The ophiolite of northern Oman

The Ophiolite of Northern Oman

Frontispiece. The northern Oman mountains as imaged by the Multispectral Scanner (MSS) carried by the Landsat satellite, using near infrared light (band 7). Mosaic made from parts of two MSS scenes: path 170 row 44 and path 171 row 43. The ophiolite forms the dark terrain occupying most of ~he mountain-belt. The darkest, most rugged unit is the mantle sequence; crustal rocks are slightly paler. The palest units are sediments with autochthonous Mesozoic limestones to the west, and the Batinah coastal plain sediments to the east, of the ophiolite. For reference to ophiolite block names and geographic localities refer to the left-hand inset map on the 1:250,000 geological map that accompanies this memoir.

The Ophiolite of Northern Oman S.J. L I P P A R D , A.W. S H E L T O N & I.G. GASS Department of Earth Sciences, The Open University, Milton Keynes, MK7 6AA, UK

Memoir No 11

1986 Published for The Geological Society by Blackwell Scientific Publications Oxford London Edinburgh Boston PaloAlto Melbourne

Published for The Geological Society by Blackwell Scientific Publications Editorial offices: Osney Mead, Oxford, OX2 0EL 8 John Street, London, WC1N 2ES 23 Ainslie Place, Edinburgh, EH3 6AJ 52 Beacon Street, Boston, Massachusetts 02108~ USA 667 Lytton Avenue, Palo Alto, California 94301, USA 107 Barry Street, Carlton, Victoria 3053, Australia 9 1986 The Geological Society. Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by The Geological Society for libraries and other users registered with the Copyright Clearance Center (CCC) Transactional Reporting Service, provided that a base fee of $02.00 per copy is paid directly to CCC, 21 Congress Street, Salem, MA 01970 USA, 0305-8719/84 $02.00 First published 1986 Printed and bound in Great Britain by William Clowes Limited, Beccles and London

DISTRIBUTORS

USA and Canada Blackwell Scientific Publications Inc PO Box 50009, Palo Alto, California 94303 Australia Blackwell Scientific Publications (Australia) Pty Ltd 107 Barry Street, Carlton, Victoria 3053

British Library Cataloguing in Publication Data Lippard, S.J. The ophiolite of Northern Oman. 1. Rocks, Igneous--Oman I. Title II. Shelton, A.W. Ill. Gass, I.G. Society of London 552'. 1'095353 QE461

IV. Geological

ISBN 0-632-01587-X Library of Congress Cataloging-in-Publication Data Lippard, S.J. The ophiolite of northern Oman. (Memoirs/Geological Society: no. 11) Bibliography: p. Includes index. 1. Geoiogy--Oman. 2. Ophiolites--Oman. 1. Shelton, A.W. II. Gass, I.G. (lan Graham) III. Title. IV. Series: Memoirs (Geological Society of London): no. 11. QE291.05L56 1 9 8 6 555.3'53 86-11737 ISBN (,L632-01587-X

This memoir is dedicated to the memory of the late D R I S M A E L EL BOUSHI

former geological adviser to the Sultan of Oman, without whose help and whole-hearted co-operation this research project would not have been possible.

Contents

Preface

ix

The Geological Background 1.1 Introduction 1 1.2 Regional Tectonic Setting 2 1.2.1 The Zagros 2 1.2.2 The Makran 3 1.2.3 The Gulf of Oman 4 1.3 Tethyan Ophiolites 4 1.4 History of Geological Investigations 7 1.5 Tectonostratigraphy of the Oman Mountains 9 1.5.1 The Basement Rocks of the Arabian platform 1.5.2 The Allochthonous Units 12 1.5.3 Late Cretaceous Nappe Emplacement 14 1.5.4 Neoautochthonous Sediments and Tertiary Structures 16

11

Evolution of the Oman Tethys

17 2.1 Introduction 17 2.2 Rifting and Continental Break-up 18 2.2.1 Permian Sediments 18 2.2.2 Triassic Sediments 20 2.2.3 The Haybi Volcanics; Triassic Volcanism at the Opening of the Oman Tethys 22 2.3 The Oman Passive Margin 31 2.3.1 Shelf and Slope Facies Sediments 31 2.3.2 Rise and Basin Facies Sediments; the Middle and Upper Hawasina 32 2.3.3 Late Mesozoic Alkaline Igneous Activity 34 2.4 Mid-late Cretaceous Syn-tectonic Events on the Continental Margin - a prelude to late Cretaceous Nappe Emplacement 36 2.4.1 The Aruma Group 36 2.4.2 The Hawasina Melange 36 2.5 The Oman Tethys - a Summary 37

The Semail Ophiolite

39

3.1 Introduction 39 3.1.1 Internal Subdivisions 40 3.1.2 External Relations of the Semail Nappe 41 3.2 The Mantle Sequence 43 3.2.1 Harzburgites 46 3.2.1.1 The Basal Lherzolite 48 3.2.1.2 Geothermometry and Geobarometry 51 3.2.2 Dunites and Chromitites 52 3.2.2.1 Chromite Mineralization 55 3.2.3 Mantle Dykes 55 3.2.4 Structure 59 3.2.5 Alteration 62 3.2.5.1 Alteration Products 62 3.2.5.2 Physical and Chemical Effects of Serpentinization 63 3.2.6 Summary 64 3.3 The Petrological Moho 65 3.4 The Layered Series 66 3.4.1 The Layered Series of the Semail Ophiolite 68 3.4.1.1 Structure 68 3.4.1.2 Petrology and Mineralogy 72 3.4.1.3 Geochemistry 76

3.4.2 Large Scale Cyclic Units, Crystallization Sequences and Cryptic Variations 76 3.4.3 Magma Chamber Processes 78 3.4.3.1 Origin of the Cyclic Units 78 3.4.3.2 Magma Chamber Size and Shape 78 3.4.4 Fractionation Trends and Primary Magma Compositions 79 3.5 The High-Level Intrusives 79 3.5.1 Field Relations 79 3.5.2 Petrology 81 3.5.3 Geochemistry 82 3.5.4 Petrogenesis 86 3.6 Late Intrusive Complexes 86 3.6.1 Large Gabbro-diorite-plagiogranite bodies 87 3.6.2 Peridotite-gabbro Complexes 90 3.6.3 Geochemistry and Petrogenesis 91 3.7 The Sheeted Dyke Complex 94 3.7.1 Field Relations and Structure 95 3.7.2 Dyke Trends 100 3.7.3 Petrography and Mineralogy 100 3.7.4 Geochemistry 101 3.8 The Extrusive Sequence 103 3.8.1 Volcanic Stratigraphy 103 3.8.2 Geochemistry 107 3.8.3 Tectonic Setting 113 3.8.4 Clinopyroxene Compositions 115 3.8.5 Rare-earth Elements 116 3.9 Petrogenesis 117 3.9.1 Convective Processes 120 3.9.2 Partial Melting Processes 120 3.9.3 Magma Fractionation 123 3.10 Ocean-floor Metamorphism 124 3.10.1 Metamorphic Facies: Nomenclature and Conditions 124 3.10.2 Ocean-floor Metamorphism in the Semail Ophiolite 124 3.10.3 Summary 127 3.10.4 Massive Sulphide Deposits 127 3.11 Metalliferous and Pelagic Sediments 128 3.11.1 Fauna and Age 128 3.11.2 Stratigraphy and Field Relations 128 3.11.3 Geochemistry of the Pelagic Sediments 132 3.12 Isotopic and Magnetic Studies 133 3.12.1 Isotopic Studies 133 3.12.2 Magnetic Studies 135

Ophiolite Detachment, Emplacement and Subsequent Deformation 140 4.1 Introduction 140 4.2 Detachment 140 4.2.1 The Metamorphic Sheet 140 4.2.2 Peridotite Banded Unit 148 4.3 Emplacement and Related Processes 149 4.3.1 Metamorphism 149 4.3.2 Igneous Activity Associated with Emplacement (Biotite Granites) 151 4.3.3 Structural Evolution 153

