OXFORD
Geographic Information Systems Applications in Natural Resource Management
1
Second Edition
Geographic Infor...
116 downloads
1826 Views
47MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
OXFORD
Geographic Information Systems Applications in Natural Resource Management
1
Second Edition
Geographic Information Systems Applications in Natural Resource Management
Michael G. Wing Pete Bettinger
OXFORD UNIVERS ITY PRESS UNIVERSITY
2
OXFORD U:-:lvr SIV I RSITV ASITY Pitt's!> l'Rt'SS
rt' 204, Don Mills, Omario OH5 8H Sampson Mcw$, Mews. Sui Suite O lllario M3 M3C OHS www.oupcanada.com Oxford Uni University ress is a department of the Uni\'cr.)ity versiry ('Press Unive rsity of Oxfo rd. rd . II furt hers (he rsity's objective objccrivc of excellence in resea rch. rch . scholarship. if urthcrs rhe Unive rsiry's scholarship, :md cd uc.1.r ion by puhlishing worldwidee in and cduc uion publi shing worldwid
Oxford New York Yo rk Auckland Cape Town Dar cs Hong Karach i es Salaam H ong Kong Karachi Kuala Lumpur LUIll I)Uf Madrid Melbourne Mexico Ciry C iry Nairobi New Delhi Shanghai Taipl·j Taipei 1'0(0111'0 Toronto Wi \'(Iith th oHiees offices in AIgcmina Argentin a Auscria Austria Brnil Braz.il C hiJ hilec Czech 7.Cch Republic Fl":1ncc France Greece G reece GuatemaJa IraJy Japan Poland Portugal Sin ingapore ga pore Guatemala Hungary Italy South Korea Swirtcrland Swirlcrl and Thailand T hailand Turkey Ukraine Ukrain e Viclnam ViclIlam
Oxford is i~ a.a lradc Hade m:ak mark of orOxford xford Universiry Press in the UK and in cenai certainn other counrrics couillfies rublished Publi~ht'd in Can:1da Canada by Oxfo Ox fo rll rd Un Ullivcr!liry iversity Press Copyri opyright ght @ Oxford Unive rsity rsiTY Press Canada an:ld:l 2008
The T he 111 moral r.tl fight:. rights of the author have been asscncd asserted Database Unii\'crsiry versity Press (maker) D:uaba.se right Oxford Un Fi rst published 2008 First
All riglus righrs rcserw(l. rese rved. Non pari of nf tlli thi s publication puhlicatioll may m3Ybe reproduced. r cp roduc~d . :.wred retrieval SY:.fem, system , or transmitt transm iu ed oo., in any fortn form or by any means, merub. sto red in a relrleval wit hout the prior permission perm issio n in wri(ing Oxfo rd University U niversity Press. without writing of Oxford o r 3.) a ~ ex perrnill ed hy by Jaw. law, oorr und er terms agreed or cxppressly rc.ss ly pcrmitlccl under agre~d with the appropriate reprographics n:prographics right rightss organization, organizati on. Enquiries Etlgui ries concern co ncernin ingg reproduction Ilit side sidc the scope of th thee above ahove should IX" be :.em sc nt to Ihe the Ri Right ghtss Depanmel1l Department . out Oxford Ox ford U Uni\'ersit"y niversiry PrC.)$, Press. al at lil th et address above. 110 t circulate ircuiale thi.) this book in any other OIh er bind binding ing oorr cover You must musl 1101 any :lequirer. acgu iTer. must impose this and you muSl Ihis sallle condition condirion on :lily
Librnry Li brary ~nd and Archives Canada Cataloguing in Puhlic:1 Publication lioll Wing. Mid13d M ichael G
Gcogr.tphic SYSWllS : applications in forestry and n:1IlIral nafUr.l.1 Geographic information synems: resources management J/ M Michael ichael G. G. W Wing ing & Pete Pele Benin I3cltinger. ger.- -lnd 2nd cd c:d.. Previous Pr('ViollS: cds. t'd->ouo.lv _ boo p' ....... bIrt "'-0
"!wl"~""
nIL ....... ;,,;....,..... _
o. _
wi,,,,""'1hoo
u,.,.a.f..w.. "f'M .... 5""'tw;! .. iIhon cIo+r
1orIoiI.
~
""' ... bot ~'II
I"""" ..."'" rho- - ' . . , . - l ""... th."'I!h ,,..., . _ .... \ bot __. 1«.1 dootIj wpon Ihr ...plftlh;.h.oIIbt ................ ltt.tn ' ........uIhe""'lOoninw...·.oI tr.
2. V............"')'.... ~
.... ..-.- .....". 1ft...
,Iw..,. "'" oPf>M""
wttIC4II ...... _
....
-..",.,.,;1.,..
~.
~~- .... ,..,..........~ .......... • ... r~~.".,"'""" MftI."'Y ., .., ..., .or ... -.II t... """f'U"'I "'" , . - '" "'....... ",,,,- Ioc.o_ .. ...... _ ..., .......n ap........ ~II _~ r - - .........."'"""-'d by ............. 4 hr~ • ",........ rht. """~ to. nwde l') ,.... pn>d ........ ~ .~~ ...."'.. whkh .ho.lI ...... Jo1s • ....r
...
'Injr (-...-..... p'l"''''.f'''W'"''td "" ~""""" ·ill.. ""'I' It ... ~.!I"_ 01. I~.tm
~~"...,tup_ ~ . r.!4.00J..c.0Io,.....IIN .... p of It.Ml< h.~1oot "" ... ,. c-Irudlooto _
... .. 011 r...kt.r o.nol ~ <mi' Io.:F"". _ ......... .......-_~~ ....... ... '"*"-..-J~."'~,-. }..J. ~ _ ... _
1>IIiCrif'I .... "" _ ......... •.............._""_rlO . . pu/>Ikht
sand board feel feet (MBF) per ac acre re of timber volume' volume? Amibure: Amibute: MBF Relational Rela[ional operaro operamc: r:
per hectare (TP (TPH) H)?' Amibute: TPH Relational operawc: operator: >> (greater than) than) Threshold value: value: 700 Query: TPH > 700 Answer: 5 stands (4.5.7.9, (4.5.7.9. and 10)
f) How many dmber timber srands stands have at least 950 95 0 trees per hectare? per Anribure: Anribule: TPH
hectares in size? AttriblHe: hectares Attribute: Relational operator: >> (greater than) Threshold value: 100 Query: hectares > 100 Answer: I stand (3)
Multiple criteria queries criteriaa queries are com combinations Multiple criteri binations of single criterion queries, queries. held together rogether by logical operators (and. or,
1/ot). not). They allow you to ro develop a complex query wi without thout havi ng [0 having to perform several single cr criterio iterion n queries in mulriple criteri criteria a queries mat that sequence. Below are several multiple
relate to the rhe dara data fou found nd in Table 5.1. 5. !. a) How many c.imber timber stands are less than or equal to 20 years yea rs of age, and contain more than rhan 950 trees
per hectare (TPH)? Attribu(es: Attributes: age. TPH Relational operarors: operators: Age: S; S (less than or equal to)
TPH : > (grea (greater TPH: ter than) Threshold values: Age: 20 TPH : 950 TPH: Logical operator: and Query: (age S; Query: S 20) and (TPH > 950) Answer: Answer: 4 srands stands (5.7,9, (5.7.9. and 10) b) How many ti mber stands are ar old, at least 25 years old. co ntain at least 10 thousand board feer or contain feet (10 MBF) per ac re of timber rimber volume? volume? Attributes: age, age. MBF Relational operators: Age: ~ (greater [han Age: than or equal to) MBF: ~ (greater than or equal to) MBF: Threshold values: val ues: Age: 25 MBF: 10 Logical ooperator: peraror: or Query: (age
Chapter 5 Selecting landscape Features c) How many timber stands are at least 20 years old, and are no older than 30 years old, and contain more than 500 trees per hectare? Amibutes: age, age, TPH Relational operators: Age: ;" (greater than or equal to) Age: S (less than or equal to) TPH : > (greater than) T hreshold values: Age: 20 Age: 30 TPH: 500 Logical operators: and, and Query: (age;" 20) and (age S 30) and (TI'H > 500) Answer: I srand (7) To illustrate the use of a complex query, we wi ll ask a few questions regarding the polygons contained in the Brown Tract stands GIS database. First, assume that the managers of the Brown Tract are interested in managing the forest for timber production, and maximizing the growth potential of (he [fees in the forest. One way ro achieve this goal may be to use precommercial thinning. As a result, (hey need to understand whether any potencial commercial thinning opportunities exist. Assume that the criteria developed by the managers of the Brown Tract [Q assist in (he analysis was based on four ideas:
be between 30 and 40 years old, the land allocation should include only the even-aged Stands, and the timber volume prior to thinning must be above 9 MBF per acre. The criteria, placed within the structure of a query then becomes: (age;" 30) and (age S 40) and (MBF ;" 9) and (land allocation = 'even-aged') The resulting eight stands (42 hecta res) on the Brown Tract that conform to this query are illusuated in Figure 5.1. These areas can be considered. the poremial commercial thinn ing opportun ities for the fo rest in the near future.
Selecting features from a previously selected set of features Rather than develop a long, complex q uery containing multiple criteria. you can design a set of less complex quer ies that are hierarchical in nature and that reduce the landscape features contained in the set of selected landscape features with each additional query. This process may help you stay organized and prevent the occurren ce of mistakes that may be difficult to understand when usi ng a long and complex query. To selec, landscape feacures from a previously selected set of landscape features, a number of single criterion queries are assembled .
l. Thinning should occur about 10 to 15 years prior to the fin al harvest age assumed by the organization
(45-50 years). 2. Enough crop crees should remain un-cut in the thinned stands so that they (the residual trees) sufficiently respond (within increased growth rates) to the increased ava ilabili ty of light, water, and nutrients for the remaining 10-15 years prior to final harvest. 3 . Commercial thinning will only be applied to evenaged forested stands. 4 . Commercial thinning operations should remove, at a mInimUm, 10 MBF per hectare (abou, 4 MBF per acre). Because the managers have specified a minimum residual volume level the dmber volume per unit area prior to thinning should be substantially greater. The criteria for the query that the managers of the forest decide to use includes the age of the stands that could be thinned muSt
95
Figure 5.1 Stands on the Brown T net that meet the following criteria: age 2: 30 and age :5 40 and MBF 2: 9 and land allocation . 'even-aged'.
106
96
Part 2 Applying GIS to Natural Resource Management
Example c, presemed ea rlier. involved the following multiple criteria query,
(age ~ 20) and (age
s: 30) and (TPH > 500)
which could be subdivided intO three single criterion quenes: age ~ 20 age s: 30 TPH > 500 Each of these can be performed in sequence; the first from the full set of stands GIS database landscape features, age
~
20
(6 stands [1,2,3,6,7 and 8])
the second from the set of 6 landsca pe features that were selected (sta nds 1,2,3,6,7 and 8), age s: 30
(5 stands [1,3,6,7 and 8])
and the third from the remaining 5 landscape features (stands 1,3,6,7 and 8),
TPH > 500
(I stand [7])
resulting in the same landscape feature selected as when
the multiple criteria query was used. The preference for a particular technique (selecting landscape features from a previously selected set or selecting landscape features using a multiple criteria query) wi ll vary from user [0 user, depending on each user's confidence and experience. If you were co try this hierarchical process of selecting landscape features on the Brown Tract thinning example from above, w here rhe criteria was,
Figure 5.2 Stands on {he Brown Tract that meet the following criterion: age' ~ 30.
Breaking down a complex query into smaller, single criterion queries may not work when the logical operawr
involved is 'or'. In the following example, the complex query cannor be broken down into rhree single criterion queries.
(age
= 29) or (age = 30) and (TPH
> 500)
The set of stands that might comprise TPH > 500 can be su bdivided into those that are 29 years old. However, the resulring set cannot further be subdivided into sra nds (har
(age ~ 30) and (age s: 40) and (MBF ~ 9) and (land allocation = 'even-aged') you could subdivide the querying process into four steps. Ste p I: Select from rhe entire set of stands those
stands where age ~ 30 (result: 212 stands shown in Figure 5.2). Step 2: Select from the 212 previously selected stands, those stands where age s: 40 (result: 23 stands shown in Figure 5.3). Step 3: Select from the 23 previously selected stands, those stands where MBF ~ 9 (result: 9 stands shown in Figure 5.4). Step 4: Select from the 9 previously selected stands, those scands where the land allocation is even-aged (resu le: 8 stands shown in Figure
5.1).
Figure 5.3 Stands from the previously sciecte'd set (age 2: 30) on the Brown Tract that meet the fo llowing criterion: age S 40.
107
Chapter 5 Selecting Landscape Features
(age
97
= 29) or (soil_rype = 'PR') and (TPH > 500)
Again, we can locate the stands where TPH > 500, and from those we can locate the stands that have an age of29 years. However. there may be many other stands beyond those in the resulting set that have TPH > 500 and a so il
rype of 'PR' (yee an age that is nOt 29 years). However, the following multiple criteria query could be broken down into three single criterion queries:
(age> 28) or (age < 31) and (TPH > 500) Here, the set of stands that might comprise TPH > 500 can be subdivided into those that are greater than 28 years
old. The resulting Set can furrher be subdivided into stands that are less than 31 years old. Figure 5.4 Stands &om the previously ~Iected set (age O!: 30 and age :5 40) on the Brown Tract that meet the following criterion: MBF O!: 9]
are 30 years old (they are 29 years old) . Similarly, in the following example, the complex query cannot be broken down into three single criterion queries.
One of the most common mistakes made when asking questions of databases is that results are often accepted as 'truth' without considering whether results are reasonable. For example, the Brown Tract timber scands
GIS database contains a number of polygons th at, when summed, describe a 2,123 hectare area. Within the Brown Tract a variety of ages of forests, ranging from recent c1earcuts (age = 0) to older stands. are present. To describe the current structure of the
Inverting a selection Occasionally, you may find yourself in a situation where you need to understand {wo aspects of the spatial features contained in a GIS database: what is the condition (state or characteristic) of one set of features, and what is the
forest, as in this case, the sum of area represented by
the multiple queries should equal the sum of the resources in the original GIS database. If the sum of the area in the age classes is greater than the size of the Brown Tract, some areas were double-counted, perhaps using queries such as these,
Age class I: Age class 2:
(age (age
~
0) and (age" (0) ~ 10) and (age" 20)
Brown Tract, you could develop an age class distribution that indicates the area within. say. la-year age classes. After performing queries of the various forest age classes, the sum of the area queried should nor
result in more or less than 2,123 hectares (the size of the properry). You should always ask yourself whether the results obtained seem reasonab le, given the resources being queried. Whenever possible , if a method of verifying results is available, it is advisable to check yo ur work or have a colleague check your work. If multiple queries are performed that are designed to completely describe the resources of the
where the area of I O-year-old stands is included in both classes. If the sum of the area in the age classes is less than the size of the Brown Tract. some areas were nor counted. perhaps using queries such as these.
