its study and preservation and for making it known and accessible to a larger audience. 3.

Conclusions

The procedures described above have been developed, tested and validated by means of the experimental activity performed so far in the framework of the EU research project VENUS in several missions. Sites located at different

193

depths have been explored in the Tyrrhenian Sea (near the Island of Pianosa, Tuscany Archipelagos, Italy), in the Adriatic Sea (Tremiti Islands, Italy) and in he Atlantic Ocean (mouth of Sado River, Portugal). In those missions, the ROV has acquired images, respectively, of a large, scattered group of Roman amphorae and fragments on an almost flat, sandy sea bottom and of heaps of tiles, representing the cargo of sunken vessels, on a rough, rocky sea bottom, at depths of about 30m and 60m . Data association and processing have been refined and formalized using the data collected in the missions and the experience acquired in the field operations. Sample of the maps constructed on line and by post processing are visible on the VENUS web site [1]. The operational scheme described in Section 2.4 has been implemented, tested and validated in several sea trials. In conclusion, the results obtained in the experimental activity validate the method and they represent a fundamental step in the development of semiautomatic procedures for exploiting efficiently ROV’s and marine robotic technology in the exploration of submerged archaeological sites. Progress in this direction will provide new means for archaeological studies and will eventually contribute to increase the knowledge and the preservation of important aspects of cultural heritage. In the framework of the EU research project VENUS, the work will continue to increase the performances and the level of automation in data acquisition and processing and to realize efficient procedures for reconstructing archaeological submerged sites in virtually real environments.

References 1. http://www.venus-project.eu 2. http://piccard.esil.univmed.fr/venus/deliverable.html, Public deliverables 3. D.G. Lowe, Object recognition from local scale-invariant features, Proc. ICCV 1999, Corfu, Greece (1999). 4. S. M. Zanoli and D. Scaradozzi, Automatic Control of a Low-Cost Commercial ROV, Proc. UUST 2003, Durham, NH (2003). 5. G. Conte, A. Caiti, G. Casalino and S. M. Zanoli, Underwater archaeology: available techniques and open problems in fully automated search and inspection, Proc. Workshop on Innovative Technologies for Underwater Archaeology, Prato, Italy, (2004). 6. P. Chapman et al., VENUS, Virtual ExploratioN of Underwater Site, Proc. XX CIPA /VAST 2006, Nicosia, Cyprus (2006) 7. G. Conte, A. Caiti, G. Casalino and S. M. Zanoli, Innovative technologies in underwater archaeology: field experience, open problems, research lines, Chem and Ecol., 22 (2006).

194

8. G. Conte, S. M. Zanoli, D. Scaradozzi and L. Gambella, Underwater Archaeological Data Collection by means of ROVs, Proc. IFAC CAMS 2006, Ancona, Italy (2006). 9. G. Conte, S.M. Zanoli, D. Scaradozzi, L. Gambella and A. Caiti, Data Gathering in Underwater Archaeology by means of a Remotely Operated Vehicle, Proc. XXI CIPA, Athens, Greece, (2007). 10. P. Drap et al., Photogrammetry for virtual exploration of underwater archeological sites, Proc. XXI CIPA, Athens, Greece, (2007). 11. R. Jeansoulin and O. Papini, Underwater archaeological knowledge analysis and representation in the VENUS project: a preliminary draft, Proc. XXI CIPA, Athens, Greece, (2007). 12. S. M. Zanoli, G. Conte, D. Scaradozzi, L. Gambella and A. Caiti, Proc. UUST 2007, Durham, NH (2007). 13. G. Conte, S. M. Zanoli, D. Scaradozzi and L. Gambella and V. Calabrò, Underwater Archeology Missions Design for Data Gathering Automation, Proc. MED'08, Ajaccio , France (2008). 14. P. Drap et al., Underwater cartography for archaeology in the VENUS project, to appear on Geomatica.