References

167

Preface

This memoir is based on studies by Open University (OU) and associated personnel between 1975-85 and represents a fullsome precis, correlation and evaluation of work presented in eleven Ph.D. theses and numerous scientific publications. In total, it represents some 48 man-years of effort. The names of persons associated with the research project at various times, together with their status and affiliation, are listed in the table below. Early in the reign of Sultan Quaboos the Oman Government began to welcome overseas scientists. Until then, only geologists of the PDO (Petroleum Development Oman) a part-Oman Government part-Shell International Company, had been allowed to work in the north of the Sultanate. Work by PDO scientists, particularly the classic study of the Oman mountains by K.W. Glennie and his coileagues (Glennie et al. 1974), although primarily concerned with sedimentary formations, provided a firm foundation for this ophiolite-orientated project. Our own association with the Oman began in 1975 when one of us (IGG), on a visit to King Abdulaziz University (KAU) in Jeddah, Saudi Arabia, discussed with Dr A.O. Nasseef, PresiOpen University

dent of that University and now Secretary General of the Muslim World League, the scientific desirability of investigating the Oman Ophiolite and its potential as a post-graduate training ground. Dr Nasseef contacted Dr El Boushi, then Geological Advisor to the Sultan of Oman, and as a result Gass & Dr Abdurrazzak Bakor (KAU), together with Drs J.D. Smewing and A.D. Lewis (OU Research Fellows) visited the Sultanate later that year. It took only a few days to realize that the Oman mountains are a magnificent ophiolite, similar in many respects to the then better known Troodos massif of Cyprus. All present were convinced that it was a research area of the highest potential. It was envisaged that the area would be used as a training ground for Saudi, Omani and British graduate students who would work under the supervision of Dr I.M. El Boushi, and academic staff from the Open and King Abdulaziz Universities. Geological maps on the scale of 1:100,000 would be produced under an informal agreement with no contractual obligations on any of the parties. A grant of s from the Royal Society of London paved the way for a detailed feasibility study of the ophiolite in 1976 81

82

83

84

85

86

ACADEMIC STAFF

I.G. Gass J.A. Pearce A.J. Fleet, R.M. Shackleton RESEARCH FELLOWS

J.D. Smewing (OU) K.O. Simonian (OU) A.D. Lewis (OU) S.J. Lippard (OU) C.R. Neary (BGS: ODA; chromites) D.I.J. Mallick (BGS; ODA; remote sensing) R.B. Evans (BGS, ODA; geophysics) RESEARCHSTUDENTS:(abbreviated thesis titles and date D.T. Aidiss (NERC) M.A. Brown (ODA) G.M. Graham (Shell) M.P. Searle (OU) T. Alabaster (ODP) A.W. Shelton (ODA) D.A. Rothery (ODA) P. Browning (NERC) G. Stranger (Private) S. Roberts (EEC) I.D. Bartholomew (NERC)

(field leader 1976-7~

II

(field leader 197%81)

)resentanon given m parenthesis) ] (Granitic rocks of ophiolites: 1978) (Chromite studies: 1983) (AIIochthonous sediments. Hawasina Windo'~v: 19811) (AIIochthonous ttavbi complex and metamorphic sole: 19811) - - (Volcanic and hvdrothermal processes: 1982) ~ (Gravity and palaeomagnetic studies: 19841 (Remote sensing: 1982) (Plutonic rocks in Rustaq area: 1982) ~ (Groundwatcr hydrogcochemistrv: 1986) (Chromite metallogenesis: 1986) - - (Mantle sequence structures: 1983)

ASSOCIATED SCIENTISTS

I.M. Elboushi (Oman Government) N.I. Christengen (Seattle) J. Malpas (Memorial) A.H.F. Robertson (Edinburgh) A.G. Smith (Cambridge) N.H. Woodcock (Cambridge) D. Elliot (John's Hopkins) P. Tippitt (Texas)

(Mineralization) I (Seismic and related studies) (Field studies) I I [ [ [(Sedimentology) (Structure) ] I ] --% [ (Sedimentology and structure) (Structure) I I ] I (Radiolarian dating) 9

Preface by J.D. Sinewing, A.D. Lewis, K.O. Simonian and D.T. Aldiss. This fully confirmed that high quality geological studies could be made in this well exposed and relatively undeformed terrain. When it became likely that no KAU personnel would be further involved, Dr El Boushi agreed that the OU group should proceed alone. Subsequent to the initial Royal Society grant, the Natural Environmental Research Council (NERC) provided, over the period 1976-81, a research grant of s as well as three research studentships. However, by far the largest support (s 1977-82) came from the British Government's Ministry of Overseas Development (now Overseas Development Administration: ODA). Through the good offices of Dr David Bleakley, then the Director of the Overseas Division of the British Institute of Geological Sciences (now BGS: British Geological Survey) and Professor E. Machens, the project was put to the OEDC (Organization for Economic Cooperation and Development) whose enthusiastic support resulted in the ODA grant. These monies funded four research students and three BGS specialists and allowed the production of four 1:100,000 coloured geological maps as well as the 1:250,000 geological and gravity maps that accompany this memoir. Another major contributor was the Open University which funded 15 man-years of Research Fellowship and one research student. Shell and the European Economic Community (EEC) each funded a research studentship whilst the American Oil Company (AMOCO) funded a Research Fellow for two years. We are most grateful to all these funding agencies for their support. Throughout the project we have received invaluable assistance from numerous groups and individuals. In Oman, we are particularly indebted to the Department of Petroleum & Minerals (DPM) whose Director, Mr Mohammed Kassim, sponsored the project and provided the necessary documentation for project personnel to work in the Sultanate. We also acknowledge with many thanks the provision, by the DPM, of a house in Sohar that acted as the project's base for many

years. We were particularly fortunate that between 1977-79, our peak period of field work, we had the company in Oman of a University of California - United States Geological Survey research team who were making a detailed study of the ophiolite along a 25 km wide N-S strip extending south from Muscat. We benefited greatly from discussions with our American colleagues as we did from the PDO geologists who gave freely of their time and expertise. But, perhaps most of all, we will remember the unfailing and uniquely generous hospitality of the Omanis whose mountains we invaded. On the more practical side, the RAF and SOAF (Sultans Own Air Force) provided invaluable assistance in air lifting two landrovers to the Oman, and bringing back to the UK several tons of rock specimens. At the OU we received the fullest support and cooperation from the Department of Earth Sciences" technical and secretarial staff. We are most grateful to John Holbrook, Ian Chaplin, Andy Tindle & John Watson who provided prompt and efficient curating, specimen preparation and analytical services, to Helen Boxall who prepared all the text figures in this memoir and to John Taylor for the 1:250,000 geological and gravity maps that accompany it and to Pare Owen and Carol Whale who provided a first class secretarial service that saw the manuscript through numerous stages. The principle objective of this memoir is to present the major findings of the project under one cover as a coherent whole. We are greatly indebted to our former colleagues for allowing us to use their data in this way and also for commenting on and correcting those parts of the manuscript based on their findings. We are particularly grateful to the following who checked and improved earlier versions of the manuscript, Julian Pearce, Robert Shackleton, John Sinewing, Chris Neary, Tony Alabaster, Dave Rothery, Paul Browning, Gordon Stanger, Steve Roberts, lain Bartholomew, Micky Brown, John Malpas, Aiastair Robertson, Alan Smith & Martin Menzies.

Chapter 1 The Geological Background of which lies between 500 and 1500 m, is bare and rocky with little or no vegetative cover. Although some wadis have perennial streams, most are dry except during severe winter storms when they are subject to flash flooding. East of the central part of the mountains there is a narrow coastal plain (the Batinah), underlain by recent sands and gravels and coastal deposits; whereas inland to the west and south are the desert wastes of the interior of Arabia, known as the Empty Quarter or Rub al Khali. Most of the mountain area is in the Sultanate of Oman, but between Wadis Hatta and Dibba an eastward extension of the United Arab Emirates (U.A.E.) separates the Musandam

1.1 I n t r o d u c t i o n The Oman Mountains, also known as the Hajar or Eastern Hajar (Hajar al Gharbi) range, lie at the eastern extremity of the Arabian sub-continent and run in a broad arc parallel to the Gulf of Oman coastline (Fig. 1.1). They extend from Ras Musandam, facing the Straits of Hormuz in the north, to Ras al Hadd in the east, a distance of nearly 800 km. The mountains have an average width of about 75 km, reaching a maximum of 130 km in the central part, and rise to a height of about 3000 m on Jebel Akhdar. The rugged mountain terrain, most 1

56 ~E

.~

N

o~

I

I,-~4~~4~/~~ MTt~s

~

[

~.o,~ Qo; ~ ", MAKRAN

\

,,,

RLJus al Jibal (2087 m)

26 ~N I.

/

~-"--~--J/

~)_GULF

ARABIA

rv"

,>

p,,,.~., "

//

W/k -n---, i

jl -j

::s

/'-' ~i Khor

1500-2000 m

~.O

I

Iloo0-15o0m

1

J 500-1000 m

[

] Land b e l o w 500 m Main c o m m u n i c a t i o n routes t h r o u g h the m o u n t a i n s International b o u n d a r y Major t o w n s Main m o u n t a i n peaks

.....

ohar _X

'~~'q!~,

INDIAN OCEAN

Land over 2000 m

\

/Buraimi",,

.sz

"4

Fakkan ~ ~13 Fujairah ~

z~

/

F~

,f

g

~ .f ! / ,4,1

OF OMAN

.