Age class I: Age class 2:
(age> 0) and (age < 10) (age> 10) and (age < 20)
where the area of O-year-old stands (c1earcurs) is not
included in age class I , and the area of IO-yea r-old srands is nor included in either age class.
108
98
Part 2 Applying GIS to Natural Resource Management
condition of everyth ing else. Two sets of queries can be
developed to identifY these two sets of features; howeve r, if rhe second set contains 'everything else', simply inverting rhe selected fearures after rhe first query will produce rhe second set. For example, if you were in terested in
available for the Pheasant Hill planning area of the QU'Appdle River Valley in central Saskatchewan. Here, we have created a GIS database that contains so il s, topography, and land classification information . There are 168
polygons with in the GIS database. Assume that we, as nat-
understanding how much land area was considered 'reserved' on the Brown Tract, and rhen how much land area remained 'un-reserved', you could first develop rhe
ural resource managers, are interested in understanding the land areas that contain clayey soils, have steep or
query for rhe reserved areas,
having no significant lim ita cion as they pertain to agricul-
undulating topography, and that have been categorized as tural practices. Initially, we could develop a multiple cri-
(land allocarion = 'meadow') or (land allocatio n = 'research') or (land allocation = 'oak woodland') or (land allocation = 'rock pit') and find that it contains 42 stands covering about 229 hectares. By then inverting the selection, you will find that what rema ins is a ser of 241 stands covering abom 1,893 hectares. A second query for even-aged. unevenaged. and shelrerwood stands was not necessary. The inven selection technique simply switches the GIS database selections, so that featu res previously selected are no longer selected. and vice versa. Some G IS software will
teria query to select from the larger set of features only those that have clayey soil types. There are a number of
soil types in the Pheasant Hill planning area, thus the query would be designed something like this: (Soil_type = ' Indian Head Clay') or (Soil_type = 'Indian Head Clay Loam') or (Soil_type = ' Indian Head Heavy Clay') or (Soil_ type = 'Oxbow Clay Loam') or (Soil_type = 'Rocanville Clay Loam')
make this capability ava ilable through a menu choice in
Given that a polygon is assigned on ly one soil type. we needed to use the relational operator 'or' in the query rathe r than 'and'. As a result of this multiple criteria
the cabular database window while ocher programs may
query, we find that only 69 of the original 168 polygons
make this capability available through a menu or button
have a clay component in their assoc iated soil type. In order to locate those areas within this sub-set of the landscape features that are also located on undulating or steep slopes. we can perform a second multiple criteria query.
in the spatial database viewing window. Some GIS software programs include borh capabi licies.
Example 1: Find the landscape features in one GIS database by using single and multiple criteria queries and by selecting features from a previously selected set of features A co mbination of query processes can be used if you believe that they are necessary to accurately arrive at the desired set of GIS database features. In this example, we
use a GIS database created from the set of GIS databases
_
(Topography = 'STEEP') or (Topography = 'UN DULATING') and here only select features from the previously selected sub-set oflandscape features (not from the larger, o riginal set of landscape features) . In this case, we find that 28 of the polygons have both the soil characteristics and topO-
Areas that meet the query specifications
C=:J Other areas that do not meet the query specificatioos Figure 5.5 The result of a query for areas with dayey soils, located on steep or undulating topography, and with no limitations for agriculrural practices using GIS databases developed for the Pheasant Hill planning area of the Qu 'Appelle River Valley, Saskatchewan ( 1980).
109
Chapter 5 Selecting Landscape Features graphic characreriscics characteriscics of imerest (0 to us. Finally. Finally, we perform a single criteria query to determine how many of the remaining remain ing 28 polygons also have no sign significant ificant limitalimica dons tions for agri agricultural cultural praccices: practices:
me w
' No Significam Significant Limitations) (Land Class = 'No Again, this query is made by selecting from the previously selected set of28 selecred of 28 polygons. As a result of this final query, query. we find lind that thar 18 of the ooriginal riginal 168 polygons have soil, topographic, topographic. and land classification characteristics suitable for our original original na(Unatural resource management managemem analysis a nal ysis (Figure (Figu re 5.5). We could have arrived at the same answer by performing one long, long. mulciple multiple criteria query. Alternatively. we could have arrived at the same answer by using single criteria queries to build up the selected so soilil rypes (adding to the selected selecred ser set each time rime addirional additional polygons that thar have soil rype imeres[ (Q us) attributes of interest us)., then selecting from the previously selected set that have the desired topographic se r those rhose thac ropographic and land classification amibmes. attributes.
Selecting features within some proximity of other features In addition to selecting selecring landscape features fearures based on the set of attributes available within the tabular portion of a GIS database, you can select landscape features based on their spatial relationsh ip to other landscape features. fearures. This ask, for example. example, whidh fearures allows you to ask. which landscape features are within a threshold distance of. adjacenr (0, roo o orr in close proximity proximiry of other landscape features. feamres. For example. you may wam want {Q to know which research resea rch plots are in older forest stands. srands. what forest stands are next (0 to research areas, or which water sources are wirhin within a certain cerrain distance of a road. The abil abiliry ity [Q (0 ask questions quesdons in spatial spadaJ terms rerms is but bur one indication of the power of GIS. The following three [hree exa examples mples provide a description of three common fo forms rms of spatial spadal queries. Example 1: Find the landscape features in one GIS database that are inside landscape features datahase (polygons) contained contamed in another GIS database rhis example, a3 narural interIn this natural resource manager may be ineerested in examining [wo (wo GIS databases: databases; one has landscape features that he oorr she is ineerested imerested in knowing something orner has landscape features that about; the other thar represent areas within which he or she is only concerned. Obviously the second GIS database suggests sugges" that it consists of polygons
99
fe-drures, fearures. since line or point poine feat features ures do nO( not describe areas. Alternatively. A1ternarively. the first firsr GIS database could comain conrain point. point, line. or polygon features. feamres. In this example. example, assume that the first GIS G1S database, the one containing features feamres the manager wishes to know something about. contains points. From a natural resource perspective. managers of the Tracr may be interested in understanding [he Brown Tract the habitat conditions within which certain cerrain wildlife species reside. In order (Q CO collect coUee[ habitat information ir it may be necessary to locate and install forest inventory plots. and Permanent forest sample the characteristics of the forests. Pcrmanem inventory plms. plots. those rhose that have already been installed and are periodically re-measured. re-measured, can also be used for this purpose. From a review of the natu natural ral history hi story of the rhe wildlife interest, you may decide that species of interest. thar only [hose those research plots that resea rch plotS thar are contained within wirhin older forest sstands rands (those at ac least leaSt 100 years old) require measurement. measuremem. Thus the problem becomes one of selecting the plots that reside within w ith in older Stands. stands. T The he GIS database that thai contains co ntains the rhe landscape features fearures of interest inte rest is the research plot GIS database. and the th e GIS GIS database that represents rep resents the older fo foreSt rest areas is the forest stands GIS database. In order ro to complete complere the spatial sparial query. query , you would first firSt select the older o lder stands in the forest stands GIS database of the Brown Tract, using a single criterion query: age ~ 100. The focus of the analysis is then sh ifted ifred to the research plot GIS database. darabase. where the question is posed: how many plots are located within the selected selecred landscape features of the srands stands GIS database (rhe (the older Stands)? stands)? The entire sparial query process, in generic terms, terms. can be described as this two-step two-Step process: (I) select the older stands from the forest stands GIS G1S database using a single si ngle crieerion criterion query, and (2) develop a spa spatial rial query on the rhe researdh research plm plot GIS database daeabase where [he the selection seleceion is performed using the spatial location of selected landscape features from within the fo rest reSt stands GIS database. This spatial selection abiliry may be described as 'selection by location' or some other similarly named menu choice or burton, bu([on. depending on the GIS softwa software re being used. Landscape featu fearures res in the inrersect the research plot GIS database are selected if they intersect fearures in the space covered by the selected landscape features stands GIS database. darabase. The result of this spatial query process should yield 40 research plotS plots that full within the rh e boundaries of older forest stands interested in seands (Figure (Figu re 55.6). .6) . Similarly. Similarly, if you were imeresred locared in you yo ung knowing how many research plots were located ng 110
100
Part 2 Applying GIS to Natural Resource Management
Figure 5.6 Permanent plot point locations within older stands on the Brown Tract.
stands. you would first query the stands database for young srands (perhaps age!> 30). then perform the spatial query simila r [0 the process noted above. The result should yield 1 research plot.
Example 2: Find the landscape features in one GIS database that are close to the landscape features contained within another GIS database In this example. the imerest is again in examining two GIS databases: one has landscape feacures of interest; (he other contains landscape features [hat represent those a reas around which (nor JUSt within which) there is concern. The seco nd GIS dambase suggestS that it consists of polygon features. but here it could also consist of point o r line features. since [he area of concern is the area represented by a zone of proximity arou nd landsca pe features. The first GIS database could also contain point. line, or polygon features. Assume that the GIS database of interest contains point featu res, and that the GIS database that will represenr the area of interest contains line features. The managers of the Brown Tract may be inrerested in developing a fire management plan for the fores t, an d thus would need to understand the types of water resources that are in close proximity to roads. Therefore, the problem becomes one of selecting the water sou rces rh at are within some distance of a road. The GIS database thar contains the landscape features of interest is the water sources GIS database (beca use of the need to know where the approp riate water sources are located). The GIS database that rep resents landscape features around which one
can define an area of concern is the roads GIS database. In order to perform (his spatial query. you muse first determine the distance from the roads that is cri tical for meering the needs of the fire management plan. Assume here th at it is 30 merers, suggesting that water sources within 30 meters of a road may be of benefit to forest fire fighting efforts. This assumes that fi re-fighting vehicles can draw water from these sources and transport the water to the fire area. In the development of the fire managemenr plan, you may have also assumed that only certain rypes of roads can support fire-fighting vehicles. although th is example will proceed under the assumption that aU roads on the Brown Tract can suppOrt these vehicles. A generic description of the spatial query process might then include the following two steps: (1) select all of the landscape features in the water sources GIS database, a nd (2) develop a spatial query on the water sources GIS database where the selection is performed usi ng the spatial location of landscape features con tained in the roads GIS database. Landscape features in the water sources GIS database are selected if they are located within 30 mete rs of any road contained in the roads GIS database. The result of this spatial query process yields 5 water so urces th at lie within 30 meters of a road (Figure 5.7). As you might imagine, this example, as well as the previous example. would also be helpful to those concerned with the proximity of certain resources (water sources, home sites) to potential management activities (herbicide or fertilization applications) , or even to potential sites for wildlife or fisheries studies.
o Figure 5.7 Water source point locations within 30 meters of roads on the Brown T tact.
111
Chapter 5 Selecting Landscape Features
Example 3: Find landscape features from one GIS database tbat are adjacent to otber landscape features in tbe same GIS database In this example, the inrerest is in performing a spatial query that uses landscape features within a single GIS database. Adjacency issues in natural resource management usually concern the placemenr of harvests or the location of habitat) and imply that activities may be prohibited from being implemented next to (or nearby) other receody implemented activities. In the case of habitat development. natural resource managers may desire co develop habitat next to (or nearby) other good wildlife habitat areas. Alternativdy. an invesrmem in research may need co be protected by limiting activity in nearby or su rrounding areas. Since this example concerns adjacency issues, the GIS database used also suggests that it conrains polygon features. This example will assume that a natural resource manager is interested in understanding the extenr and number of stands that are adjacent to research areas. Since the Brown Tract is a working forest that contains some research areas, coo rdin ation of both research and harvesting act ivities is paramount. particularly if the harvestin g activities affect a resource being smdied in the research areas (for example, species of wildlife or hydrologic conditions under canopy). A generic description of the spatial query process might then include the following steps: (1) select the stands in the stands G IS database that are designated a 'research' land allocation, and (2) develop a spatial query on the stands GIS database where the selection is perfo rmed based on how far away other stands are from the research areas. In this case. you can assume that the stands to be queried are 0 meters away from the research areas, and essentially touch a research polygon. Depending on the GIS softwa re program used, the resul , of this spatial
Structured Query Language, or SQL, is the most popular com puter language for querying and manipulating data contained in relational databases. Sometimes simply called 'sequel ', the language allows you to develop quer ies similar to those presented here to access data from large data bases. Although IBM , Oracle. and Microsoft have led the recent developments of SQL. and many other organizations have
101
query process may yield 44 stands, including the research a reas . To remove the research areas from this set of selected landscape featu res. you can perform a single criterion query from the previously selected set of landscape feamres. such as, Land allocation ' Research ' where attribute: land all ocation relatio nal operator: (not equal to) threshold value: research areas By performing this single criterion query on the previously selected set oflandscape featu res. you can select and identify JUSt those stands adjacent to the research areas. Figure 5.8 illustrates the spatial location of the 37 stands that are adjacent to resea rch areas on the Brown Tract.
Figure 5.8 Stands adjacenllo research areas on the Brown Trace .
tailored the SQL language for various applications. the American Nationa l Standards Institute (ANSi) and the International Organization for Standardization (ISO) have developed standard versions and offer them for sale. Some GIS software programs supPOrt the use of the SQL language, and extend it to the management and manipulation of spatial data features. 112
102
Part 2 Applying GIS to Natural Resource Management
Advanced Query Applications Advanced app lications of GIS-related database queries have concentrated on limiting the focus of queries only
discourage some users of GIS. MoS[ of these problems occur because brackets or parentheses are missing from a query; this results in an incomplete query. such as in the two cases shown below from ArcMap que ries.
to rhe features inside a spatial or temporal window
defined by the user. In addition
to
simply providing a
summary of rhe resources contained within a specific area, as the user-defined window (location or time frame) sl ides is expa nded. or is contracted, the query is updated (Q reflect those features that have left rhe window and those [hat have entered rhe window. Queries can be completely re-eva luated as a window slides (or
otherwise changes). or iteratively evaluated by updating the query by considering only the changes that have occurred (Edelsbrunner & Overmars. 1987; Ghanem et al.. 2007). Dynamic queries can be designed that allow users to adjust questions asked of GIS databases by incorporating slider objec[S in a window rather than ask-
in g the user
to
redefine (by typing) the adj ustmentS
needed. For example. a graphical user interface can be
designed to allow users to easily adjust the upper and lower bounds of a quantitative query along a scale, using
a computer mouse (Domingue et al.. 2003). and vide those results quickly.
to
pro-
Syntax Errors Syntax errors that occur when developing queries can
[Height] >= 50} and ([Age] >= 25
(beginning and ending
( [Age] >= 25) and 30 )
{the attribuce 'age' is missing from the second
( S using corporale database
Select and copy newly digitized features
t Paste new features Inlo corporate database
1 Attribute the new spatialfealures
Figure 10.1 acquisition.