(B)

(D)

(C)

(A)

(E) (L)

(I)

(G)

(F)

(H)

(A) Surface vessels/UUV/Divers survey ; (B) Preliminary map of the site; (C) Archaeologists ; (D) Definition of survey’s goals and waypoints ;(E) Engineers; (F) Mission preparation and design;(G) Photos/videos and navigation data acquisition by ROV; (H) JPEG/EXIF data generation: Photo [JPEG area] and Navigation Data (position, attitude, depth, distance from bottom, sonar return)[EXIF area]; (I) Data processing tools and supervisors; (L) 2D mosaicmaps (3D maps from post- processing) Figure 1. Operational scheme in survey and exploration missions on underwater archaeological sites

195

ACCELERATORS AND RADIATION FOR ART AND ARCHAEOLOGY CLAUDIO TUNIZ The Abdus Salam International Centre for Theoretical Physics, Trieste, Italy Accelerators, high energy particles and radiation provide advanced scientific tools and procedures, mainly developed in physics research, which can be used for the nondestructive characterisation of cultural heritage materials.

1. Introduction New microscopes based on synchrotron radiation, neutrons, ion beams, lasers and other radiations, or particles can reveal non-destructively the structure and composition of art objects and archaeological remains. The analyses are applied to a variety of ‘hard’ materials, such as artefacts in metal, ceramics, stone or fossil human teeth and ‘soft’ materials, such as textiles, wood and paper. Each kind of material requires a different analytical strategy and the use of a suitable probe. The morphological, elemental and isotopic composition inferred from the analyses of these materials is important to art history, archaeology, anthropology and other areas of research relevant to cultural heritage. This in-depth characterisation can be used to develop appropriate strategies for the conservation of cultural heritage sites and objects. 2. Characterisation of cultural heritage materials Cultural heritage materials need to be characterised in the four dimensions of time and space, spanning scales of many orders of magnitude. Chronologies from decades to million years can be measured by ‘clocks’ based on radioactivity and other phenomena characterised by predictable changes with time. Satellite imaging and laser scans provide tools of increasing sophistication for prospecting cultural heritage sites on space dimensions from kilometres down to centimetres. Composition and structure of cultural heritage materials is analyzed down to the nanometre scale using new microscopes based on synchrotron radiations and high energy ions. The analysis of isotopic ratios for elements such as carbon, oxygen, nitrogen, calcium, strontium and other elements provides useful information to reconstruct migration patterns and diet of ancient human populations.

196

Genetic science offers novel approaches to cultural heritage studies. For example, the analysis of ‘ancient DNA’ has been applied to investigations of Egyptian mummies and Neanderthal remains. In the following we will discuss more details on advanced instruments and methods used in studies of cultural heritage.

3. Dating An important objective in cultural heritage studies is to order chronologically past events by analysing materials associated with past human activities. Relative chronologies can be deduced from circumstantial evidence, such as change of style and manufacturing technique. Relative chronological information can also be obtained using methods based on time-dependent geological and chemical changes (e.g., stratigraphy, sedimentation rate, weathering, hydration and magnetism). Certain kinds of cyclic phenomena, such as tree ring or varve formation, will yield very precise chronologies if stringent precautions are followed. Finally, many methods providing absolute chronologies are based on time-dependent phenomena related to natural radioactivity, and include: 1. 2. 3. 4.

decay of long-lived radionuclides produced by cosmic rays, as in the radiocarbon method; in-situ production by cosmic rays of long-lived radionuclides, such as 10Be, 26 Al and 36Cl, which can be used for dating rock surfaces and stone artefacts; build-up of radiation exposure effects, in thermoluminescence (TL), optically stimulated luminescence (OSL), electron spin resonance (ESR) and fission track dating; build-up of a radiogenic daughter from a primordial radionuclide, in K-Ar, Ar-Ar and U-series dating. 14