.~,

--,

I

",...1

0~4%~4L

9

o

4~

r4 uvs

@ ,.,\\ t e /

Seeb P

I a

i

n

,,

i '--Jebel Ka~ ~,,(2960 , ~

K F-~IQ-A - ~ / ~ ' ~ 1 N a p p e ~

Crop

BATINAH COMPLEX

~ ~: "Batinah Melange" (KU)

SIMSIMA Fm

Maa

.

Fm (LJ-KM) Sakhin Fm (UT~-L J)

,~i ~.

"Sheeted Dyke Complex"

9 ',' " "High Level Intrusives"

~ - ~ _ . ."~~

wa.s,~ G ~ : : ~

"Cumulate Sequence"

"Late Intrusive Complexes" occur w~lhin these units

~. *Crustal Sequence"

SEMAIL NAPPE

=-~:7~--:=--" Pet r olo gic a I Moho" Kahmah Gp ~

"Mantle

[-~-r Unit ,4 -~ : z

Sequence"

~: "Banded Unit" -'". '-- "Metamorphic Sheet" (KM-KU)

~.e

Ghalilah ~alilah Fm :~ ~:

uo

M ~

L

~

~-Fm

Huqf Gpl Kharus Frn ~ Late

Mi'aidin Fm Qahlah

~.

V. .,~ ~ " . - \ \ ~ ..~,,,~)

Sur

Rasal Hadd

I / I

Fig. 1.9. Tertiary outcrops and structures.

"

/

!

x6

Chapter 1

tectonically higher ones. The last movement on the Semail Thrust appears to have been relatively "late" so that it cuts across imbricate structures in the underlying Haybi and Hawasina complexes (Searle 1980). Emplacement of the nappes was preceded by a rapid collapse of the continental margin which allowed it to accommodate over 10 km thickness of allochthonous rocks. As emplacement proceeded, the Aruma basin foredeep developed and migrated in front of the advancing nappe pile locally accumulating up to 3-4 km of deep water clastic sediments. Rapid uplift (isostatic rebound?) of the area occurred during the last stage of emplacement. 1.5.4 Neoautochthonous sediments and tertiary structures

Maastrichtian, Palaeocene and Eocene sediments, mainly limestones with a total thickness up to 600 m, unconformably overlie the late Cretaceous allochthonous units of the Oman Mountains (Lees 1928; Morton 1959; Glennie et al. 1974) along the eastern and western flanks of the mountains (Fig. 1.9). There are two facies: fluviatile conglomerates, sands and gypsiferous marls (Qahlah Formation) and fossiliferous marine limestones (Simsima Formation). The coarse grained clastics are largely restricted to the eastern side of the mountains; whereas, by contrast, on the western side, the limestones rest directly on older rocks. In places subaerial exposure and lateritic weathering of the ophiolite occurred before Maastrichtian sedimentation took place (Coleman 1981). The Maastrichtian-Lower Tertiary sediments were deformed during the mid-Tertiary, post-Middle Eocene. The tectonism was particularly severe in the north where, along the west side of the Musandam Peninsula, the Mesozoic limestones of the Ruus al Jibal massif were displaced several

kilometres westward along the Hagab thrust (Hudson et al. 1954b; Allemann & Peters 1972). Along the line of the thrust the limestones are locally overturned and thrust over Eocene shales (Searle et al. 1983). In general, the Tertiary fold axes trend parallel to the mountains from N-S in the north, to NW-SE and E-W further south although there are local variations and changes in strike. It has been suggested that the Tertiary folding is the result of reactivation of basement faults and local uplift, perhaps partly due to salt doming (Tschopp 1967; Glennie et al. 1974). A regional origin is however, suggested by their widespread development and the consistency of trends and directions of the fold axes (Searle et al. 1983) (Fig. 1.9). The post-Miocene history of the area is one of repeated uplift of the mountains and subsidence of the flanks with thick sediment deposition. Glennie et al. (1974) report over 4000 m of Neogene sediments in an offshore borehole on the Oman continental margin where extensive slumping of the thick semi-consolidated sediments has been identified on seismic profiles (White & Ross 1979). Cycles of repeated uplift are recorded by a series of gravel-capped terraces on the flanks of the mountains. Elevated terraces up to 300 m above sea level are found on Jebel Nakhl and on the northern edge of Saih Hatat just south of Muscat, but are as yet undated. In the major mountain wadis cemented gravels are "perched" up to 50 m above the present valley floors. Raised beaches and marine terraces characterize most of the coastal area from Dibba southwards (Lees (1928) reports "sub-recent" shells west of Shinas at 375 m above sea level), but to the north the Musandam Peninsula has a drowned, ria-type coastline and has been sinking in recent times (Vita-Finzi 1973).

Chapter 2 Evolution of the Oman Tethys 2.1 Introduction

Gondwanaland supercontinent, to produce a narrow ocean basin by the end of the Triassic, followed by the Jurassic to mid-Cretaceous development of a northeast-facing passivetype continental margin. The history of the Oman margin, as recorded by the sediments that overlie the pre-Permian basement, begins after a marine transgression in the Middle Permian that followed Palaeozoic (post-Ordovician) folding, uplift and erosion. The Triassic marks the first appearance of a deep water basin in which the earliest hemi-pelagic and pelagic Hawasina sediments were deposited. Volcanicity in the late Triassic (Haybi Volcanics) shows a trend from early alkaline through transitional to later tholeiitic types, reflecting the further development of the ocean basin. At the end of the Triassic in the present area of the Oman Mountains there was a broad continental shelf bordered to the northeast by a deep-water (Hawasina) basin and beyond this an outer margin "high" composed of shallow-water limestones (now represented by the "Oman Exotics") which formed as carbonate platforms developed on a basement of the late Triassic volcanics. In the mid-Jurassic the deposition of largely quartz-rich terrigenous turbidites in the Hawasina basin, which had begun in the mid-Triassic, was succeeded by calciturbidite-chert deposition. The rift-related volcanicity ceased at the end of the Triassic and the outer margin carbonate platforms subsided and were blanketed by pelagic sediments. In the late Jurassic-early Cretaceous there were relatively deep-water, sediment-starved conditions across the whole continental margin from the shelf to the ocean basin. By the mid-Cretaceous (Cenomanian) continued subsidence had resulted in the deposition of over 3000 m of shelf (Hajar Supergroup), 2000 m of slope (Sumeini Group) and at least 1000 m of continental rise (proximal Hawasina facies) sediments, whilst only a few tens of metres of pelagic limestones and cherts (distal Hawasina sediments) had been deposited in an abyssal oceanic environment between the late Triassic and mid-Cretaceous. Contemporaneous magmatism is represented by alkali dolerite and wehrlite sills (dated at 16093 Ma) that intrude the Hawasina sediments. In the mid-Cretaceous tectonic instability of the continental margin was reflected by uplift and erosion (the Turonian "Wasia-Aruma break") followed by subsidence and relatively deep-water sedimentation in the Coniacian to Campanian (Aruma Group). In the outer part of the continental margin, sedimentary melanges (olistostromes) formed in the midCretaceous as a response to rapid differential uplift and subsidence probably marking the beginning of subduction. These events heralded the emplacement of the continental margin sediments and volcanics as a series of thrust nappes (Hawasina Assemblage and Haybi complex). The Semail Nappe represents a slice of oceanic lithosphere that was generated at a midCretaceous spreading axis to the NE of the Oman margin and which was probably formed above a northward-dipping subduction zone in a marginal basin tectonic setting (see Chapter III). The ophiolite nappe was detached by intra-oceanic thrusting in the Turonian (ca. 90 Ma) and was then emplaced (obducted) southwestwards over the Hawasina and Haybi allochthons onto the continental margin in the Santonian to Campanian (87-72 Ma). Late in the emplacement sequence tectonic melanges formed beneath the Semail Nappe and were

The "Oman Tethys" was part of an ocean area that extended from the Mediterranean to eastern Asia between the ancient supercontinents of Gondwanaland (Africa-Arabia-India) to the south and Eurasia to the north (Fig. 2.1). The ocean appears to have formed in the early Mesozoic by the rifting off and northward migration of several small continental fragments or "microplates"; including Anatolia, Central Iran, Afghanistan and Central Tibet, away from the remainder of Gondwanaland (Dewey et al. 1973; Stoneley 1974; Boulin 1981). This seaway has been called the "Southern Tethys" or "Neo-Tethys" to distinguish it from the older "Palaeo-Tethys'" that lay to the north of the microplates and which was largely destroyed by pre-Liassic plate collision (Takin 1972; Sengor 1985; Teleki 1981). On a local scale, Glennie et al. (1974) referred to the Mesozoic ocean to the north of Oman as the Hawasina Ocean; here it is called the Oman Tethys. The evolution of the Oman Tethys throughout the late Palaeozoic and Mesozoic, up until its destruction by ophiolite obduction and nappe emplacement onto the Arabian platform in the late Cretaceous, is here discussed in terms of the rifting of the northern margin of the Arabian continent, then part of a

30~N ~ . ~ ~

Y

~

============================================ ~':,:,iiiiiiii!iii!!!!!iiiiiiiii ! !i ! : : Ap \4:q::::::::::: i i i i i ~_.