1 Edit spatial features (e.g., remove muttipath)
1
Select and delete features to be updated (il necessary)
1 AddGPS features Into corporale database
1 Attribute the new spatial features
Three examples of update processes related to land
from a few hours to a few months. Users of GIS databases will ultimately suggest a variety of enhancements to the databases that would facilitate further ana lyses. For example, the Brown T ract vegetation GIS database could be modified to show more explicitly (he riparian areas, or could include more anributes that describe forest stand structure. A roads GIS database might also be enhanced to show the type of road surfacing or all of the trails (unauthorized roads) that weave through the property. Updating these GIS databases ro include all of the informarion (hat is necessary to make natural resource management decisions may be, however, limited by the time and budget available ro make the changes, the quality of (he information available to make the changes. and other organizational data standards. The needs of namral resource managers. with regard to GIS databases. must eventually be considered along with (he costs of data development.
The Need for Keeping GIS Databases Updated Natural resource managers generally base managemem decisions on the best ava ilable data. The qua lity of data 169
Chapter 10 Updating GIS Databases
can range from very precise and accurate (collected with a high q ual ity GPS receiver) to somewhat imprecise and inaccurate {drawn by hand from memory}. Keeping the data used for making decisions accurate and updated is therefore imponant, and thus the interval between updates becomes important. For example, the update interval that a resource managemem organization uses to refresh the spatial extent {history} of their management activities and the growth of their forest inventory is imponant, since subsequent management decis ions might be affected by previously implemented managemem decisions. The imerval chosen can range from six momhs. to a year, or even (WO years between updates, depending on the GIS database considered. The interval chosen depends on the organization's perception of the usefulness and cost-effectiveness of such an update on a GIS database. For example, if the goal of an organization (e.g., a southern US forest management organization) were to generate revenue for its stockholders, the need for updating the data related to its primary resource (pine forest stands) may be more important, and updated more frequently than data related to secondary resources (hiking trails). Other resources, such as roads, streams, culvens, water resources, and wildlife may be more or less imporram, depending on the goals of the organization, thus the frequency with which these GIS databases are updated may vary according to the organization's perceived need to do so. At the extreme end of the spectrum. every GIS database could be updated continuously; however, the cost of doing so may be qu ite high and the task would require employees {or consultants} dedicated to the task. Two ofche more important questions an organization must address, beyond determining when a GIS database should be updated, are how the update process will be accomplished, and who will do the work. As mentioned earlier, the methods by which a GIS database could be updated vary considerably; the fo rm of input cou ld range from hand-drawn maps to LiDAR-derived measurements, and the GIS processes could involve scanning, digitizing. attributing, and other methods {Table 10.2} . As you may have gathered from chapter 3, when a GIS database is being updated, rhe database is being edited. In some form or fashion , the imem is to change something about a GIS database-either the landscape features or their underlying attributes, or both. Two examples of GIS update processes are now presemed, one related to a forest stands GIS database maimained by a forest industry organization in Florida, and the other related to a streams GIS database
TABLE 10.2
159
Inputs and processes that can be used to a ssist a GIS database update
Input Hand-drawn maps GPS features T2bu lar datab2Ses Field nOles A person's memory GIS Features developed by field personnel Digital orthophotographs and subsequent interpretation GIS processes Scanning Digitiring Updating Joining Linking Copying J p2Sting Attributing Importing Querying and verification
mai ntained and distributed by the Washington State Department of Natural Resou rces.
Example 1: Updating a forest stand GIS database managed by a forest management company A typ ical forest management company in Florida might update their forest stand GIS database on an annual basis. Their field personnel collect information related {Q changes in irs forest land ownership throughout a calendar year, and the forest stand GIS database is updated near the end of the calendar year. Why would they update the forest stand GIS database once a yea r? The forest stand GIS database is arguably the most important GIS database for assisting industrial forest management activiries, and field-level managers require high quality data (maps and inventory data) to make managemem decisions. In addition, most corporations require an annual estimate of the value and volume of resources, for planning and tax reporting purposes. A less frequent updating interval may not be appropriate given the short rotacions typical of southern industrial forestry operations. For example, waiting two years between updates of a timber stand GIS database may represent 8- 10 per cent of the lengrh of a forest harvest rotation. A more frequent imerval, say six months, may provide field personnel with higher qualiry information with which to make management decisions, particularly in cases where a large amount of activity takes place over a six-month period. Some have argued (hat continuously updating GIS databases may be appropriate. but the: time and cost required to update a GIS database may make a nearly continuous update process impractical. Further, field personnel could easily become confused when faced with a cominuously changing set of GIS data170
160
Part 2 Applying GIS to Natural Resource Management
bases. thus updating databases and leaving a window of time {a year. perhaps} between changes may be perceived as more des irable. The changes to the forest stand GIS database that are recommended by field foreSters and other natural resource managers may be indicated o n hard-copy maps and timber cruise forms. or they may be comai ned in d igital databases created in GIS or with GPS. Field foresters. timber procurement managers. or other profess ionals responsible for managing la nd will typically indicate {draw} on maps the changes {e.g .• harvest and regeneration activi ti es} that have occurred on [he forest land base as these act ivit ies have been completed. This informacion is usually sent to a central office {Figure 10.2}. which takes ownership of the timber Stand GIS database. The cemral office checks rhe new data for mistakes and omissions according to a set of organizational standards, and may ask for clarification from the field staff. The information is rhen digitized, either in-house or by an external con n actor. The resuhing digitized GIS database is checked again for mistakes and omissions, and then imegrated into the official {sometimes called 'corporate} GIS
Field office
Central office
.r------ ---- ------- ----,, Delineate changes to be made to a database
,~ , , ,, , ,, , ,, , Make management ,, , decisions using ,, the database , ,, , ,, , ,, , ,, , , ,, Check data for mistakes and omissions
Check data for mistakes and omissions
L Digitize changes
~
~ Check data for mistakes and omissions
r-
,, ,, ,
,, ,
J
L Integrate into corporate database
~ -,
~ Check data for mistakes and omissions
Figurc 10.2 A gcnerali"lcd proces.s for updating a forest stand C IS database.
,
-J
database. and again checked for mistakes and omissions. Finally. the forest stand GIS database is distr ibuted back to the field office. either as a GIS database. or as hard-copy maps and tables. The field office may then have itS own verification procedures for checking the updated database (or maps) fo r m istakes and omissio ns that may have arisen during the update process. Processes such as these. with a systematic method for data collection. entry, and verification, are designed to ensure that high-qual icy data will be developed and avai lable for use in natural resource decision-making contexts.
Example 2: Updating a streams GIS database managed by a state agency In rhe State of Washington, all fo rest harvest plans must be subm itted to t he Department of Natu ral Resources {DNR} for review and approval. A map must accompany each plan, and illustrate the juxtaposition of proposed activities in relation to, among other landscape features, the stream system. To ensure a consistent definition of the 'stream system', the DNR provides (at a minimal COSt, as was illustrated in chapter 3) a streams GIS database for the entire state. This database is conti nuously updated by the DNR as new info rmat ion is co llected. However, processes and protocols exist that are related [Q each potential change to the GIS database. For example, assume a pri vate landowner surveyed a stream reach and noted that the type of stream {and perhaps location of the stream} on the landscape is different than the type of stream illustrated in rhe DNR streams GIS database. The landowner has the option to submit ce rtain documentation [Q the DNR in suppOrt of a request to change the DNR streams GIS database. The DNR d irects each request through a review process, and based on the outcome of the reviews, decides [0 ei ther accept or reject the proposed changes suggested by the landowner. The amount of time required to make a change in the streams GIS database, from initial submission by the la ndowne r to official incorporation in the streams GIS database, may require several months. The process is considered a continuous one since approved changes to [he streams GIS database can be made at any time during a calendar year. Therefore, landowne rs may need to continuously review the statuS of the DNR streams GIS database in the areas where they own or manage land, and acquire updated data as they deem necessary to reflect the latest stream info rmation. 171
Chapter 10 Updating GIS Databases
Updating an Existing GIS Database by Adding New Landscape Features
Updating a stands GIS database
me
Assume that (he owners of Daniel Pickett forest have pu rchased 80 acres (32.38 hectares) of land adjacent to the southwest corner of the original forest boundary (Figure 10.3). Following process B illuStrated in Figure 10. 1. the Stand bounda ri es of this area have been digitized inco a new GIS database chat is separate from the o riginal stands GIS database. and these features have been attributed wim data fields similar to that in the original stands GIS database {Table 10.3}. The edge between the newly digitized stands and original stands is seamless, implying that there is no gap between the polygons of the two GIS databases. and no overlap if the two sets of poly-
D D
Smnd
2
Dani~ 1 Pick~[( for~$[
Vegetation Basal type area"
Hectares
Acus
17.24
42.6
A
15.t4
37.4
8
Age
MBP
190
55
21.3
15
7
0.8
• squar~ f«t per acr~ h
thousand board fttt per
ac r~
gons were placed together. By simply copying the landscape features from the land purchase GIS database into the stands GIS darabase. it is possible to bring the newly digitized land purchase polygons into the stands GIS database, however. the attributes of the new stands may not be present. depending on the GIS software program being used (Figure 10.4) . The new fo rest stand polygons would then need to be attributed a second dme, after they have been pasted intO (he original stands GIS database. To avoid duplication of effort in update processes, three options are clear: (l) digitize the new landscape features directly into the original stands GIS database. (2) use a merge process to com bine the newly digitized stands with the origi nal stands GIS database. or (3) if available in
D
Forest stands
Original stands
Stand
VegType
Basal Area
Age
MBF
Stands in land purchase area
1
A
200
50
21.2
2
C
175
40
12.9
30
c c
190
45
17.3
110
25
4.1
o
o
o
o
o o
o
31
Figure 10.3
Attributes of stands in a 32.38 hectare (80 acre) land purchase adjacent to the Daniel Pickett forest
TABLE 10.3
GIS databases can be updated with new landscape features (points. lines. or polygons) by either adding the new landscape features (Q an existing GIS database, or by editing the existing landscape features. or boch . Two examples are provided below to illuStrate updating a GIS database by adding new landscape features . The firSt example involves a land purchase and subsequenr addition of twO forest stands co a stands GIS database. The second example involves the addition of new trails to a trails GIS database. In each case, assume that the new landscape features were either digitized or collected with a GPS SYStem and are available in a GIS format. Prior to [he iniciarion of the update process, you should assume that the new data are comained in GIS databases that are separate from the GIS databases that need updating. Refer back to chapter 3 for a review of methods and tools for development of a new GIS database.
stands and land purchase area.
161
0
Figur~ 10.4 Daniel Pick~n forest stands and land pu rchas~ a r~a aft~r copying and pasting landscape fntures from th~ land purch3# GIS datah3# to th~ stands GIS database.
172
162
Part 2 Applying GIS to Natural Resource Management
rhe GIS software program being used, use an 'updare' funcrion . ArcMap and ArcView 3.x, for example, both have (he ability to use an update function made available
rhrough the XTools extension (Data Easr, 2007; Oregon Department of Foresrry, 2003). When using a merge process or an update funC[ion, the stands GIS database
will be updared with rhe new polygon dara contained in the land purchase GIS darabase. If rhe land purchase GIS database includes fields named and formarred exactly as
purchase polygons beyond che excent needed in the new land purchase GIS database, creating an area of overlap
with the polygons in [he "ands GIS database (Figure 10.6). Then, erase from [he land purchase GIS da[abase the area of overlap with the stands GIS database, creating
a second (new) land purchase GIS database. In [his new land purchase GIS database, [he edges of the new polygons seamlessly match the edges of [he associated polygons in the stands GIS database (Figu re 10.7).
those in the forest stands GIS database, the 3ttrihme data
wi[hin [he land purchase GIS da[abase will be moved (along wi[h rhe associa red purchased polygons) to rhe updared srands GIS darabase (Figure 10.5). The assumprion was made rhac [he polygons in che land purchase GIS darabase seamlessly marched [he edges of polygons in rhe Daniel Picke" foresr scands GIS da[abase. How is this possible? Matching rhe spacial juxtaposition of the new landscape features
to
the landscape fea-
cures in [he GIS da[abase being updaced (the stands GIS da[abase) can be accomplished using one of at leasc cwo mechods, depending on what rype of process is available within the GIS software program being used: (I) copy [he new po lygons into the original stands GIS database and use snapping tools [0 properly match rhe new polygons
wi[h the original scands polygons, or (2) use a process as described in the fi rst cwo sceps of processes A and B of Figure 10.1. Here, you might firsc digi[ize the new land
D
Updating a trails GIS database The existing trails system for the Brown Tract was digi-
tized several years ago using hard-copy maps provided by the forest recreation planner (Figure 10.9). While suitab le for recreation planning and the development of recreation maps to guide visitors around the area, the trail
syscem described in the [rails GIS database co uld very well be considered out-of-da[e. The [rail system, like ocher fea[Ures of a landscape, evolves as the managers of the forest
develop new trails, or as people find different hiking or mountain biking rouces through the landscape. The latter Overlap
D
Or~inal stands
Stands in land acquisition area
Timber stands
Figure 10.6 Overlap of new land purchase polygons with a GIS database that will be updated.
D
Stand
VegType
Basal Area
Age
MBF
1 2
A
200
50
C
175
40
21.2 12.9
31
c c
45 25
o o
A
190 110 190
B
15
7
30
Figu~
10.5
55
o
Original stands Stands in land acquisition area
17.3 4.1 21.3 0.8
Daniel Pickett forest stands and land purchase ana after
updlUing the stands GIS database wing the land purchase GIS database.
Figure 10.7 Land purchase GIS dat:lbasc after erasi ng the overlap with the stands GIS database.