C is the most widely used of these chronometers. In the late 1940s, the development of radiocarbon dating by detection of the 14C residual activity revolutionised archaeology providing a precise and direct measurement of the time scale for the development of human activities during the late Quaternary. In particular, radiocarbon dating had a strong impact on the understanding of European prehistory, previously dated only by correlation with the historical chronology of the Near East. In the late 1970s , the development of direct atom counting by AMS enhanced more than a million-fold the sensitivity of 14C analysis. Extensive AMS work followed, particularly in the analysis of radiocarbon and other cosmogenic radionuclides for archaeological, geological and environmental applications [18, 19]. Through the non-invasive analysis of famous artefacts and

197

findings such as the Shroud of Turin [3] and the Ice Man [11], AMS has gained widespread public recognition as a dating technique. Also 10 Be has been used for dating in archaeology [1].

Figure 1. Comparison of the datable time span of different dating techniques [21] .

4. Accelerator microanalysis Particle accelerators were developed more than seventy years ago for basic research. There has been a major shift in the past twenty years towards their use in the analysis of materials composition and structure for interdisciplinary applications, including cultural heritage. Low-energy ion accelerators, originally constructed for nuclear physics, wereturned to other uses as the effort in their initial application faded. They have evolved into specialised tools for ion beam analysis (IBA) and accelerator mass spectrometry (AMS) [18]. An IBA facility totally devoted to cultural heritage studies has been operational for nearly 20 years at the Louvre museum [9]. A laboratory has been established in Florence with the main purpose of performing applications of nuclear techniques to solve problems related to cultural heritage (http://labec.fi.infn.it). Synchrotron accelerators have become dedicated facilities, optimised for emission of bright

198

electromagnetic radiation, an ideal microanalytical probe. A large fraction of beamtime on one of the beamlines at the European Syncrotron Radiation Facility in Grenoble is dedicated to palaeoanthropology. Finally, high-energy proton accelerators are used in spallation sources for producing pulsed beams of neutron to characterise the structure of materials. The neutron tomography at the neutron spallation source of the Paul Scherrer Institute in Switzerland has been used in a several cultural heritage applications. 4.1. Ion beam analysis Please preserve the style of the headings, text font and line spacing in order to provide a uniform style for the proceedings volume. Ion beams lose energy by ionisation of the atoms composing the target material caused by the interaction of the Coulomb field of the projectile with the atomic electrons and also by nuclear scattering from the nuclei of the atoms. The range of ion beams - in materials is short, with relatively well defined end point. By comparison, x rays are attenuated according to an exponential law and sample a much greater amount of material. Ion beams are used for trace element determination using the characteristic x-rays produced in the ionization process. Ion beams can also interact directly with atomic nuclei. Nuclear reactions, including elastic and inelastic scattering or Coulomb excitation, are useful to identify specific elements and nuclides in the sample. Concentration of individual elements or isotopes as a function of depth is possible using narrow nuclear resonances and energy loss of ions as they travel in the material. Detection methods for x-rays, !-rays, charged particles and neutrons have been developed in parallel with the development of accelerators, ion sources and other instruments necessary for the production of ion beams. Combination of different ion beam analysis techniques such as PIXE (particle induced x-ray emission) and NRA (nuclear reaction analysis) can be used to determine elemental composition for elements from hydrogen to transuranic elements. PIXE is by far the most widely applied of all ion-beam related techniques used in analysis of cultural heritage materials. It is used for routine detection ofelements with atomic numbers greater than perhaps 13, using simple energy dispersive x-ray detectors. The detection limits are not constant across the periodic table, but are extremely good in many critical regions such as for the transition elements and for heavy elements such as lead and mercury. It can be used in different modes: broad beam for analysis of bulk samples and microbeam for measurement of individual features. Maps of the composition of heterogeneous samples can be obtained by rastering the beam across the sample and making a point-by-point determination of the element present. NRA is used to make sensitive determinations of many specific isotopes. In general, nuclear reactions and elastic scattering are used for detecting specific elements/isotopes throughout the periodic table. However, nuclear reaction

199

analysis is particularly helpful for elements with Z