//

/ / / / / / / 7"--,._

An

~'--... >-

"~::::::::::: ........ : :--,

~_

~

~i~ ~

~ .

~ ~

~ .

.

.

.

.

.

.

.

.

~ ~

~

~,~

3C

.

.

.

.

~

~ ~

.3~

~.

.

~

~ =

.~.

~: ~

6.~

~

-.-

~

c_:=

.

tr~, ".C

~'~6 3C

~

~,c,=

._

.--

-~

~.~

0 0 0 0 ~ > ~ .

,r"4

~t-,9 3 c

~c

ur;

~.1

.,.c. ~

~

g

~

'

. ~ ~


1300 m (Glennie et al. 1974) which is nowhere exposed in the mountains but found in the subsurface to the west and southwest. The coarse grained Juweiza Formation (c. 3000 m) of late Campanian to early Maastrichtian age overlies the Fiqa beds and contains abundant Hawasina and Semail clasts (Glennie et al. op. cit.). The relationships and lithologies of these Aruma Group sediments were interpreted by Glennie et al. (1974) as indicating the deposits of a deep water ensialic basin or "foredeep" that migrated westwards diachronously in front of the advancing nappes of the Oman Mountains. 2.4.2 The Hawasina Melange

0

1

I

La Ce

I

Nd

I

I

I

I

Sm Eu Gd Tb

I

1

1

Tm Yb Lu

Fig. 2.16. Chondrite-normalized REE patterns for alkali peridotite and gabbro sills. From Lippard (1984). northern U.A.E. part of the mountains, in the Dibba zone, there are alkaline ankaramitic tufts, dated as 96 + 4 Ma (Lippard & Rex 1982), which contain blocks of alkali peridotite and syenite. These volcanics are interbedded with red radiolarian cherts and occur in a melange of probable mid-late Cretaceous age equivalent to the Aruma Group (Searle et al. 1983). It is possible that this Cenomanian alkaline igneous activity is related to tectonic instability of the Oman margin due to the onset of subduction and ocean basin closure prior to ophiolite obduction and nappe emplacement (see next section). The late Mesozoic alkaline igneous activity in the Oman Mountains is petrologically similar to the mid-late Triassic Haybi alkaline volcanics, but represents a later and quite distinct magmatic event. Incontrast to the early phase, which was related to rifting and continental break-up, the second and longer phase seems to have occurred largely during passive margin development. Alkaline rocks are presently found off some passive margins, such as the Tertiary to Recent volcanics of the Canary and Cape Verde island groups off northwest Africa which are formed of Cenozoic alkaline rocks built partly upon an older Mesozoic oceanic foundation (Robertson & Stillman 1979; Stillman et al. 1982).

2.4 Mid-late Cretaceous Syn-tectonic Events on the Continental Margin - a Prelude to Late Cretaceous Nappe Emplacement 2.4.1 The Aruma Group An important break in sedimentation occurred in the Arabian shelf sequence between the Cenomanian limestones of the Wasia Group, at the top of the autochthonous carbonate

The Hawasina Melange forms a discontinuous unit above the Hawasina Assemblage at the base of or imbricated within the Haybi complex. The boundaries are generally tectonic and thickness estimates are difficult because of tectonic disruption and thickening due to late Cretaceous thrusting and folding; however, Graham (1980a) suggests that locally it reaches 1500 m although thicknesses of 50 to 250 m are typical. The deposits range from chaotic "block-on-block" megabreccias with blocks ranging from metre- up to kilometre-size to olistostromes composed of smaller (~ I

High-level Intrusives

Gabbro

..'"

~

mainly melange

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

WADI JlZl

Layer 3b 7.1 km/s

i!!!

k I I \ I I I 7 I I Vp km/s

.

! !/!!i!:: ~

-,,

/ / Perid ,_/~, otite //

.

56~176 E . . . . . . . . . . . ....

Ampiib~

"Seismic Moho"

'*Petrological Moho" Tectonite ectonite harzburgite ~rzburgite Mantle Sequence + dunite 85 I I

..:

[.

Layered S e r i e s ~ 6-

50'N

Ze i

Layer 3a 6.7km/s 4

WADI HATTA

5:6: '0'E

-

lite \

43

1

Layer 4

8.1 km/s

9 I

Fig. 3.4. Stratigraphic column (1), P-wave velocity profile (Christensen & Sinewing 1981) (2) and metamorphic facies (3) of the Semail

ophiolite compared to the seismic structure of oceanic crust (Raitt 1963). Details of the metamorphic facies overlap can be found in Section 3.10. melange has been called the "Basal Serpentinite" by Searle (1980) and contains thrust slices of weakly metamorphosed (sub-greenschist facies), but highly deformed, Haybi volcanics, Hawasina sediments and Exotic limestones in a complex tectonic unit known as the Haybi Complex (Searle 1980, Section 4.3.3.2). The Haybi Complex not only structurally underlies the Semail Nappe but also occupies cross-strike fault zones or "corridors", such as Wadi Jizi, between some of the ophiolite blocks (Fig. 3.5). In some of the corridors, notably in Wadi Sakhin just to the south of Wadi Ahin, it is continuous with the Batinah Melange that locally overlies the Semail Nappe along its eastern margin (Lippard et al. 1983; Robertson & Woodcock 1983b). In some places; for example, at Suhaylah in Wadi Jizi, the melange is in depositional contact with underlying pelagic sediments that in turn lie on top of the ophiolite (Woodcock & Robertson 1982a), elsewhere it has a tectonic contact. The Semail Nappe and this discontinuous melange cover are in turn tectonically overlain by several thrust slices of Hawasina-type sediments, the Batinah Sediment Sheets (Woodcock & Robertson 1982b), which form the upper part of Woodcock & Robertson's "Batinah Complex". On the flank of the mountains the Semail Nappe and the other allochthonous units are unconformably overlain by Maastrichtian sediments. In the sub-surface to the west of the mountains the late Campanian Juweiza Formation (Glennie et al. 1974) contains abundant clasts of igneous rocks probably derived from the Semail Nappe suggesting that it had been emplaced into its present position on the Arabian continental edge by that time. Coleman (1981) describes evidence for lateritic weathering of the ophiolite before Maastrichtian sedimentation in the Ibra area of the SE Mountains, and the same is found along the western flanks of the northern mountains in the United Arab Emirates (U.A.E.) (P. W. Skelton pers. comm.).

l

56 ~ 3 0!' ,E: , v ,. : . ~9.;.~. .i ~. .b - . o .",~.

r

9 : 9 :::

:i: ::

/

iiiii:, ~

;

:i! /

~

:: :

'

;

::): : :

;:::: :.'..

~ ,:~' ~ T ' a , ,o~

::) ':,1 '

Y

0

o"

,

.!!iiiiiiiii::::::: .........

..

10 km I

0

Corridor

l

I

SEMAIL NAPPE ~ Extrusive Sequence ~Sheeted Dyke Complex Layered Series.and High level Intrusive Complexes

~ ~

WADI AHIN

Normal stratigraphic contact I

High-angle fault

9

Thrust fault

Mantle Sequence

Fig. 3.5. Three melange-filled fault zones in the northern part of the Oman M o u n t a i n s - Wadis Hatta, Jizi and Ahin.