173
Chapter 10 Updating GIS Databases
The term digitizing, as described in chapter I. means to cooven a hand-drawn (or other rype 00 map CO a digical image of a map. Normally. digitizing is performed using a digidzing table and a digitizing puck. A map is laid on the table, taped down to ensure that it doesn'[ move dur-
ing [he digitizing process. and at least four control points on the map. for which on-the-ground coordinates are known, are emceed imo the compurcr system using the digitizing puck (similar co a computer mouse) . The puck is and arc 11 was split into two pieces based on its intersectio n w ith arc 2 4, To extend the idenri ry process exam ple (() the D an iel Picken fores t, aga in examine the case of the stands GIS
database and the fire GIS database. Suppose the intent was [Q
develop a GIS database that contained the entire stands
data (geographic an d tabular). but with the fire bo undaries being integrated into the stands database. Using an ident ity process you ca n see that some stand pol ygons
defined by the boundary of o ne of the input GIS data-
have been split along the fire po lygon boundary (Figure
bases, not by the boundary of the ove rlap between the two GIS databases . Figu re 11 .6 illustrates that the resulr-
11.7), thu s the fire GIS database has an influence on the structure of the resultin g polygons. In addi tio n, 3nribure 185
Chapter 11 Overlay Processes GIS database
Spatial features
Tabular attributes
175
o
Stand bouooaries
D
Firearea
Input GIS databases attributes: basal area, volume per acre, vegetatiOil type, age
Oatabase to perform the identity process Oil : stands
Oatabase used 10 perform the identity process: fire
o
------
attributes: day, month, year
------ -----Output GIS database
Resulting database: stands, defined along original stand boundaries and along the booodaries of the fire
~
attributes: basal area, volume per acre, vegetation type, age, day, month,
,ear
Figure 11.7 Performing an identity proass on the stands GIS database using the fire GIS database.
fields are present in the resulting GIS database [0 represenr those that were presenc in the fire GIS database. However,
only the polygons within the fire boundary actually contain data related [0 the fire. The amibute fields from the fire GIS database related [0 polygons outside the fire area are empry and contain '0' values (Figure 11 .8). A key concept when performing the identiry process is obviously determining the spatial extent of the [wo GIS databases that is [0 be retained. In the above example, the fire GIS database was overlaid on the stands GIS database, and the incem was [0 rerain the spacial extent of the stands GIS database. If you were to reverse the order and overlay
m e stands GIS database onto the fire GIS database, the resulting GIS database would have quite a different look [0 it (Figure 11 .9), as the spatial extent of the resulting GIS database is defined by the spatial extent of the fire GIS database. Here, only the stand boundaries within the fire remain. While the stand-level data anributes associated with the stands in [he fire area are presenr in the resulting GIS database, no stand-level data attributes are available
for the polygons outside of the area represented by the original stands GIS database (Figure 1l.l0).
Stand V StandY StandW
Stand
VegType
Basal Area
Age
MBF
Month
Day
Year
T
C
120
5S
19.5
7
2
2002
U
A
260
70
37.7
7
2
2002
V
A
260
70
31.7
0
0
0
W
C
190
45
17.3
0
0
0
X
B
20
10
1.8
7
2
2002
y
B
20
10
1.8
0
0
0
Figure 11.8 A more detailed examination of the results of the identity process of the fire GIS databue overlaid on the stands G IS database.
Union Processes In a union process, the intent is to overlay one GIS data-
base on top of another GIS database, and re tain all of the spatial within [wo or Figure
boundaries of the landscape features contained both GIS databases, A 'union' is the act of joining more features into one (Merriam-Webster, 2007). 11.11 illustrates that when using a union process.
the resulting GIS database has a geographic extent representing the same area as both database # I and database
#2, yet the arcs that defined the polygon in database # I are present where the polygon from database #1 overlapped the polygons in database #2. The resulting GIS database now includes 5 polygons. 13 arcs, and 9 nodes, as arcs 2... 26 • and 27 were split into two pieces (11 and b) 186
176
Part 2 Applying GIS to Natural Resource Management Tabular attributes
Spatial features
GIS database
~ ~ ,, ,,
Input GIS
Input GIS databases
Database " Database to perform the identity process on: fire
attributes: day, month, year
o
Input GIS
Database .2 2,
,
,
: 21
,:, 22 ,, ,,
,, ,
,, ,
2,
... 27
... 26
8:. ---
Database used to
attributes:
perform the identity p(ocess: stands
basal area, volume
-------------
per acre, vegetation type, age
Input GIS
Database"
2,
In# 1 and #2
_____ J____ _
~
In ", outside 01112
Input GIS
output GIS database
+---t-
Database '2
Resulting
database:
" "
attributes: basal area, volume per acre, vegetation type, age, day, month, year
fire, defined along original fire boundaries and along the boundaries of the stands
./
,,'
;--- .... : I
In '2, outside
oil'
\~f--+-
In 111 aOO'2
2, Output GIS
Database Figure 11.9 Perfo rming an ide ntity process on the fire G IS database using the n ands GIS database.
/'
o
Fire area outside of Daniel Pickett forest
D
Fire area inside 01 Daniel Pickett forest
Stand M
\
Stand N
Stand
VegType
Basal Area
Age
MBF
Month
Oay
Vear
L
C
120
30
5.6
7
2
2002
0
7
2
2002
C
0 190
0
N
45
17.3
2
2002
U
A
260
37.7
2
X
B
20
70 10
7 7
1.8
7
2
2002 2002
M
Figure 11.11 An example of the processing of landscape fe atures during an union process.
Figure 11 . 10 A more detailed examination of th(' results of the identity process of the standt G IS database overlaid on the fi re G IS database.
based o n the ir intersection with arc 11. and arc II was split into twO pieces based on its imersection with arc 2 4, To ill ustrate a union process with a more realistic natural reso urce management problem. suppose a union process were to be performed on the Daniel Picket[ forest fire and stands GIS darabases. The res ul ting GIS database has the combined geograp hic extent of rhe (wo GIS databases, and co ntain s similar landscape fea tu res as we re fo und in rhe origi nal fire and stands GIS darabases (Figure I 1.12). This iliumares o ne adva ntage of using a union process: rhe sparial delinearion of rhe polygons in rhe resul ring GIS database is a fu nction of both of the origi nal GIS databases. thus polygon boundaries fro m both original GIS databases are reta ined. However, it also suggests a disadvantage of the process: rhe resulting GIS data base may comain landscape features outside of [he boundary
187
Chapter 11 Overlay Processes GIS database
Spatial features
1n
Tabular attributes
Input GIS databases
Database to perform the union process on: fire
attributes: day, month, year
Database used to perform the union process : stands
attributes: basal area, volume per acre, vegetation type, age
o o
o
Output GIS database
Resulting database: fire and stands, defined along original fire boundaries and along the boundaries of the stands
attributes : basal area, volume per acre, vegetation type, age, day, month, year
Stand attributes, no fire attributes Stand attributes, fire attributes No stand attributes, fire attributes
Figure 11.13 Illustration of completeness of a tabular database after a union process of the fire and stands GIS database.
ment analysis. In locating [he su itab le areas, you decide that the criteria should include locating a certain type of
soil (loamy soils), on a ce rtain slope condition (flat
sented in ,he OUtput G IS database. The attribute fields contained in both of the original GIS databases may also
slopes), where the area is currently wned for agricultural use, and where there are few limi tadons for using the land to grow agricultural produces. In this assessment, there are four distinct anribures about the land: so il type, slope condidon, zoning code, and land classification . These four attributes are comain ed within four different GIS databases associaeed with the Pheasant Hill planning area, and are delineated using polygons that do noc necessary
be present in the resulting GIS database, however some
coincide (spatially) from one G IS database to the next.
data cells will likely be empty in the attribute table (Figure
One way {Q accomp li sh this overlay analysis is {Q use an iterative union process to bri ng all fou r databases together so chat all of the anribures are available in a single, com-
Figure: 11.12 Performing a union process using the fire GIS database and the stands GIS database.
of interest, and thus perhaps may include unnecessary
landscape features. Although some of the fire polygon lies outside the Daniel Pickett forest, this area will be repre-
11.13). The union process is useful for those situations in
which you want to preserve all of the spatial and non-spatial data that is present in twO input GIS databases. More complex analyses can be performed using the intersect, identity, or union processes than simply bringing together the characteristics of twO GIS databases. For example, suppose you were interested in locating areas suitable for a certain type of agricultural praccice in the
Pheasant Hill planning area of the Qu'Appelle River Valley in central Saskatchewan. While the databases we will use in th is example are dated (\980), they are rich with information and they allow us to examine (he usefulness of the union process for a natural resource manage-
bined GIS database. Initially, two of the GIS databases would be unioned, then a third would be unioned to the union result of the first two. Finally, the fourth database would be unioned to the union result of the first three GIS databases. Using a query thac involved the criteria listed below, the areas suitable for the praccice you had
in mind could be identified (Figure 11.14), since the union of the four GIS databases would contain the attrib-
utes of al l of the original GIS databases, and since the polygons would be split along the boundaries of the original polygons.
188
178
Part 2 Applying GIS to Natural Resource Management
.
- ..-
-= ,pg."
.IiOi'_
_
Areas suitable for an agricultural practice
c=J
Other areas that do not meet the criteria for an agricultural practice
Figwe 11.1 4 The result of a query on the union of soils. topography, land classification, and zoning G IS databases developed for the Pheasant Hill planning area of the Qu'Appclle River Valley. Saskatchewan (1980).
Query criteria
Original database
Soil_type ", Canora Loam or Soil_type ", Indian Head Clay Loam or SoiLrype ,. Oxbow C lay Loam or Soil_ rype ,. Oxbow Loam or Soil_type" Rocanvillc C lay Loam or So il_type II: Whiresand Gravelly Loam
soils soils soils soils soi ls
Topography
topography
=
FLAT
soils
Class" No Significant Limitations or Class", Moderate Limitations
land classification (eLi)
Zoning = Agriculture Priority I
zoning
As an example. if you use a GIS database containi ng lines as an input GIS database. and then intersect it wi th a GIS database containing polygons (Figure 11. 15), the resulting GIS database wi ll be composed of line features _ The line features will be split at all intersections with the boundaries of the polygons in the pol ygon GIS database,
land classification (eLI )
Incorporating Point and Line GIS Databases into an Overlay Analysis Although the overlay examples thus fa r have focused on the analysis and manipulation of polygon GIS databases, it is also possible to inco rporate other types of features (points and lines) into overlay processes. For example. point and line GIS databases can be used in association with polygon databases when performing the intersect and idemity processes. The union process. however, requires that all GIS databases of interest be composed of polygon features. When using the imersect and identity overlays. the input GIS database can be composed of poin ts, li nes, or polygo ns but the overlay GIS database must be composed of polygons with one exception. With in some GIS software. the intersection of two line databases is possible. with the result being a new point database that has ca ptured all intersection locations. When point or line databases are involved in an overlay process with a polygon. the resulting Output GIS database will be of the same feature type as the first input GIS database (point or line).
,
,,, ,, ,,
Input GIS Database '1
.
Input GIS Database In
,
,,
2,
, ,,
,,, ,,
2,
,, , ,,, ,, ,
• • • • 2,
2,
2,
2,
2.
- - -- - -
- -- -
--- -
1, Overlay of Database #2 on Database.1
2,
2, 2,
1, 1,
2,
I,
2, 2.
2,
--- ---- -- - -
- -- -
Output GIS Database
1,
Figure 11.15 An example of the manipulation oflandsc.1pe features during an intersect overlay of line and polygon databases.
189
Chapter 11 Overlay Processes
179
and only lines that full within the extent of the polygons will be retained in the outpUt GIS database. While this represents a process similar
[0
the clip process, the resul t-
ing lines contain the information (attrib utes) of the polygons within which they fell. In the example in Figu re 11.15, a line (I) that represents a road (perhaps) is being overlaid by the two polygons from previous examples in
.
habicar. Somewhere along che 0.0-1.0 range, biologisrs will need to determine the threshold levels th at separate good habitat from poor habitat. One process that can be used to arrive at these scores is presented in Figure 12.10 .
Here, che roads are buffered chree cimes (15.24, 30.48, and 45 .72 m). Two of rhe buffer GIS databases are chen subjected
to
an erase process. resulting in a buffer band
around each road (a 15.25-30.48 mecer band and a 30.49-45.72 mecer band). The 0--15.24 mecer buffer is 1.2
then combined with these MO buffer bands (Q create a GIS database that represents three of the buffer distances
1.0
by polygons, wich no overlapping polygons presem. The
0.8
Distance from road score
0.6
I
D••
0.2
I
I
nor included in chese chree buffer discances. The buffers are then overlaid on the vegetacion GIS database. breaking
vegecacion polygons ac che buffer boundaries. The basal area, stand age. and distance from road scores can then be
calculaced in che cabular ponion of che resulcing GIS dacabase. The final HSI score can be calculaced as a funcrion
0.0 0
fourth buffer distance, as you can imagine. is everything
~
N
Distance from road (feet)
Figure 12.8 Distance from roads scores for a range of distances from the road network.
of the basal area, stand age, and distance from road scores. and a thematic map can be developed to illustrate the distribution of vole habitat across the landscape. In addition ,
che final GIS dacabase can be queried co develop a cable of area by habicar class. 203
Chapter 12 Synthesis of Techniques Applied to Advanced Topics
Iv~m~ GIS da!abase
I L
GIS R_ / database
I Roa~ I I I GIS
da!abase
Roads GIS database
i
~
~
~
Calculate basal area
Buller roads
Buffer
score
Buffer roads 15.24m
30.48 m
45.72 m
~
~
~
~
I(
Calculate stand age
BUllere,,';
(o-~~m)
score
193
r_
I( II I( III Buffered
(o-~~ m)
roads Buffered
(0-45.72 ml
~ Erase process Erase process
~
I I Bullered roads
I I ~
(3Q.48-
45.7Srn)
Buffered r_ (15.24-
30.48 m)
Combine process
~ Overlay process
I
/
~ HSI GIS database
IL II
/
Road buffers
I
Calculate
road score
calculate HSI
score
Develop mapaf HSI scores
Figure 12.10 Hierarchy of intermediate and final G IS d:ltabue$ created in the devdopmcnt of an analysis of potential w ildlife habitat suitability (HSI) areas for a vole on the Brown Tract.