3.2 The Mantle Sequence At the base of all complete and coherent ophiolite complexes there is a unit of tectonized ultramafic rocks which, because it has petrological, geochemical and geophysical characteristics similar to those envisaged for oceanic upper mantle, is often referred to as the "Mantle Sequence". This term was first used by Allen (1975) to describe the tectonized peridotites at the centre of the Troodos Massif on Cyprus. The Mantle Sequence of the Semail Ophiolite (Smewing 1980a) comprises the lower half to two thirds of the Semail Nappe and consists of variably serpentinized tectonized peridotites, dominantly harzburgites that locally grade into lherzolites and have subordinate bodies of dunite, the whole being cut by numerous, but volumetrically insignificant, dykes, veins and irregular pods of ultramafic and mafic rock types (Fig. 3.3). Precise estimates of thickness are difficult to make

44

Chapter 3

because, although the harzburgites and dunites, the two major rock types in the sequence, can be readily identified in the field, they are irregularly disposed and no major tectonic breaks or repetitions of the sequence have so far been identified by their dislocation. The best estimates that can be made are by simply extrapolating the dip of the overlying crustal units, and in particular the plane of the Petrological Moho (Section 3.3), down section across the width of the Mantle Sequence outcrop. This method assumes no major repetitions of the sequence produced by thrusting, an assumption supported by detailed studies along several well-exposed wadi sections which have failed to locate major thrusts, and also that the dip remains constant across the section (a less certain criterion in areas affected by large-scale folding). Using these assumptions, variations in the maximum thickness of the Mantle Sequence ranging from 5-7 km in some of the smaller blocks (e.g. Rustaq (Browning 1982)) to 8-10 km in the large Fizh block (Rayy-Ragmi section (Smewing 1980a)) and 9-12 km in the Ibra-Wadi Tayin area (Boudier & Coleman 1981) are indicated. Near to the base of the sequence, particularly in the Rayy-Ragmi section, the peridotites are more lherzolitic in composition; this, although not a major occurrence, is petrologically significant and is described below as the "Basal Lherzolite". Irrespective of the thickness of the Mantle Sequence, the lowermost peridotites, for as much as 500 m above the basal thrust, have been more intensely deformed than the rest of the sequence. In places these highly deformed rocks have a markedly banded appearance due to metre-thick alternations of harzburgite and dunite and have been called the "Banded Unit" by Searle (1980). The origin of the Banded Unit and its role in the obduction process are discussed in Section 4.2.2. The Mantle Sequence forms 60-70% of the outcrop area of the Semail Nappe and gives rise to a distinctive dark rugged mountainous terrain (Plate 3.1) with summit heights of 10001500 m that is deeply dissected with numerous steep gullies feeding into a dendritic to rectangular drainage pattern of steep-sided, flat-bottomed wadis. The peaks and slopes are covered in a near in situ scree of weathered blocks and rubble so that massive outcrops are largely confined to the lower slopes and floors of the larger watercourses where good waterwashed exposures occur. Where fresh, the peridotites are pale

Plate 3.1. Typical rugged Mantle Sequence (harzburgite) landscape. Distant peak is - 1200 m above sea level and about 10 km distant.

grey-green rocks but they weather to various shades of brown with an iron-rich desert varnish on exposed outer surfaces. The sequence consists of 85-95% harzburgite (locally gradational into lherzolite) and 5-15% dunite. The harzburgites are medium to coarse grained rocks composed of 75-85% olivine and 15-25% orthopyroxene. Chrome spinel (chromite) is a ubiquitous accessory forming 0.5-2% of the mode. Clinopyroxene, although generally low in abundance (98% olivine and 4 % ) than in the rest of the sequence, and, most significantly, the chrome spinels become more magnesian and aluminous marked by a colour change from dark red-black to yellow-brown (Roberts 1985). There is also a concomitant increase in the modal clinopyroxene content from < 1 to > 3 % . Major element analyses (Table 3.4) confirm a progressive change from harzburgitic to lherzolitic compositions towards the base of the sequence. In particular, the lowermost sample (OM9013), which was collected within 500 m of the base, has markedly higher Ti, AI and Ca contents than all the other analysed samples of mantle peridotites from the Semail ophiolite. Spray (1984), in discussing the decoupling of ophiolites from their in situ oceanic position, suggests that they represent

The Semail Ophiolite

estimate the temperatures and pressures of partial melting or, at least, the conditions at which the residual mantle material was last in equilibrium with the melt. But disequilibrium and lower temperature and pressure sub-solidus re-equilibration make these 'geometers' unreliable. We include a summary of the data despite having doubts as to their usefulness and accuracy. Calculated geothermometric and geobarometric data for the Semail Mantle Sequence, taken from Browning (1982), is given in Table 3.5. Individual orthopyroxene compositions, even from a single sample, produce a wide range of pressures (5-53 kb) and temperatures (900-5600~ These variations are far in excess of the likely errors (Mercier 1980), but not surprising in view of the known heterogeneity of the orthopyroxenes with respect to CaO, A1203 and Cr203 contents. Single clinopyroxene calculations give a similar wide range of values (844--1207~ 5.0-38.3 kb), although the core of a large clinopyroxene grain from the lherzolite (OM9015) gives 1207~ and 38.3kb, whereas the recrystallized groundmass grains in the same specimen give 844-906~ and 5-11.8kb. The results of two-pyroxene thermobarometry on the two most cpx-rich rocks are also given on Table 3.5. It appears that there is no correlation between the temperatures and pressures obtained by these methods and the present depth of the samples within the Mantle Sequence. Browning (1982) notes that on a P-T plot most of the data points fall on the oceanic geotherm of Mercier (1980). The olivine-spinel geothermometer gives 700-850~ using the method of Fabries (1979) and 650-800~ using that of Roeder et al. (1979). The lherzolite (OM9015) gives up to 1000~ using the latter calibration mostly because the freeenergy values adopted by Roeder will give higher values for Al-rich rocks. These relatively low temperatures are clearly the results of sub-solidus re-equilibration (Browning 1982; Brown 1982).

0.1

0.01

I

I

La Ce

I Nd

I

I

I

Sm Eu Gd

1

I

I

Dy

Er

Yb

Fig. 3.9. REE/chrondrite plots for two Semail harzburgites. relatively newly-formed oceanic lithosphere and that the plane of the peridotite solidus (c. 1200~ between depleted harzburgite above and less depleted lherzolite below is a mechanically weak horizon along which decoupling is likely to occur (see Section 4.3 on emplacement mechanisms). 3.2.1.2

5I

Geothermometry and geobarometry

The Mantle Sequence rocks contain a number of potential geothermometers and geobarometers which have been used to Table 3.4 Traverse through the Basal Lherzolite, Wadi Rayy. 0M9026

0M9022

0M9020

0M9016

0M9013

m* SiO2 TiO2 A1203 FeO* MnO MgO CaO Cr203 NiO LOI

c.2000 40.80 nd 1.03 7.67 0.12 40.20 1.07 0.39 0.35 8.42

1750-2000 40.42 nd (0.004) 1.08 7.56 0.13 40.57 0.89 0.33 0.31 9.20

c. 1 5 0 0 42.09 nd 1.31 7.59 0.13 39.16 1.12 0.37 0.29 8.06

750-10(10 41.40 nd (0.013) 1.72 7.69 0.13 39.38 1.75 0.49 0.29 7.17

< 500 40.97 0.04 (0.047) 2.16 7.49 0.13 37.70 2.36 0.37 0.28 7.48

Total

100.05

100.49

100.12

100.02

98.98

5770 8357 2691 2400

24 6300 7000 2231 2415

7518 8714 2558 2294

81 9794 13428 3327 2295

280 12494 18428 2554 2236

ppm$

Ti A1 Ca Cr Ni

m* estimated height of sample above basal thrust in metres. TiO 2 contents: nd--not detected by major element analysis (values in brackets based on recalculated trace element analysis, values given in lower table). -~ total Fe as FeO. $ parts per million, elements recalculated to 100% anhydrous.

Chapter 3

52

Table 3.5. Geothermometry and geobarometry of the Mantle Sequence.

1. Single pyroxene

Mercier (1980) Orthopyroxene

Sample number

Depth*

P(kb)

OM2307 2262 2538 6676 2306 2540 2297 2309 2315 2317 2320 2328 2333 2339 2342 9015

0.0 0.0 0.1 0.2 0.3 0.5 0.6 0.9 2.0 2.6 3.3 4.0 4.4 5.1 5.5 9.0

26.3-28.6 19.8-28.0 16.2-24.7 15.6-26.3 24.8-40.5 16.9-24.0 12.7-23.5 22.3-29.8 17.8-24.7 10.6-47.2 15.0-30.4 20.4-24.7 20.5-25.2 12.6-27.0 4.4-21.9 5.9-52.7

2. Two pyroxenes

Wells (1977)

OM6676

888~

9015

865~

T(~ 1227-1250 1060-1210 1007-1149 1098-1209 1234-1486 1016-1190 1106-1279 1199-1273 1082-1165 993-1466 990-1259 1088-1139 1111-1142 915-1227 989-1158 908-1563

Clinopyroxene P(kb)

T(~

17.3 13.3-18.4 9.1-19.2

1011 192-943 897-1054

26.2

1204

5.0-38.3

844-1207

Powell (1978) 10 kb 20 kb 30 kb 10 kb 20 kb 30 kb

1039~ I052~ 1064~ 1029~ 1045~ 1060~

*Depth within mantle sequence below petrological Moho datum (km) Data taken from Browning (1982)

3.2.2 Dunites and chromitites

Within the dominantly harzburgitic host of the Mantle Sequence are dunitic bodies of varying shapes and sizes that may, or may not, contain chromitites. Descriptions of their field relations given by various workers differ widely, although there is general agreement on the petrography, composition and origin of the dunites (Table 3.6). The authorities whose observations are recorded on Table 3.6 studied different parts of the Semail Nappe so, if they are correct, the shape, size and abundance of the dunite masses, and their distribution and occurrence within the sequence, vary considerably from place to place. It is immediately obvious that, in the Wadi Tayin-Ibra traverse studied by Boudier & Coleman (1981), the dunitic bodies are larger and far more abundant than elsewhere. For, in the northern area the dunite masses do not exceed 600 m in length, yet one body 14 km long has been mapped in the Ibra area. Just as the size and abundance of the dunite masses is variable, so also is their shape. All authors describe the masses as being irregular with anastomosing offshoots. The contacts are identified as sharp or rapidly gradational showing complex interfingering with the enclosing harzburgite; in several cases blocks of harzburgite are enclosed in the dunite. Our own observations agree most clearly with those of Bartholomew (1983) who noted that the contacts are always sharp, irregular and intricate. Most workers agree that contacts indicating intrusion of dunite into harzburgite are present but note that the shape and deformation of the dunite masses are commonly related to the harzbur-

gite foliation. Boudier & Coleman (op. cit.) identify two types of dunite masses; those that are concordant with, and those that are discordant to, the foliation. Both Browning (1982) and Brown (1982) note that some dunites interrupt the harzburgite foliation, whereas Bartholomew (1983) maintains that the foliation cuts through all the dunite bodies but is only obvious where the deformation is strongest. He identifies the shape of the dunites as being directly related to the intensity of the deformation they have suffered (Fig. 3.10). Intrusive relations have been so widely observed that there can be little doubt that the dunites were part of, or resulted from, magmatic bodies that invaded the harzburgite. The relations of these intrusive events to deformation are far less obvious. The simplest case of post-deformational dunite intrusion, as identified by Boudier & Coleman (1981), does not seem to occur in the northern area. Here, although observations are in general agreement, interpretations vary. Brown (1982), noting that the igneous features within the dunite, such as the cumulus textures of chromite layers, were undisturbed, took this as evidence that at least some of the dunites were post-deformational. Bartholomew (1983) confirms these observations but maintains that the centres of the ]east deformed dunite bodies were protected from deformation because the stress was absorbed by the easily recrystallized olivines in the marginal zones. Bartholomew (op. cit.) identified the deformation (foliation) in the harzburgite as stronger in some areas than others. The dunites were intruded at various stages of the deformation history so that those intruded earliest would usually be the most deformed. This model of

The Semail Ophiolite

(a)

53

(c) :'~-'-'." 5"-".-.".'.'.'

.'.'.

Interdigitated contact

[ Harzburgite

Anastomozing

sheets/dykes/veins of dunite

Dunite

(b)

Foliation trace Fig. 3.10. Progressive deformation of a dunite body (taken from Bartholomew (1983)).

Table

3.6. Summary of information on dunite bodies in the Mantle Sequence

Authority-area studied

Abundance, size and shape

Occurrence within sequence

Relation to harzburgite

Rock types and textures

1. Boudier & Coleman (1981) Wadi Tayin Ibra

Abundant Sharp, interforming 50% of fingering outcrop at base of sequence. Largest 14 x 3 km. Shape varies with deformation, markedly flattened at base of sequence

Most abundant near top and base of sequence

Two types identified concordant with and discordant to the foliation. Discordant type cuts foliation

2. Brown (1980, 1982) Northern mountains, mainly Wadi Ragmi and Farfar areas

5-15% of sequence, 1-600 m in length. Irregular, lenticular, anastomosing

Mainly near top of sequence in uppermost 500 m but some large bodies near base

3. Browning < 5 % of sequence, (1982) Wadi Bani 1-10 m. Irregular, Kharus, Rustaq anastomosing block

4. Bartholomew Very variable in (1983) Northern abundance part of (c. 10%) 1-600 m mountains in size. Shape controlled by deformation irregular when undeformed, elongated in foliation when deformed

Contacts

Mineralogy

Origin

Adcumulate dunites with podiform chromite

FOgl t)2

Precipitates from primitive picritic tholeiite magma with some bodies partial melt residue

Some interrupt foliation. Rarely enclose blocks of harzburgite

Complete range from dunite to chromitite. Olivinechrome spinel adcumulates with rare cpx and hbl oikocrysts

Fo~5 , ~ . 3 chrome spine cpx hbl p[ag Fe-Ni sulphides

Magmatic precipitates within rising basaltic magmas

Interfingering, Mainly within rapidly lowermost 2 km gradational but of sequence well-defined

When undeformed seem to cut harzburgite structure. Elongated in foliation when deformed

Adcumulate dunites

Fo,~2

Magmatic

Interdigitating, Very variable. sharp Five areas studied showed wide variation in size, shape and abundance

Invariably have foliation parallel to that of enclosing harzburgite

Interfingering, sharp

Chapter 3

54

magmatic emplacement into a varying and variable stress field, we believe, accounts for all the field observations. Although dunites with occasional and isolated chrome spinel grains forming only 1-2% of the mode are the most common rock type in these bodies, all petrographic varieties from dunite through chromitiferous dunite and olivine chromitite to chromitite occur. In all cases the chromitite is enclosed in a dunitic host although this may only be a few centimetres thick and intensely sheared. Usually, it is the largest dunite bodies that contain the largest chromitite masses and, in general, these lie in the uppermost 1.5 km of the Mantle Sequence just below the Petrological Moho (Brown 1980, 1982). In a detailed study of these deposits Brown (1982) identified structures and textures suggesting that both the olivines and chrome spinels were precipitated from a basic, probably picritic, melt. In particular, the chromitite commonly occurred in layers that showed igneous sedimentation features such as cross-bedding, slumping and differentiation compaction (Plate 3.8). The identification of these textures as dominantly adcumulate stands up to subsequent investigations although Christiansen & Roberts

(in press) reinterpret the "harrisitic" textures of Brown (1982) as due to the formation of olivine neoblasts, The compositional range from dunite to chromitite is due to a concomitant change in the modal abundances of olivine from 99 to 10% and chrome spinel from 1 to 90% (Brown 1982). Brown (op. cit.) also records the occasional presence (cpx---~plag sequence probably results from magmas with higher CaO/ AI20 3 ratios than those which produce the less common ol---~plag---~cpx (MORB) sequence. The early crystallization of A

Q4'/ ~

^_

20 Kbar

-~" "~ " " - ~ - - ~

CMS2 50 DIOPSIDE

40

30

~ CaO/AI203 = 0.65~

20

10

ENSTATITE

79

orthopyroxene apparently represents an even more CaO-rich and Al20~-depleted parent magma. These conclusions are based on the phase relations in the CMAS system (O'Hara 1968) (Fig. 3.31). Olivine fractionation alone cannot alter the CaO/AI20~ ratio and, as this is the main process occurring as the melts rise through the upper mantle, the contrasting primary magma compositions are most probably produced at the site of mantle partial melting. They may be caused by different degrees of melting of a homogeneous mantle or heterogeneities in the composition of the source mantle. Source heterogeneity may be an original feature or the result of an earlier melting episode which depleted the source region particularly in A1203. In general, it seems that the magmas that underwent ol----~plag-->cpx crystallization were derived from a less depleted or more '~fertile" source than those which display the ol-->cpx-->plag and particularly the ol-->cpx-->opx-*plag crystallization sequences (see Section 3.9 where a quantitative model of the partial melting process is developed).

3.5 T h e H i g h - L e v e l Intrusive In most complete ophiolite complexes, between the underlying layered cumulate sequence and the overlying sheeted dyke complex, there is a generally thin and irregular unit of "highlevel" gabbros that are distinguishable from the underlying plutonic rocks by their variable texture and the absence of well-defined igneous layering. These rocks were first described by Wilson (1959) from the Troodos Massif on Cyprus and were interpreted by Bear (1960) as having been produced by a basic melt chilling against the roof of a magma chamber. More detailed studies of these rocks on Troodos by Allen (1975) supported the roof-chill origin, and this interpretation has been extended to similar rocks in other ophiolite complexes (Coleman 1977; Dewey & Kidd 1977). Reinhardt (1969 and in Glennie et al. 1974) described a unit of medium grained, non-layered gabbros in the Semail ophiolite and termed it the "hypabyssal gabbroid unit". Aldiss (1978) showed that the Semail high-level gabbros were very similar to the corresponding rocks of the Troodos complex and that they contained small bodies of differentiated intermediate to acid plagiogranites (we follow Coleman & Peterman (1975) and use the general term plagiogranite to collectively describe the quartz diorites, tonalites and trondhjemites found in the upper levels of ophiolite plutonic complexes). Aldiss (op. cit.) proposed that these rocks formed by the crystallization of hydrous melts in a magma chamber roof zone where magmatic volatiles were concentrated and seawater convecting through the oceanic crust entered the magma via the stoping of hydrothermally altered blocks of sheeted dykes. The High-level Intrusives are distinguished from the petrologically similar but later cross-cutting plutonic bodies, the "Late Intrusive Complexes" (described in Section 3.6), by their stratigraphic position between the Layered Series and the Sheeted Dyke Complex and by their non-intrusive but generally gradational contacts with these units.