Summary This chapter illustrates juSt a few of [he more com plex
spatial analyses that may be performed (or requested) by natural resource managers. The number and arrangemem
of GIS processes could vary in addressing analyses such as
these, and may include buffering. clipping, erasing, and
need fo r quantirative rules and a log ical set of GIS processes [0 separate one set of landscape featu res from another in an analysis of ROS classes is important because a single un it of land must be assigned only o ne ROS class,
querying of landscape features. Therefore. the chapter
and all units ofland must be assigned a class. The graphical display of the result of a complex GIS analysis. such as
represents a synthesis of the tools readers have acquired
the ones illustrated in this chapter. is also important
from previous chapters in this book. It should be appar-
because land managers rypical ly use these products to
ent by now that it is important to explicitly define the quantitative rules and the GIS processes th at might be used to address complex spatial analyses. For example, the
help them visualize and make decisions regarding the management of natural resources. 204
194
Part 2 Applying GIS to Natural Resource Management
Applications 12. 1 Land classificat ion. Becky Blaylock, manager of the Brown Tract, wants you to develop a managementrelated land classification for the forest. She asks you to develop GIS databases for each of three classes using the following rules: 1. Special stewardship areas will consist of the following landscape features: Oak woodlands, Meadows, and Rock pits. 2. Focused stewardship areas will consist of the following landscape feamces: a. streams buffered according co the Oregon State Forest Practices Act (30.48 meters [100 feet] around large fish-bearing st reams [Size = 'Large' and Fishbearing = 'Yes], 21.34 meters [70 feet] arou nd medium fish-bearing streams, 15.24 meters [50 feet] aro und small fi sh-bea rin g streams, 2 1.34 meters [70 feet] around large non fish-bearing streams, 15.24 meters [50 feet] around medium non fish -bearing streams, and 6.10 meters [20 feet] around small non fishbea ring streams); b. a buffer of 100 meters around all water sources tha t are not culvert spills or water towers; c. a buffer of 100 meters around all authorized (rails; and d. research areas. 3. General stewardship areas will consist of whatever land remains. Do the following: a) Develop and illustrate a process (How chart) for accomplishing the task of defining the land classificat ions of the Brown Tract forest according to the rules listed above. b) D ete rmine how much land area is contained in the special stewardship land classification . c) Determine how much land area is contained in the focused stewardship land classification. d) Determine how much land area is contained in the general stewardship land classification . e) Produce a map of the entire Brown Tract, ill ustrating the three land classifications.
As a general strategy, yo u may want to follow [his process: • Develop a special stewardship GIS database. • Develop a focused stewa rdship GIS database by performin g the appropriate buffer and query
processes, and then intersecting these GIS databases. (Why would you nOt use a merge process here?) • Erase the special stewardship GIS database features from the focused stewardship GIS database features. • Erase both the special stewardship GIS database features and [he focused stewardship GIS database features from the stands GIS da tabase, creating the general stewardship GIS database. 12.2 Recreat ion O p portunity Sp ectru m. The Dimict Manager associated with the Brown Tracr (Becky Blaylock) would like you to determine how much area might be classified in the five recreation opportunity spectrum (ROS) classes (see T able 12.2) . Based on this subset of rhe ROS criteria, a) How much land area is contained in the primitive class? b) How much lan d area is contained in the semiprimitive, non-mororized class? c) How mu ch land area is contained in the semiprimitive, motorized class? d) How much land area is co ntained in the roaded natu ral class? e) How mu ch land area is contained in the roaded managed class? f) Develop a thematic map illustrating the five ROS classes on the Brown Tract. g) Draw a flow chart [Q describe the p rocesses lIsed [Q develop the ROS classes, including the GIS operations and all GIS databases used (original, intermediate, a nd final GIS databases). 12.3. Visual quality bu ffers. You have been asked by the manager of the Daniel Picken forest [Q evaluate the potential impact of two proposed organizational policies for the forest resou rces found there. It seems that the owners of the property are becomin g very concerned with the public perception of management on the forest, thus they are interested in the trade-off's assoc iated with alternative management pol icies. Policy #1: Buffers next to neighboring landow ners. Assume for this example that even-aged forest management is practiced across the property. This potential pol icy suggests that c1earcllt harvesting activities adjacent to neighboring landowners of the Daniel Pickett forest will be restricted. 205
Chapter 12 Synthesis ofTechniques Applied to Advanced Topics
a} If a 50-meter uncut buffer were to be left adjacent co all other property owners, how much land would this require as a volumary contribution [0 rhe pCQ[ecrion of adjacent landowners resources?
b} How much timber volume of vegetation class A would be found in the buffer (vegetation class A is the older timber class. perhaps that which can be harvested in the near- term). and what perce mage of rhe [Ocal volume in this vegetation class would
be affected? Policy #2: Buffers next to paved public roads. This potential policy suggeSts that visual quality buffers may be maintained along paved roads within rhe Daniel Pickett forest. These buffers will not be managed. bur rather treated as reserved areas, where har-
vesting is precluded. a} If a 50-meter buffer were required around all paved roads, how much land area would this involve, and
how much timber vo lume in vegetarion class A
would it affect? b} If the State decided to convert the North-South paved road on the Daniel Pickett foreSt to a highway, and required a 1DO-meter wide corridor [Q he transferred to State ownership, how much land area would be affected?
c} If bare land values were assumed to be $200 per hecta re. and timber volumes $400 per thousand
195
board feet (MBF). how much would you ask the Scare
[Q
compensate [he owners of the Daniel
Pickert forest for the loss of this land? d} If a 30-meter visual quality (i .e.• uncurl buffer was then proposed around the I OO-meter highway corridor, what is the [mal effect [Q the forest resource base, in terms of land area now affected in each vegetation type?
12.4. Habitat suitability index fo r a vole. The biologist associated w ith the Daniel Picken forest, Will Edwards. has recently become aware of a vole habi tat su itability model, and is interested in understanding the extenr of
vole habitat on the forest. Will asks you to apply the model described in the 'Habitat Suitability Model with a Road Edge Effect' section of this chapter to the Daniel Pickett forest, and m: a) Calculate the amount of land area on the Daniel Pickett forest in the following habitat suitabil ity
classes: 0.000-0 .200 (low quality). 0.201 -0.400 (low/moderate quality). 0.401-0.600 (mode rate qual ity). 0.601-0.800 (moderate/high quality). 0.801-\.000 (high quality). b} Develop a map illustrating the habitat quality for the vole by suitability class. c) Draw a flow chart of rhe process yo u used to
develop the habitat suitability classes.
References American Farmland Trust. (2006). Land classification
sysum. Washington. DC: American Farmland Trust. Retrieved February 17. 2007. from http://www. farmland.org/resources/furureisnowllanddassification system.asp. American Forest & Paper Association . (2002) . Sustainable Fomtry Initiative (SFI)"'. Washington. DC: Amer ican Forest & Paper Association. Retrieved
December 10.2007. from http: //www.afandpa.org/ ContentfNavigation Menu/ Environment and Recycling!
SFIISFl.htm. Brooks. R.P. (l997). Improving habitat suitabil ity index models. Wildlifi Society Bulletin. 25. 163-7. Butler. R.W .• & Waldbrook. L.A. (l991). A new planning tool: The tou rism opportunity spectrum. JournaL of Tourism Studies. 2(1). 2-14.
Canadian Forest Service. (2007). Ecological land classifications. Onawa. ON: Canadian Forest Service, Natural Resources Canada. Retrieved February 17. 2007. from hup:11ecosys.d1.scf. rnea n.gc.cal dassifl i n rro_strat_e. asp. Clark. R.N .• & Stankey. G.H . (I 979}. The recreation opportunity spectrum: A framework for planning, management, and research. GeneraL Technical Report PNW98. Portland . OR: Pacifi c Northwest Forest and Range Experiment Sta tion, USDA Forest Service.
Frayer. W.E .• Davis. L.S .. & Risser. P.G. (I 978}. Uses of land classification. Journal ofForestry. 76. 647-9. Klingebiel. A.A .. & Montgomety. P.H . (I 973}. Land capability classification. USDA agricultural handbook 210. Washington. DC: US Government Printing
Office. Retrieved February 17. 2007. from http:// 206
196
Part 2 Applying Applying GIS to Natural Resource M Management anagement
soils. soils.usda.gov/tech usda.govltochnicaUhandbookicontents/part622p2. nical/handbookicontents/partG22p2. html#ex2. Morrison Marcot, B.G B.G..•, & Mannan, Morrison., M.L. ., Marcot. Mannan. R.W. 92). Wildlife-habitat relationships: Conupts and ((19 1992). applications. Madison. Madison, WI : University Universiry of Wisco nsin Press. Natural Natu ral Resources Canada. (2000). OverviewofclassificaOverview o[classificadeurmining land capability for tion methodology for determining forestry foustry.. Onawa, Ottawa, ON: GeoGrads GeoG racis Cli C li enc ent Services. Services, Nacural Narurai Resources Reso urces Canada. Rerrieved Retri eved Augus[ Au gust 10, 2007, from http: //geogratis.cgdi.gc.ca/C LI /frames. 2007. hrtp:lIgeogratis.cgdi.gc.ca/CLl html. htm!' Depareme", of Forestry. Fo rest ry. (2007). Oregon adminOregon Department iJtrah'vt! ruiLs, nlus, department offorestry. istrative o/forestry, Division 35. manngeuullt agement of o[ statt state forest lands. Salem, Salem , OR: Oregon Depart ment of Fore5[ry. Depanme", Forestry. Retrieved March 12. 12, 2007, 2007. from http://a rcweb.sos.sta te.or.us/ te.or.us/rules/OARS_6001 rules/OARS_GOO/ h((p:llarcweb.sos.sta OAR_G29/629_035.html. OAR_629/629_035.html.
Rempel. R.S R.S.,.. & Kaufmann Kaufmann., C.K. G.K. (2003) (2003).. Spatial modelmoddRempel, ing of harvest co nsrra nsuainrs inrs on wood supp supply ly ve rsus wildlife wildli fe habi habicat tat objectives. objecrives. £Ilvironmeruo Environmental/ Management, 32. 32, 646-59. ment. US Bureau of Reclamatio n. (195 1). Land classification. Bureau of reclamation man manual ull~ vol. V, irrigated land use, Denver, CO: US Bureau of Reclamation. Reclamati on. use. part 2. Denver. forest, USDA Forest Service. (2006). Hiawatha national form. 2006 forest plan. Milwaukee, WI: USDA Forest fomt Milwaukee. Service, Eastern Region. Retrieved February 18.2007. 18,2007, Service. from http://www.fs.fed.us/r9/hiawarhalrevision/2006/ from http://www.fs.fed.us/r9/hiawatha/revision/2006/ ForPlan .pdf. Washington State Scate Parks and Recrearion Rec reatio n Com Commission. mission. (2006) (2006).. WAC 352-16-020 land classification. Olympia. Olympia,
Recrea(ion ComWA: Washington Srate State Parks and Recreation Reuieved February mission. Retrieved Feb rua.ry 17. 17,2007, 2007. from http://www. htrp:/iwww. parks.wa.gov/ plansllowerhoodcanaI/State%20Pa rks% parks.wa.gov/plans/lowerhoodcanaI/Srare%20Parks% 20Land%20C 20 Land%20Classifications. lassificat ions.pdf. pdf.
207
Chapter 13
Raster GIS Database Analysis Objectives
the next chapter to other raster database applications. the primary raster GIS database that is considered in th is
The skills and techniques you'll learn in this chapter
chapter is a digital elevation model (OEM) . Many difTer-
should provide insight inco the examination and applica-
em types of landscape information can be cultivated from
don of raster GIS databases for natural resou rces research,
a single OEM database.
and how raster GIS databases might be included in suppaning natural resource management decision-making.
Digital Elevation Models (OEMs)
At the conclusion of this chapter, you should have an understanding of:
As their name implies. OEMs comain information related
1. how landscape comour GIS databases are created from a OEM; 2. how landscape shaded relief GIS databases are created
to the elevation of a landscape above sea level or relative to some other datum point. T hey are different than the typical USGS Quadrangle maps discussed in chapter 4, in that they are in digital form. As with other raster GIS databases,
from a OEM; 3. how slope GIS databases are created from a OEM; 4. how to calculate slope gradients for a linear landscape feature, such as a road, trail, or stream;
5. how
(0
conduct a viewshed analysis for a parrion of a
landscape; and 6. how to create a watershed boundary based on digital elevation data.
each unit on the landscape is typically represented by a landscape-related value (or set to a null or 'no data' value), and each unit is exactly the same s ize and shape as the
other units (F igure 13. 1). The most prevalent OEM databases available in the US are the USGS 30 meter OEMs (US Department of Interior, US Geological Survey, 2007). Within Canada, Natural Resources Canada (2007) provides access to digital topographic data. Raster databases are onen described in terms of their spatial resolu-
As mentioned in chapters 1 and 2, the re are {Voto general types of data structures used in GIS coday: vector and raster. Unci l now, we have focused on vec[Qr GIS data-
bases and the GIS operations related to the typical kind of applications performed in natural resource organizat ion field offices. This chapter now delves into rhe use of raster GIS databases for namral resource applications. and a few of the GIS operations that can be performed using them. An emphasis is placed on how raster GIS databases mi ght be used in field offices to support natural resource management decisions . Although we will rurn our attemion in
tion, as in the phrase '30 m OEM'. This infers that each grid cell in the OEM database is 30 m by 30 m in size in terms of on-the-ground area that it represenrs . Many
regions in the US also have 10m OEMs available for areas within federal and state agency administrative boundaries. In some cases OEMs for states. provinces. or other large regions can be purchased from commercial entities. DEMs can be used for a variety of analytical purposes, bur the most general of these purposes is simp ly ro view
the rel ief of a landscape. OEMs can use shades of color or gray rones co illustrate differences in e1evadon through a 208
198
Part 2 Applying GIS to Natural Resource Management One notable category rep resented in the lege nd in
Figure 13.2 is the 'No Data' category. This category is necessary because raster GIS data must be stored as a set of grid cells that combine ro form a rectangular or square shape-
the width and heigh t of the image is defined by the number of grid cells. Therefore, when landscape fearures of interest do not match a rectangular shape (e.g., the shape of the Brown Tract), the grid cells that are nOt associated with [he landscape features of interest are given a nuil, or
No Data, value. For example, all of the grid cells that represent areas outside the boundary of the Brown T ract contain no data. For mapping purposes, the symbo lization
used to display cells with a null value can be assigned a transparent shade. Almost all raster GIS software programs allow the recognition of a null value, yet are designed to ignore this va lue in analytical computations. When performing a multiple GIS database overlay ana lysis, some raster GIS software programs are also designed ro ignore Figure 13. 1 Ruler grid cells from a digital elevation mode (OEM).
separation and c1assificarion of elevation values . Figure
13.2 illuscrates a gray tone color-shading scheme app lied a 10 m OEM of the Brown Tract, and uses a twelve-
to
category equal-inrerval classification scheme CO highl ight (he cha nges in elevation. The equal-imerval classificat ion takes the disrriburion of elevadon data values found
within the 10m OEM and divides it equally into twelve sub-sets of elevat ion ranges. Most GIS software will allow users to define [he number of elevation range categories shown visually. and offer choices for color and gray tone schemes [0 illusrrate distinctions between elevation C3rcgones. Brown Tract OEM Values (feet)
D
136-150
0
151 - 200
D
201-25O
D
251 - 300
~ 30 1- J50
IH;11 351 -"OO
111 401_ 450 _ _
451 -500 501- 550
_ 5 5 1-600 _ 8 0 1-850 _651-700 _ No Data
Figure 13.2
OEM.