3.5.1 Field relations Fig. 3.31. Projection from olivine onto the CS-MS-A plane of the CMAS tetrahedron (O'Hara 1968) showing the compositions of the Semail harzburgites (dots), Lherzolite (open circle) and Browning's primary magma composition (triangle). 1 bar and 20 kbar phase boundaries shown and projected fractional path of the Semail magmas (dashed arrowed lines) (from Browning 1982).

The Semail High-level Intrusives form a discontinuous unit, up to 700 m thick but generally less than 200 m and sometimes only a few metres thick or absent, at the top of the plutonic sequence (Fig. 3.32). The sequence of rock types is often complex and variable from section to section. The transition

80

Chapter 3

-F

111111 7.';',"-'

T i

"1 \ I - L

-i

i,

+++4-+

-;:b~

",, t ' - x

~

'L" ( 5 - ' , "

t

2,'_.'2'z~l

T

,-2 ,',, C

I i " ; l ~ / Ii

;' -",)~d

:;.,',.1 r,-',,-t.

x/i

_-

i

_~.~,

,?;7:7? + +§ ~ § .,~ 9,4"I- i , l ^ \ \\\\

\', \ \ "1

+

Xx\\\/

"-'\ \ \ \

I "/~',',",\

t

it

.'-',jiLl

mt

"- 7".. '.',q

'.: ,','.d

I

ol

ol ~ - ~ Dykes

ol

•

Isotropic gabbro

~

Laminatedgabbro

ol

Layeredgabbro

I magnetite-bearing gabbros mt

m ~ Plagiogranite ~ Xenoliths ~ ~ Coarsepegmatitic gabbro-diorite

I olivine-bearinggabbros

from layered to non-layered rocks occurs gradational!y over a transition zone a few tens of metres thick where the compositional layering, which is often steeply dipping in the uppermost cumulates, dies out up section. There usually are no marked mineralogical changes across the contact but olivine is rare in the high-level gabbros and disappears up section where titanomagnetite becomes more abundant giving rise to ferrogabbros with up to 20% opaque oxides. Hopson & Pallister (1981) recognize an intermediate unit of laminated but non-layered olivine gabbros and gabbros between the cumulate gabbros

Plate 3.20. (a) Wadi Salahi, plagiogranite bodies at the base of the Sheeted Dyke Complex. High-level gabbros poorly exposed in the foreground.

Fig. 3.32. Representative sections through the high-level intrusive sequence (modified after Browning 1982).

proper and the high-level gabbros but we find that patches of laminated gabbro occur throughout the high-level sequence and frequently grade into non-laminated rocks. The plagiogranites occur as light-coloured veins and dykes throughout the gabbros increasing in abundance upwards where they coalesce to form larger 5-50 m thick sheet-like bodies in the upper part of the sequence (Plate 3.20(a)). Here they locally intrude into the base of the sheeted dyke complex although they may in turn be cut by later basic dykes (Plate 3.20(c)). The largest plagiogranite bodies are characterized by sharp,

The Semail Ophiolite

intrusive contacts and often contain xenoliths of gabbros or, more commonly, metadolerite with angular to rounded shapes and representing partially assimilated blocks of dyke stoped from the overlying sheeted complex. In addition, some plagiogranite bodies consist of several distinct intrusions that can be identified by differences in grain size, textures and xenolith populations.

3.5.2 Petrology 3.5.2.1 Gabbros and diorites The high-level gabbros are characterized for the most part by medium grained (1-5 mm) hypidiomorphic to ophitic textures (Plate 3.21). They are non-layered but occasionally laminated with planar orientations of the constituent minerals, particularly plagioclase. Pegmatitic segregation patches and veins, with up to 5 cm grain size, are common and have diffuse, gradational or sharp and cross-cutting contacts with the finer grained host. The pegmatites contain patches of leucogabbro and diorite which Aldiss (1978) interprets as residual liquids that became trapped in the solidifying gabbro. The high-level gabbros are composed largely of unaltered zoned plagioclase (An84_5~) and partly altered diopsidic augite (Wo46.5._49.5En4~48Fs5.5_lO.5). In contrast to the cumulate gabbros, where zoning of the mineral phases is very limited and

Plate 3.20 (Cont.) (b) Fine-grained gabbro-diorite netveined by plagiogranite. (c) Plagiogranite cut by sheeted dykes near Fujayg, Wadi Sarami (Figure for scale). Plate 3.21. High-level gabbro. Zoned plagioclase showing patchy alteration and clinopyroxene showing partial marginal replacement by actinolite. Hypidiomorphic granular texture.

8I

82

Chapter 3

often reversed, the feldspars, and to a lesser extent the pyroxenes, in the high-level gabbros are strongly normally zoned. Brown and green hornblendes with a wide range of compositions (Fig. 3.33) occur as an interstitial phase or as pyroxene overgrowths. The Ti and Al-rich compositions reflect primary magmatic crystallization. The pyroxenes are partly altered to fibrous green actinolite (uralitization) containing inclusions of fine grained sphene and magnetite as additional secondary alteration products. Small (

0O

1

:u

--t

q m

m

.< q

-~ q

m

m

_ -4

q m

m

,

Depth km SALAHI UNIT

I I

Cpx-~ UNIT ALLEY UNIT

I LASAIL UNIT

q rn

I

I

i

i

J I

I

I

I

200-250~

I

S

Lower Greenschist facies

i I

GEOTIMES UNIT

I

0

i i

o

c. 350 ~ C

3-

SHEETED DYKE COMPLEX

1 HIGH LEVEL INTRUSIVES

T

A

1::u

Upper Greenschist facies

~ ,~ S

c. 475~

o ~. O

i i

Actinolite facies

'

__

~~

I

Fig. 3.67. Metamorphic facies and distribution of mineral phases in the upper part of the Semail ophiolite.

morphism of ocean-floor type (Coleman 1977; Smewing 1980a; Pallister 1981; Pallister & Hopson 1981). Gregory & Taylor (1981) have shown, on the basis of ~sO isotope studies, that seawater alteration or re-equilibration affected much of the layered sequence but left little or no mineralogical record. Alabaster et al. (1980), Alabaster (1982) and Alabaster & Pearce (1985) have demonstrated a complex metamorphic history for the Semail lava sequence involving several phases of alteration related to cycles of magmatic/hydrothermal activity. The metamorphic assemblages are here described from lowest to highest grade which, in general, represent increasing depth in the ophiolite succession (Fig. 3.67). 3.10.2.1 Brownstone facies

The brownstone facies is characterized by low temperature clays, particularly a K-rich dioctahedral Fe-illite, formed by the alteration of olivine and interstitial glass under oxidizing conditions, and Mg-rich trioctahedral smectite formed under more reducing conditions (Cann 1979). Alabaster (1982) has identified iron smectites in the groundmass of the Geotimes lavas which he claims are relicts of an early brownstone facies now mostly overprinted by the lower greenschist facies assemblage that is typical of these rocks.