Elevation ca t~ori ~ for the Brown Tract using a 10 m
any cells that overlap null cells in any other GIS database being analyzed. For example, if twO spatially coincident raster databases were overlaid on each orher and any portion of either dacabase comained cells with no darn, any database resuhing from an overlay or compara tive analys is that involved borh databases would also comain no data cells at rhe same locations. This result would occur even if actual values were present in a portion of one of the raster databases. Whi le th is functiona lity may or may not be appropriate for particular G IS analyses, users should be
cognizant of how null values assigned to grid cells will be handled within their selected GIS software program{s).
Elevation Contours OEMs ca n be used to create devation contours, or lines that indicate a constant or nearly constam elevation across a landscape . Contour lines are created adjacent ro each other such that elevation represented in even in crements, such as every 30, 50, or 100 m. An elevation contour GIS database is usually represemed through a vector dara srruCture based on a user-defined elevation interva l. Contour lines allow YOll to examine the relarive relief of a landscape and to make inferences about landscape ropography ro support managemem decisions. For example. co ntour intervals can be used ro delineate likely hydrologic drainage patterns and watershed boundaries. When
designing road systems, engineers rypically need to keep the slope of each road below some maximum gradient. since slopes too steep will either prevem the movement of
certain rypes of veh icles (if the vehicle is travelling uphill), 209
Chapter 13 Raster GIS Database Analysis
A contour. [Q most people. represents the ou dine of some figure or body. Contour intervals. as used on maps. represent the oudine of all areas thar have the same elevation. It would be as if you were to slice (horizontally) the landscape every 100 feet (or whatever interval was chosen) in verrical elevation. Contour plowing is a common practice in agriculture. where
or will be too dangerous to travel (if the vehicle is travelling downhill) . With many raster GIS software programs, users have the ability [Q choose a contour interval and the starting elevation value at which contours will be created. To creare contour lines a process such as rhar described in Figure 13.3 can be used. In cases where e1e-
Select contour interval
Setect base elevation value to start contours
199
plow lines are laid parallel to the contour of the landscape [Q reduce the erosion potential of rhe agricultural practice. This also reduces strain on farm machinery; this practice encourages moving laterally rather than perpendicularly through elevation gradients. Each section of a plow line is. theoretically. at about the same elevation as every other section of the line.
vatlon umts (i.e. meters. feet) do not match horizontal mapping units. a unit conversion factor may be needed in order to bring the units into agreement. Within ArcGIS a OEM must be opened as a layer, the Spatial Analyst Extension must be activated. and the Spatial Analyst menu must be opened. From the Spatial Analyst menu. choose Surface Analysis. then choose the Contour option. This should open the Contour dialog box, which will prompt you for the Input Surface. conrour dimensions, and an output database name. Using the Brown Tracr 10m DEM, a COntour interval of 50 feet with a base contour elevation of 100 feet was chosen (Figure 13.4). Using (he 100-foot base elevation should result in comour lines that o ri ginate from 100 feet and incremem in 50-foot steps. The contour line GIS database that is created is a vector GIS database, and each line contains an att ribute describing the elevation . Users can then modify this vector GIS database to display different color shades or line thickness for differenr contour lines of interest.
Brown Tract ContoIS Interval
- 50'''' OEM Value. (feet)
Verify elevation units
0
o
100 - 150 1S1-200
W
2{)1 -25O
_
2S1-lOO 301 - 350
_
351--400
_4()1 -~
_ . S l -5OO
Figure 13.3 A general process for the development of a contour line GIS database from a OEM.
_
501-S50
_
551-600
_
601- MO
_
6S1 -700
Figure 13.4 A contour line GIS database for the Brown Tract displayed on top of the Brown Tract 10 m OEM.
210
200
Part 2 Applying GIS to Natural Resource Management
Whenever boch ho rizontal coo rdinate positions and ve rtical elevations are processed simultaneously. as is {he case in the creation of contour lines discussed above . it shou ld be ascenained. [hat both cypes of measurements use rhe same units. TypicaUy coordinates and elevations will he recorded using meters, international feet, US survey feet, or some combination of these units. It is not uncommon within rhe US, however, to discover OEMs chat have coordi nate values in meters bur (hat srore elevation val ues in survey feec. A GIS analyst might mistake rhe resulring co ntour lin e va lu es (a represent meters, rhus over-rep rese nting (he elevat ions alo ng co ntou rs. The Spac ial Analyst Contour option dialog box conta ins an input box w here users can specify whe ther elevation units differ from coordinate units within a OEM.
Shaded Relief Maps Another product that can be derived from a OEM is a
Selecl azimuth that represents the sun's location
Select altitude of the sun in the sky
(raslerl Figure: 13.5 A gene:ral process for the: devdopme:nt of a shade:d re:lie:f GIS database: &om a OEM .
shaded relief map. Shaded relief maps are intended ro simulate the su n-lit and shaded areas of a landsca pe when assuming thar the sun is positioned at some location in
tal map unit (Z factor), an ompur cell size. and an output database nam e. Usi ng the Brown T ract 10m OEM, an
rhe sky. La ndscape fearures that face roward the sun will appear more brightly lit than objects facing away. For
azimu rh of 210·, and an altitude of 45·, a shaded rel ief map is created (Figu re 13.6) that shows (rel atively speaking) how much sunl ight reaches each parr of the landsca pe
raster GIS software program s that provide the abiliry co create a shaded reli ef map . the resulr of performing a shaded reli ef map process is a raster GIS database, and
in th e late afternoon (the su n azimurh of 21 0° indicates [hat the sun is located directly co the southwes t of the
each grid cell typically contains an attribute value describing a gray cone rangi ng from light (facing cowards rhe
sun) ro dark (f..ci ng away from the sun). The shaded relief map is useful for illustraring rhe ropography and provides a th ree-dimensional perspect ive of the landscape.
Shaded relief maps can be created with the genera l process described in Figure 13.5. The azimuth selected specifies the d irection from which rhe sun is sh in ing. An azimuth of 90°, for examp le, indicates that the su n is positioned in rhe eaSt, and an azimuth of 1800 ind icates that rhe sun is positi oned in the so uth. The altitude defines the
angle of the sun above the landscape. An alti tude of O· typicall y indicates that the s un is located directly over-
head, whereas an altitude of 90· wou ld indicare thar the sun is at the hori zo n. With in ArcGIS a OEM must be opened as a laye r, the Spatial Analyst Extension must be act ivated, and the Spatial Analyst menu must be o pened. From [he Spatial Analyst menu, choose Surface Anal ys is.
the n choose rhe Hillshade option. This should o pen rhe Hillshade dialog box, which will prompt you for th e Input Surface. Azimuth, and Altitude amOunts. Options are also provided for vertical unit conversion to horizon-
FigllR 13.6 Shade:d relief map of the: Brown Tract using a 10m OEM, an illumination azimuth of 210°, and an illumination altitude of 45°.
211
Chapter 13 Raster GIS Database Analysis
The oriemacion and presentation of 'direction ' has not been discussed co great extent in this book, however, it is important for readers to know the difference between an azimuth and a bearing. Why? Because compasses used in fieldwork either represent di recdon as azimuths or bearings . In some cases, both types of measurements will be represented. Azim uths are degrees of a circle, with No rth being 0° (or 360°), East being 90°, South being 180°, and West being 270°. A compass line indicating an azimuth of 353°, therefore, indicates a d irect ion of almost due North. A bearing is represented as any angle of 90 0 or less from eithet the North or South
landscape, and the 45° altitude indicates a sun position halfway between 'direcdy overhead' and 'setting'). With this shaded relief analys is, you can obtain a sense of the varied ropography of the Brown T ract. Other landscape features. such as study areas, roads, and streams, might [hen be displayed on tOp of the shaded rel ief GIS database to allow an exam ination of how chese resources might be infl uenced by landsca pe tOpography.
Slope Class Maps A third product that can be derived from an analysis of DEMs are GIS databases that represent the slope class, or gradient, of each portion of a landscape. Slope class values are measurements chat indicate me steepness of a landscape, and provide insight into the rate at which other resources, such as water, vehicles. or people. are likely to travel over [hose portions of [he landscape. Since each grid cell in a DEM contains both ho riwntal (e.g., latitude and longitude) and vertical (elevat ion) measurements, the slope of each grid cell can be computed based on the position and he ight of the ne ighboring grid cells. MoSt raster GIS software programs have the abi li ty to compute slope classes. and are able to express slope class as an angle (degrees) or as a percentage of the difference in elevat ion of each grid cell as compared to the neighboring grid cells. lr is important to understand that there are a number of different methods used in choosing the values for slope class calculations. In a raster GIS database. each raster grid cell will have eight neighbors that share a portion (a side or a point) of its boundary: fou r neighboring cells will share
201
(and directed towards the East or West). Thus an azimuth of 353° represents a bearing of N7°W. since the angle would arise from the North half of a compass, and is directed towards the West Similarly, an azimuth of 89° represents a bearing ofN89°E (the angle arises from the North half of the compass and is directed towards the East 89°). and an azimuth of 190° rep resents a bearing of S I OOW (t he angle arises from the South half of the compass and is d irected towards the West 10°). Property deeds, commonly used within North America co legally state ownership of a land area, often use bearings to describe the land boundary locations.
r.
a corner point and four neighbo ring cells will share a side. Theoretically, eight possible grid cell values can be used to calculate the slope class. Many raster GIS software programs lise a formula that takes into account me values of these neighbo ri ng grid cells in calculating the average slope elass of a single grid cell (Burrough & McDonnell, 1998, pp. 190-3). In Figure 13.7, the weighted average slope gradient between the cell of interest (the center cell with a 293 m elevation) and the eight neighbors can be calculated to determine the slope class change by computing the elevation change among the cells. You can imagine. however. that perhaps only the cells that share a side might be used to calculate the slope class for the cell of inte rest, or a broader window can be used. (e.g .. one more ring of cells around the cell of interest, or 24 neighboring cells) . To create a layer representing slopes within ArcGIS a DEM must be available, the Spatial AnalySt Extension muSt be activated, and the Spatial Analyst menu mUSt be opened. From the Spatial Analyst menu, choose Surface Analysis, then choose the Slope option. T his should open the Slope d ialog box, which will prompt you for the Input Surface
1 2 3 302 m 300m 298 m
5 293 m 290 m 295m 4
7 6 8 287 m 288m 290 m
Legend
12~m l
Neighboring cell (3) and elevation (298 m) Cell lor which slope
1293 m1 class will be computed
Figure 13.7 Slope class computation within a raster GIS environment.
212
202
Part 2 Applying GIS to Natural Resource Management
and whether degrees or percent slope is desired. Additional options include vertical unit conversion unit
{Q
Tree height =
horiwnral map
tan (30') - 50 feet
(Z faccoc), Output cell size, and OUtput database name.
or 28.9 feet
G iven [hat slope is a d irect function of distance and elevation comparisons, it is imperative with the slope process rhat users know whether (he measurement units of coordi-
nates and elevations are [he same. If the measurement units ate nO[ the same (e.g. meters for coordinates and feet for ele-
50 feet
vations), the Z factor input can be used co reconcile differ-
ences. T he slope class GIS database created from the Brown Tract 10 m DEM (Figu re 13.8) shows that slopes are repom::d in degrees, and are divided into nine categories.
Angle (degrees) = 30' Angle (percent) = (28.9 feet /50 feet) = 57.7%
The darker-shaded slope class categories represent areas
where slopes are steep, and the lighter-shaded slope class categories represent areas where slopes are gentle. Many natural resource management organizations prefe r CO work with slope classes expressed as a percentage, and thus it may be important for CO know how to perform th is conversio n:
tan (30°) • 100 = 57.7, providing a Quick conversion from degrees to percent slope Figure 13.9 A simple example of the' con'lo'ersion process from degrttS to percent slope.
T o prove th is rather simp le conversion fro m degrees co percent slope, assume that a person was standing on flat
slope class (percent) = tan (a) X 100
ground (Figure 13.9) and needed to determine the pe rcent slope from their location to the cop of a tree. By knowing
where
the angle (30') from their position to the top of the tree, tan = tangent tr igonometric function 0: = slope in degrees
and the distance from their location to the tree, the person
can calculate the height of the tree (28 .9 feet). The slope from the person to the top of the tree, as expressed in percentage terms, is then the rise (the height of the tree, or
28.9 feet) divided by the run (the distance the person is from the tree, or 50 feet), or 57.7 per cent. And, by simply inserting the angle into the equation noted above,
slope class (percent) = tan(30') X 100 yo u can arrive at the same conclusion, D epending on (he GIS sofrware program being used, you may need to conven between degrees and rad ians, since the angle reponed after a slope class calculation may be reported as a radian. An examination of the software documentation will reveal whether this consideration is necessary,
Interaction with Vector GIS Databases
--- ---
Brown Tract slope (degrees) 0
0-2.3
0
2.4-3.8 3.9-5.2
5.3-6.5
9.2-10.8
6.6-7.8
10.9- 13.1
7.9-9.1
13.2- 21.8
Figure 13.8 Brown Tract slope class G IS database' created from aIOmDEM.
There are a number of methods by which you can perfo rm a GIS analysis using both vector and raster GIS databases simultaneously. This ability has traditionally been uncommon in many desktop CIS software programs, bur as technology progresses you will see the expansion of these capab ili ties. and field personnel (those with access
primarily to desktop GIS software programs) will be able to
perform more complex analyses. Two types of analyses 213
Chapter 13 Raster GIS Database Analysis
203
that combine vector and raster GIS databases will now be explored: an examination of the slope class characteristics of land management units, and an examination of the slope class characteristics of streams.
Suppose you were ass igned rhe rask of developing a
Selecl attribute that uniquely identifies
management plan for an area the size of the Brown Tract, and one where there was significant amount of rel ief associated with the landscape. The set of management activities appropriate to each management unit defined on the
stands
landscape may vary based on rhe slope class wirhin each Summarize
un ir. For example, if you were to consider planning a forest thinning operation on the Brown T ract, it would be useful to know the locations of areas where thinning oper-
slope conditions
arions should use a ground-based logging sysrem (e.g., fell-bunchers skidders, harvesrers, forwarders, ere.) and
Slope class
the locations of areas where the thi nning operations
report
should use a cable-based logging sysrem. Since groundbased logging sysrems are appropriare fo r rhe gender slopes, slope class measuremenrs will help identifY those
FigUJ"~ 13.10 A g~n~ra1 prouss fo r tb~ developm~nt of a sJop~ class condition information for each stand ( manag~m~nt unit) on a landscape.
management units that have the steepe r slopes more
appropriare for cable logging sysrems. The slope class condition of a management unit ca n be measured in the field with clinometers or other surveying instruments or hypsometers, or the slope class cond ition can be computed using a OEM in conjunction with the vecto r GIS database that describes the management units. In the case of the Brown Tract, rather than having field crews spend several days collecting slope meas urements, the average slope class of each management uni t can be calculated with GIS
TABLE 13.1
using a process similar to that described in Figu re 13 .10.
A rabular reporr is generared by the process described in Figure 13.10, and provides a summary of the slope class condition for each of the management units. An annotated vers ion of the output, showing information for the first ten stands of the Brown Tract, is provided in Table 13. 1. The first variable in the table represents the a((ribure that
was selecred ro uniquely identifY each srand (rhe srand
Output of percent slope values for management units
Count
A=-
M;"
Max
Range
Mean
Std
Sum
3 19
343603
0.11
15.44
15.33
5.3 1
3.81
1692.78
2
2186
2354595
0.34
23.55
23.21
9.41
3.76
20564.20
3
770
829386
0.44
22.46
22.02
10.22
4.15
7866.61
4
2884
3 106428
0.28
23.01
22.73
9.54
3.66
27521.07
5
533
574107
1.71
19.80
18.09
8.34
3.14
4446.68
6
1195
1287164
0.44
23.72
23.28
8.51
4.24
10168.51
7
338
364068
0.20
15.15
14.95
6.20
3.52
2096.76
8
2494
2686349
0.15
26. 11
25.95
13.65
4.27
34040.15
9
337
362991
3.20
25.4 1
22 .21
15.03
3.9 1
5066.74
2395
2579714
1.55
24.2 5
22.70
11.52
3.90
27591.07
S.... d
10
number units (10 m grid cells) in the database Area .. squar~ feet Min. minimum valu~ in [h ~ da[abas~ Max .. maximum valu~ in me darabas~ CoUnt:o
Range ,. (maximum value - minimum val u~) Mean z averag~ slope Srd '" standard deviation o f values in [he database Sum = sum of the slope for aU units
214
204
Part 2 Applying GIS to Natural Resource Management
number}. With this value, you could join the tabular data co the stands GIS database. using a one-to-one join process (see chapter 9 for a review of join processes), and facilitate a graphical display of the slope class for each managemenr unit. The variables 'Co unr' and 'Area' list the number of grid cells from the slope class GIS database that F...l1 within each management unit, and the area mat the grid cells represent. The 'M in', 'Max', and 'Range' provide the minimum slope class value within each management unit, the maximum slope class value, and the difference becween these [wo values for each management unit. The 'Mean' is the average percent slope (what was hoped to be obtained for the thinning opportunity analysis) and the 'Std' vari-
able is
me standard deviation of the slope class values of rhe
grid cells located within each managemem unic. The stan-
dard deviation provides information on rhe disrcibucion and variat ion of slopes classes within each management unit, Large standard deviations indicate a wide variado n of slope class values whereas small standard deviations indicate a narrow varia don. In the managemenr of namral resources, the condition of a stream system may also be imporram [0 know from
both a hydro logic and fisheries perspective. For example, you might need [0 understand the abilicy of Streams to support fish populations, o r to understand the potential water runoff im plicar ions from extreme rainfall evems. Stream slope class (gradient) is one common measure of the condition of a stream system. Stream slope class can be calculated by field personnel using clinometers or other surveying instruments or hypsometers, yet this requires a visit to each stream to provide measures for the e ntire landscape , a very costly and time co nsum ing proposition. The slope class conditions of streams across a landscape can, alternatively, be estim ated rather quickly if a OEM and a streams GIS database is available for the landscape. The straightforward approach to calculating slope values for streams would be to fo llow the previous example of supporting a thinning operation. and use the slope class GIS database for the entire Brown Tract. However. si nce the slope class GIS database was created for the entire landscape and only a small portion of the landscape is of interest (the streams), a different approach might be approp riate (F igure 13 .1 1). One solution would be to create a raster GIS database of the streams . A raster database of
Conversion to raster database
Overlay
analysis
Select attribute that uniquely identifies streams
Summarize slope
conditions
Slope class
report Figun: 13.11 A general proces.s for the development of slope inform:uion for each individual stream on a landscape.
215
Chapter 13 Raster GIS Database Analysis
Intervisibility is a term used to describe the number of viewpoints with a view of each unit of land. For example, a number of homes may
be within the viewshed of
205
areas of the viewshed may be visible to a single home. From a management perspective, this information may impartanr. and allow natural resource managers to
be
a property being managed by a natural resource organ-
focus thei r public relarions efforts to rhe directly
ization. These homes could be considered viewpoims.
affected homeowners. If you we re to develop a GIS
By performing the viewshed analysis described above.
database that describes the intervisibility of a landscape.
however, you do not necessarily understand the imer-
the phrase 'cumu la tive viewshed map' (rather than
visibility of the landscape. or the number of homes that
'viewshed map') mighr be used. to illustrate the num-
acmally have a view of each unit of land. In fact, SOffie
ber of homes viewable from each land unit.
streams that matches the spatial extent and resolution of
adonal visitors are influenced by the visual appearance of
other raster GIS databases previously developed for this landscape wou ld fucilitate future overlay analyses. The grid cells th at contain ac cuai values (other than 'no data') should only be those that overlap a stream in the vector screams GIS database. An overlay ana lys is is then performed in conjunction with a OEM CO enable the creation of a raster GIS database thar describes the elevation of each
the landscape. and thus can be negatively affected by manage mem activities that leave visible impactS (Ribe. 1989) . A key to reducing (or heading off) potential public relarions
grid cell represent ing a stream. Now that only those raste r
grid cells that touch a stream have been identified. and the elevation of each is avai lable. you can calculate the average slope class values for each of the streams . Si milar to the previous analysis, the res ult of this analysis is also a tabular database. By using the attribme that uniquely identifies each stream. yo u could join this tabular data with the streams GIS database to facilitate the
display of the slope classes for each stream . A similar processing approach could be used to identifY slope or elevation characteristics for any vector GIS database that con-
problems in natural resource management is to ascertain what pardons of a landscape are visible to recrearional visitors, and to adjus t the management plans associated with
those areas accordingly (Wing & Johnson. 2001). Vicwshed analysis can facilitate an understanding of the portions of a landscape that are visible from specific landscape features of interest. For example. observation or viewing si tes (overlooks) may be represented in a GIS database by poinrs. and the location and elevation of the points would be used to determine which other parts of
the landscape would be visible. If rhe viewing sites were described by lines, the vert ices (points where the line
direction changes) of each line would be used for the viewshed analysis. Fo r example. lines that represent a road
or trail through a landscape. could be used to determine
tains lines (e.g .. roads, trails, facility corridors, etc.) .
which other parts of the landscape can be seen from those featu res.
Viewshed Analysis
The landscape in a viewshed analysis is represemed by a OEM or a TIN (see chapter 2 for a description of a TIN). The objective of a viewshed analysis is to calculare rhe line
The maintenance or enhancement of aesthetic values is becoming increasingly important in natural resource management. Research has shown that the experiences of recre-
of sight berween the viewing sites (e.g., observation points,
homes) and other landscape features (F igure 13. 12) .
Figure 13.12 Line of sight from a viewing site to the surrounding landscape.
216
206
Part 2 Applying GIS to Natural Resource Management
Features [hat are identified as being in [he line of sight of
To illumate [he development of a viewshed analysis, a
the viewing sites are considered visible, whereas all other
GIS database rep resenti ng home locat ions surrounding the Brown Tract will be used as the viewing sites. A cur-
landscape feamces are not considered visible. A number of considerations must be taken into account when conducting a viewshed analys is . Ahhough a OEM assoc iated with a landscape may have been
sory in vestigation of [he homes GIS database reveals [hat [here are multiple homes along rhe eastern half of [he Brown Tracr (Figure 13.13). A viewshed analysis will
acquired (perhaps from rhe US Geological Survey), an
allow yo u to determine what portions of the Brown Tract
assessmem of its fitness
represent the landscape in a
are visible by residents of [he n... rby homes. One of [he
viewshed analysis should be performed. I n heavily
first steps in the viewshed analys is is to define the observer height. You might assume that the average person has a
(Q
forested landscapes. the OEM may nor represent the effect of tree heights on your view, as [ree canopies extend above the elevation surface (t he ground. as represented in the
OEM) . The current management of rhe landscape will also affect a viewshed analysis because different aged stands have different heights, and therefore [he (Op of rhe current canopy (which is what is usually seen in a scenic view) may be misrepresenced. One solution to this problem is co acquire a vegetation GIS database that contains tree height information , and [Q incorporate these measurements inco the DEM elevations. Another consideration is to make sure that the DEM surface cove rs the entire landscape area between a viewing site and the landscape bein g analyzed. In some applicat ions. a managec might be interested in determin ing the visibility of a resoucce from surrounding viewing sites (homes oc roadways) but only has access to a OEM that describes the area managed (such as the OEM for the Brown Tract, as shown in Figure
view height (above ground level) of 5.5 fee 10 per cent) . Although the choice ofa range between 1 a nd 10 is somewhat arbitrary. choosing values from a larger range will help differentiate possible routes more distinctly than will values from a smaller range . The cwo layers are then added to each other through raster map algebra to form a si ngle database with coSt values for each cell. COst weighting and direction are then calculated for the Brown Tract road ne(Work in order to reach the rock pit. After the supporting databases have been developed a shortesr path function can be used to identify a preferred transportation route for the fi ll material (Figure 14.8) . Although th e road system represented in the Brown Trace database is not substantially large in extent, many forest and other natural systems have exrensive transportation ne[Works. 231
Chapter 14 Raster GIS Database Analysis II
Slope GIS database
Conversion to raster database
slope categories
Reclassify road type
Combine values
221
Rock pit GIS database (vector)
Reclassify
Develop cost path & cost direction (raster)
Best path algorithm
(vector) Figure 14.7 A general process
lO
identify the shortest path between two locations on the
Brown Tract.
In these simations, the number of potenrial
fOUCes
can
surpass (he ability of transportation planners to sysremar· ically evaluate and select from a full range of options.
Shorrest path algorithms can assist planners and managers
in making sound decisions.
Creating a Density Surface for the Number of Trees Per Acre
Brown Tract • Rock pit •
Sootheast entrance
-
Shortest path
-
Roads
Figure 14.8 Shortest path between rock pit and southC:l$t entrance of the Brown Tract, given cost weights for road surfaa and road slope.
Density functions can be used to demonstrate [he relative abundance or strength of the locations of features an d anributes, The stands database for the Brown Tract contains an attribute named 'trees_acre', This attribute has a relative weighting of the trees that you would ex pect to find in each stand within the Brown Tract on a per ac re basis. It may be of interest to determine the areas in the forest where this field is strongest, indicating where higher numbers o f trees are more likely to be fou nd. Deter-
mining this info rmation could be done through plotting the polygons and using shaded symbols to demonsttate intensity values of individual polygons. This approach, however, would neglect the influence of neighboring 232
222
Part 2 Applying GIS to Natural Resource Management
What is a centroid? A centroid is a coordinate pair that is intended co represent mid-point of a feature or group of fe-drures. A centroid could be creatai co represent {he center of a group of points by taking the average of all the longirude and latitude coordinates. In terms of a line feature, a centroid position is easily determined. by dividing the tQ[al length of the line in half and using a coordinate pair [Q represent the half-way point. A polygon centroid can be more difficult to determine if the polygon shape is
me
polygons in irs representation . A more helpful approach might be
to
create a densi ty surface which would search
surrounding areas and determine inrensicies that take into account (he nllmber of trees for each stand while considering the number of trees in neighboring stands. A density surface must be created from a point or line
feature type. In order
[Q
irregular or non-homogenous (is not round, square, rec-
tangular, triangular, etc.), patricularly if some or all of the boundaty that make up a polygon contains curves, as is orren the case when describing narural features. The cen-
troid of such a polygon is determined through mathematical integration (calculus) with the goal of determining where the center of gravity of the polygon is located. The center of gravity can be thought of as the point at which the polygon would balance if set flat upon a pole.
strate the detected densities. Figure 14. 10 shows the Output that results for a smoothed density surface using the same stand centroids and a 1,000 ft search radius for trees per acre.
apply the density function to the
stands layer in the Brown Tract, we'll need ro conven the
stands polygon feamce rype. A point representation is probably preferred over a line feature type for the con-
Simple density surface
--
Trees per Icre
o
vened stands. A common method for representing polygons as poims is {Q calculate the cemroid, or middle of a
polygon's extent, which is determined geometrically if the shape is basic, such as ,hat described by round, square, or rectangular fearures. For irregular polygon shapes, centroid determinat ion must be accomplished with more rigorous mathematical techniques. Most GIS software systems will offer routines for cemroid determination and can quickly create a cemroid represemation of a polygon or line feature with one oUCput point created for each
lOw density
CJ _
moderate density
_
hIgh density
Figure 14.9 Simple density surface for num~r of trees per acrc based on a 1,000 ft search rndius.
input feature. In addition, all of the attribute values will be carried into the poim anribuce table. Within [he
ArcGIS software, ,he ArcToolbox has a 'Featu re to Point' conversion command that will accomplish this cransformation. The XTools extension software, a popular lowCOSt ArcGIS extension program, has commands that also support centroid creation. After the stand polygons have been converted to poims, the density surface can be created. Figure 14 .9 shows the resuh of a simple density surface based on the
Smoothed density surface
Trees per Icre
•• i ..
number of trees per acre. The darker shaded areas highlight the areas where greater numbers of crees would be expected. The search rad ius was set to 1,000 ft and density circles were created for each stand cencroid to demon-
Figure 14.10 Smoothed dcns ity sUrhce for based on a 1,000 ft search radius.
o
Iowdenslty
CJ
II!llI _
moderate density
_
high denslty
num~ r
of trees per acre
233
Chapter 14 Raster GIS Database Analysis II
223
Summary We demonstrated in (his chapter how raster databases can be manipulated and analyzed (0 solve questions related [0
bases. The procedures facilitate spatial analys is and su p-
natural resource app licado ns. A host of functions are
are su ited for specific analytical purposes. In add itio n, we presented [Wo potential applications in which some of the
available ro supporc analysis including distance, sracisrica1 search summary. and density functions . The functions
differ in their application and in the types of database structures that can be used for analysis. Omput may be tabu lar, veC(Of, or rast er depending on the function . Several common procedures within most raster-based software include raster reclass ifi cation, raster map algebra.
and conversion routines between vector and raster dar3-
POrt
,he ability
[Q
prepare spa,ial daubases so ,har ,hey
functions and procedures discussed earlier in the chapter were applied. The raster analysis p rocesses and examp les we presented by no means represent the extent of the porential of raster analysis. Rather, these processes an d functions describe so me of the more usefu l and co mmo n commands and processes for namra1 resource analys is with raster data.