3.10.2.2 Zeolite facies The Alley Unit basalts contain abundant zeolite minerals, such as stilbite and mesolite, along with celadonite they are found infilling and lining vesicles and sometimes replacing igneous minerals. Stilbite is particularly common and replaces plagioclase and occurs as large sheaf-like radiating aggregates in the groundmass of these lavas. It is also sometimes found as vesicle and cavity infillings in the Geotimes Unit. Mesolite and celadonite are confined to the Alley Unit, the latter occurring as a dark green groundmass mineral and as vesicle lining. In these rocks chlorite is partly replaced by smectite and it appears that the zeolite facies assemblage overprints an earlier lower greenschist facies metamorphism (Alabaster & Pearce 1985). The calcium zeolite laumontite is present in cavities in the upper part of the Lasail Unit which indicates retrogressive metamorphism of these rocks at temperatures of 150-250~ (Liou 1971). 3.10.2.3 Greenschist facies

Most of the sheeted dykes and the extrusive rocks of the Semail ophiolite have been affected by greenschist facies metamorphism. The metamorphic mineral assemblages are complex and there is a general heterogeneity and irregular

I26

Chapter 3

distribution of many of the secondary minerals. This, combined with the presence of relict igneous minerals (mainly clinopyroxenes) in relatively low grade metamorphic assemblages suggests that equilibrium in the temperature range 250450~ was only rarely and locally attained. Olivines are altered to mixtures of chlorite and epidote, clinopyroxenes to actinolite and chlorite and plagioclase to albite, prehnite, epidote and quartz. Sphene and disseminated Fe-oxides replace titanomagnetites, and epidote, chlorite, prehnite, Fe-oxides and pyrite occur in the groundmass and infill cavities and veins. The actinolites in the upper greenschist facies rocks have AI203 c o n t e n t s of 2.06-4.36% which overlap with the range of values in the upper actinolite facies rocks. Actinolite replacing clinopyroxene is common in the sheeted dykes but less so in the extrusives where chlorite is the dominant alteration product of ferromagnesian minerals. The chlorites have a wide range of compositions (Fe/(Fe+Mg) 0.2-0.6); fibrous green Fe-chlorites (Fe/(Fe+Mg) >0.4) mainly replace pyroxene and/ or actinolite. The chlorite that replaces olivine is less iron-rich (Fe/Fe+Mg 0.2-0.4) (Fig. 3.68). The general absence of actinolite and the greater preservation of clinopyroxenes in the volcanics suggests that temperatures of metamorphism were lower than in the sheeted dykes. Ellis & Green (1979) suggest temperatures of about 350~ for the transition although secondary amphibole formation from clinopyroxene is also suppressed by high f 9 2 (Liou et al. 1974). Epidotes are common throughout the sheeted dykes and lower lava units (Geotimes and Lasail Units). In the dykes they are most abundant in the green epidosite dykes and breccias that were channelways for hydrothermal fluids. In the Geotimes lavas epidote occurs mainly as cavity and vein infil-

0.7 cr

0.6

O5

1O 1~ O1

O5

o

0.5 Fe Fe + Mg

a

2 ~F-

4 0 ~ 94

z
50) and suggest that they probably form as the upflow limbs of hydrothermal circulation systems where the fluids are highly channelized. 3.10.2.4 Actinolite facies

Most of the high-level gabbros of both the axis sequence and the late intrusive complexes and some of the dykes at the base of the sheeted complex are characterized by an association of primary calcic plagioclase, the unaltered nature of which is

The Semail

evident from optical and microprobe studies which show normal compositional zoning in the range Ango_so, and a secondary Al-poor brown, or more commonly green, actinolite or actinolitic hornblende replacing clinopyroxene or, in rare cases, hornblende. Primary Fe-Ti oxides, mainly titanomagnetites or magnetites with exsolution lamellae of ilmenite, are largely unaltered in these rocks. The AI203 contents of the secondary amphiboles range from 2.4 to 5.9% suggesting that the boundary between the upper and lower actinolite (greenschist) facies amphiboles placed at 5% AI20 3 by Elthon & Stern (1978) does not apply to the Semail rocks. Analyses of clinopyroxene-actinolite pairs from a high-level gabbro show increases in AI, Fe, Ti, Na and K and a decrease in Ca content with alteration (Alabaster 1982). Mg/(Mg+Fe) decreases from 0.79 in the pyroxenes to 0.73-0.63 in the actinolites. The secondary amphiboles are often fibrous and clearly replacive although they occasionally have subhedral prismatic forms and are difficult to distinguish from primary brown-green hornblendes, although the primary amphiboles usually have markedly higher A120 3 and TiO2 contents than the secondary ones. The high-level gabbros often show patchy replacement by greenschist facies assemblages such as saussuritization of the feldspars and by the presence of cavity and vein infillings of epidote, prehnite and quartz. In the associated diorites and plagiogranites greenschist facies assemblages predominate and the feldspars are usually completely altered except for oligoclase-albite rims and overgrowths.

3.10.3 Summary In the upper part of the Semail ophiolite there is a general increase in metamorphic grade from Brownstone and lower greenschist facies (350-250~ in the lavas and from upper greenschist (350-475~ to actinolite (475-550~ facies in the sheeted dykes and high-level gabbros (Fig. 3.67). The probable temperatures of the facies boundaries suggest an overall temperature gradient through the 4 km of section of about 150~ However, the pattern is complicated by several features: (i) Most of the rocks show evidence of several phases or pulses of metamorphism, usually of a retrogressive nature with lower grade assemblages overprinting higher grade ones. Commonly the highest grade event is a bulk rock alteration which is partly overprinted by lower grade alteration confined to dispersed vein and cavity infilling. For example, most of the zeolite facies alteration in the Alley Unit lavas is a late-stage vein and cavity alteration whereas the groundmass of the rock remains in lower greenschist facies. Likewise, the actinolite facies gabbros often contain greenschist facies vein assemblages. These features indicate several episodes of different temperature hydrothermal fluids and renewed convection triggered by new cycles of intrusive and extrusive activity. Alabaster (1982) and Alabaster & Pearce (1985) suggest that the metamorphic history of the lavas is complex because of the intense but localized hydrothermal activity related to the Lasail and Alley Unit magmatism. (ii) The presence of lower greenschist assemblages (chloritealbite-epidote) in all the lavas units, up to and including the uppermost Salahi Unit, shows that the whole volcanic section was subjected to temperatures of at least 200-250~ This implies a high geothermal gradient near to the surface. Alabaster & Pearce (1985) suggest that the late pervasive greenschist facies overprint of the entire eruptive sequence occurred at a

Ophiolite

I27

late stage and was caused by the "sealing in" of the hydrothermal system by the sedimentary cap overlying the lavas. They cite Davis & Lister (1977) who showed theoretically that temperatures at the base of an impervious sediment cap only a few metres thick overlying young oceanic crust could reach 200~ (iii) There are areas of unusually high grade metamorphism where stratigraphically higher units show mineral assemblages typical of lower ones. For example, in the Lasail area, the Lasail basalts which form the hanging-wall to the massive sulphide ore-body contain actinolite and show upper greenschist facies alteration more typical of the Sheeted Dyke Complex. In most other areas they are characterized by an epidoteprehnite-chlorite lower greenschist assemblage.

3.10.4 Massive sulphide deposits The Extrusive Sequence contains numerous gossans and other showings (including slag heaps and other evidence of ancient Sumarian working) of massive base metal sulphide mineralization (Bailey & Coleman 1975: Coleman et al. 1979). Following on from a prospecting programme using geophysical techniques and borehole sampling, three of the largest deposits in the Wadi Jizi area, at Lasail, Bayda and Aarja, are currently being mined by the Oman Mining Company. The geological setting of the ores has been described b)' Alabaster et al. (1980) and Alabaster (1982). Ixer et al. (1984) have made a detailed study of mineralogy and geochemistry of the ores and Alabaster and Pearce (1985) discuss the relationship between the hydrothermal metamorphism of the lavas and ore deposition. The deposits are typical exhalative '~Cyprus-type", such as those described from the Troodos massif (e.g. Constantiniou & Govett 1973) and subsequently recognized as forming at hydrothermal vents on modern spreading axes (Francheteau et al. 1979; Oudin et al. 1981). Haymon et al. (1984) have described fossil worm tubes, similar to those found in modern vent communities, in the Bayda massive sulphides. Chen & Pallister (1981) show that lead isotope data from the Fe-Cu sulphides indicate a close genetic relation with the igneous rocks of the ophiolite suggesting that the source of the sulphide is the oceanic crust (Section 3.12.1). The gossans overlying the exposed ore bodies form areas of brightly coloured ground up to 100 m across (Lasail) and contain a variety of secondary minerals, including limonite, jarosite, haematite, azurite, malachite and native copper. The underlying mineralized stockwork zones of altered volcanics beneath the ore bodies are characterized by quartz-pyritesphene-chalcopyrite mineralization in amygdales and crosscutting veinlets. The primary ore bodies themselves are essentially massive pyritic ores containing varied amounts of chalcopyrite and sphalerite. They consist largely of massive early-formed euhedral or colloform pyrite with some associated haematite-magnetite and chalcopyrite, secondary chalcopyrite-sphalerite-bornite (mostly in fractures in the pyrite), and late-stage replacements or veins of haematite-quartz (Ixer et al. 1984). Apart from quartz, other gangue minerals present are calcite, gypsum, chlorite and epidote. Within the ore, values of iron and copper are erratic and copper contents >0.5 wt % only occur where chalcopyrite is a conspicuous phase. Except locally, where sphalerite is present, zinc values are generally low (