Applications 14.1 Straight line discance function for points. A!; parr of your position as a natural resource manager. you manage the research areas on the Brown Tract. Concerned abom
their dimibu
Technical and Institutional Challenges One of the most expensive and time-consuming aspects effore (hat related [0 to using GIS is the effon that is required to creatc create a GIS database. darabase. Dupl Duplicating icating previous data co collection llection effores in [he efrorts rhe creation crearion of a GIS database should always be avoided, avoided. thus a lack of awareness of existing GIS databases is a serious challenge chaJlenge [0 to confront confrom.. To prevem prevent duplica duplicating ting GIS datab.se database development efrons. efforts, GIS GIS users within an the types of GIS organization should be made aware of [he databases that (he the organization can easily eas il y access. This might include GIS products [hac that we re developed GIS database produces imernaJly within an oorganizadon. rganization, GIS databases develinternally oped by GIS contractors or land surveyors, surveyors. or GIS databases that are availab available le through agreements or relationships with wiTh other nacural natural resource managemem management organizations. avai lable GIS dataorganizarions. Information regarding available
241
bases could be stored seo red in a searchable database or it might ca[alogued in a less formalized manner. Regardless of be catalogued how ffiis this information informadon is g:nhered gathered. and stored, stored. personnel in an organization who use GIS G IS should be able co to easily identifY and loca[e locate exiseing existing GIS da databases tabases [hac that might facilita[e lacilitate [ify the tasks that [heir their jobs require. [he rasks [hac Meradara, or information Metadata, informadon documenting documeming [he the specifica[ions database, tions and quality oflandscape feacures features in a GIS database. have become an important aspect of GIS databases in the pase decade (Dobson & Durfee. Durfee, 1999). In order co de[erpast to derermine the fitness of a GIS database for a particular use, use. the me[ada[a related co to the me GIS database should be considmetadata ered. In panicular, particular. when a GIS database is acquired from anomer organization, another organization. me the me[ada[a metadata should be relied upon co verify [hac condition of [he to verifY that [he the condirion the GIS database is what was expected when acquired. acqui red. In many instances, instances. however, Iirde mecadata GIS datalittle metadata exises ex ists to describe the qualities of GIS bases. The twO [wo hypothetical foresrs forests used extensively in examp les. In faCt, lact. more often than this book are prime examples. not you may find [hac that G GIS IS databases developed or maintained by non-federal organizations lack. [ained lack, or have insufficient, mecad:Ha. cient. metadata . Thus, Thus. nacural natural resource professionals musr must be careful when basing decisions on GIS databases where the level of qualiry is uncertain. uncenain. For organ izarions that are involved in producing produci ng and organizations distributing GIS databases [0 to orner other namral natural resource management organizations, guidelines or protocols should be in place that address all aspects aspecrs related to the distribution disrriburion of the GIS databases. WithoUt Without guidel ines. organizations guidelines. organizarions inef-hciem distribution are likely ro to be working with an inefficient syseem , and may be prone co system, to liability liab ility problems. struccure for all availGuidelines should include a pricing Structure ab le GIS databases; databasesi (his chis structure will need [0 to reflect reAecr [he the oorganization's rganization's views on cost COSt recovery. In me the case of publie organizations. rganizations, {here there may be no need or desire to co lic o recover more than than me the delivery costs costs.. Some public organizations, however. do utilize contractors conuacmrs m to collecr collect and izarions. develop GIS darabases, databases. and may need co to recoup some of the coses COSts of doing so. For most private natural resource management organizat ions, the ing srrucrure struccure will w ill managemem orga niza tions. rhe pric pricing likely need co to [eflecr reflect [he the actual accual coses COstS of collecting [he the dara. data. O rganizations [hat that distribUte Organizations distribute GIS databases will also need co comprehensive to develop a comp rehensive liability policy [0 to pro[ecr protect themselves from li[iga[ion. t11emselves litigation. A liability policy will likely need co to be [ailored tailored [0 to each particular GIS database because the content. accuracy. and uses of databases will vary. A [he conrent, method of providing GIS GIS darabases (Q to cuStomers will also need [0 [Q be identified. As discussed in chap[er chapter 15 IS,, organ organiizations zacions [hac that provide GIS data [0 to [he the publ public ic should make 252
242
Part 3 Contemporary Issues in GIS
GIS databases avai lable in an expedient manner and use cu rrent Internet technologies such as creating new wehsites chat allow users (0 browse darabase offerings and download both data and metadata. In addition. file transfer protocol (FrP) can be used to make data available. but this data-sharing system is less user-friendly. For organizations that are nor involved in producing and developing GIS databases for other namrai resource management organizations. the high COSt assoc iated with crearing GIS databases can result in a reluctance [0 share databases with other organizations. For example. certain GIS databases may contai n sensitive information. an d might reveal the location of landscape features (such as endange red species nesting locations or a rcheological sites) that might be disturbed or destroyed should the locations become public information. Two examples of these GIS databases include the location of endangered species nest sites, and the locacion of genetically-modified [fee field trials. The GIS databases may also comain information about the status oflandscape resources that would be of value to another organization with which it competes, providing the other organization an advantage in the market place. This is clearly importam tOday as the number of land sales has skyrocketed. and potential (a nd perhaps hostile) investors desire complete information regarding a land asset. A reluctance on the part of federal public organizations in the US to share or publicize data holdings is that all federal agencies are subject to the Freedom of Information Act (FOIA). FOIA was signed into law in 1966 and later ame nded in 2002. FOIA makes it possible for individuals (Q request access to federal agency records,
To facilitate a recem landscape analysis research project, a private natural resource managemem organization agreed to provide a highly detailed GIS database that desc ribed the management units within their ownership boundaries. The GIS database had been assembled at great COSt and effort, and revealed considerable information abom the natural resources that the private organization managed. The private organization required that a confidentiality ag reem ent be signed prior to making the GIS database available (Q the resea rch project. As noted in the confidemialicy
exce pt in cases where records are protected from disclosure. These requestS can also be made for spatial databases [hat were produced by federal agencies. In so me cases, spatial databases may comain information that the agency considers to be sensicive or potenrially damaging (Q the resources it manages. One example of sensitive information might be spatial records of vandalism that occur in public recreation areas. Evidence of high vandalism occurrence might deter visitors from staying in affected a reas, and potemialiy reduce revenue that is gathered from day-use permits. In addition, there may be hesitation to draw additional arremion to 'hot-spotS' of criminal activiry in case the arrenrion may encourage others to visit these locations our of curiosiry or to cause additional vandalism. Other hesitations may involve the presence of unusual features, such as special habitat areas, or archeological sites, and the potencial damage that too many visimrs may bring to these areas. For various reasons, federal agenc ies may not openly advertise the types of spatial databases that they have produced. A primary hesitation to do so is likely because of the uncertainty of what will happe n with the information within the databases should rhey become widely circulated. Understanding that there are factors that may hinder an organization's willingness to share a GIS database is important prior m requesti ng the database. For data of a sensitive nature, it may be possible to enter into a confidentiality agreement to gain access (Q the GIS database. Ultimately, it may be necessa ry (Q pay a large sum of money ($5.000 to 10.000) for a GIS database. and for some organizations, it may be difficult m locate the necessary budget resources appropriate for this type of purchase.
agreement, access to the GIS database was limited to the primary scientists of the research project, the sha ring of the GIS database with others was prohibited. and protocols for distribming information drawn from the GIS database were outlined. Without the confidentiality agreement, which facilitated the sharing of the GIS database, knowledge of [he status of resources located within the private natural resource management organization 's ownership would likely have been less accurate, reduci ng the confidence placed on the landscape analysis results. 253
Chapter 16 Instrtutional Challenges and Opportunities Related to GIS
Benefits of Implementing a GIS Program The decision to implement a GIS program (the entire suite of hardware, software, and personnel related CO the use of GIS within an organization) can be intimidating for natural resou rce management organizations. There are many factors to consider, including investments in software, hardware, personnel, a nd CIS databases. Since natural resource management organizations cypically rely on maps or mapped data to assist in making decisions, GIS ca n allow an efficiem storage of maps, and can fac ilitate the generation of multiple versions of maps in a timely manner. In addition, GIS allows landscape features to be measured, analyzed, and integrated with other GIS databases in an expedient manner. New technology has provided [he pQ[entiai to convey information [Q field managers very quickly. These capabilities, if managed properly, can allow natural resource management organizat ions to make bener management decisions, more accurately gauge the effore and cost of potencial namral resource managemem projec(S, and increase the efficiency of tasks performed by their employees.
Successful GIS Implementation Saving money, reducing the amo unt of time spenr in the office analyzing options. and thus saving resources for Q[her [aSks and management activities are common goals of namral resource managers. Distributing GIS capabilities to field offices has been suggested as one way to address some of these issues. Successfully implememing and managing a GIS program can be difficult, as the costs of implementation and ma nagement vary from one organization to the next. Perhaps the strongest ingredient for success in implementing a GIS program is in esrablishing an organizational commitmenr within the upper levels of managemenr of an organization . This commitmenr needs to view the GIS program as more than JUSt a shortterm experimenr that can be discarded after early, d isappoinring results, since initial GIS products and experiences are likely to identify implementation problems. Unfortunately, upper-level managers tend to be less
243
fumiliar with GIS technology than the actual GIS users. For this reason, GIS users should communicate their supporr of GIS in terms that are comp rehensible to the upper-level managers and help th em understand that when technical difficulties do arise, program implemenracion plans must be adjusted. Upper-level managers, in turn, should promote the G IS program as a way of making more efficient the tasks required of natural resource management. Organizations must also be realistic about the time, effort, and budgetary resources that individual GIS projects or analyses wi ll require. Proper planning requ ires that project objectives be clearly defined. Objectives provide a project mission a nd ca n help keep personnel focused, should setbacks occu r. Project objectives a re also important for establishing standards that allow you to gauge the success of a GIS project or analys is. Achievement benchmarks are also critical in justifying the continued and expanded use of GIS within an organization. Allowing users of GIS to become involved in the planning and implementation of GIS projects is also importanr, since they may be among the best qualified to assess whether GIS can accomplish the approp riate project tasks, which may lead to an improved level of efficiency in the managemenr of natural resources. Finally, GIS user training is an importanr co nsideration for the success of a GIS program within a natural resource managemenr organization. Most recenr college graduates from natural resource programs will likely have a rudimenrary knowledge of how GIS can ass ist in natural resou rce management, but will likely lack the level of experience you gain from using GIS periodically for actual, on-the-ground, management purposes. To accelerate the development of personnel, organizations can provide GIS training inrernaIly, or can allow personnel to auend cominuing education courses or pursue other training opportunities. Geospatiai training courses and opportunities appear to be increasing within natural resource disciplines (Wing & Sessions, 2007). This investment in conti nuing education increases the knowl edge level of personnel an d demonstrates an organization 's commitment to its personnel, which hopefully leads to increases in work efficiency and productiviry.
Summary What would the management of natural resources be like without a few challenges? With the imroduction of
new computer-related measurement technology and accompanying GIS databases, natural resource manage254
244
Part 3 Contemporary Issues in GIS
menr o rganizat ions are facing numerous c hallen ges
come, eve n as new and va ried issues arise. As we have
related to GIS data manageme nt. Given that GIS-related
noted, many of [he challenges facing the use of GIS
technology continues (Q evo lve to meet (he increasing needs of society. an optimistic person might expect that
within natural resource manage ment organizations are
the challenges described in this chapter can be over-
zational issues .
related to GIS databases, technology, people, and organi-
Applications 16.1 Sharing GIS databases with people outside your organization. What issues sho uld a natural resou rce manage me nt o rga nization address when cons idering making G IS databases ava ilable to other organizations. including perhaps competitors?
16.2 Managing GIS technology. A seaso ned professional (Mary Swarthmore) who manages ano ther department (Acco unting) in yo ur natural resource o rga nization is considering a career cha nge. This change will result in Mary managing your organization's GIS department. Mary has never managed a GIS department before, nor been involved in the creation. acq uisition. o r use of GIS
databases. She has approached you fo r some advice regarding the to ugh issues G IS managers face when concerned abom successful implementadon of a GIS pro-
resource management organization has developed a GIS database that may conta in in fo rmation that could assist you in some part of your job. Under w hat conditi ons should you expect you r co lleague to provide you access to
the GIS database?
16.4 Distributed GIS program. You work for a large integrated forest products co mpany th at has a ce ntral office and five district offices. The company has been
attempting to shift GIS capabil ities to the field offices by pu rchas in g the appropriate hardware and software reso urces. and insist in g that field perso nnel (foresters. biologists. managers. and others) use it to make maps associated with their management activities. Afte r twO
years, only one of the five offices has successfully implemented the program.
gram. What might yo u advise Mary?
a) Why do yo u th ink the o ther four offices have been
16.3 Sharing GIS databases within an organization.
b) Why might the o ne office have been successfu l?
less than successfu l? You've heard that a colleague in anothe r private natural
References Bettinger, P. (I999). Distributing GIS capabi lities to fo restry field offices: Benefits and challenges. journal ofFomtry, 97(6), 22-6. Dobson, J.E., & Durfee, R.C. (I 998). A quarter century of GIS at Oak Ridge National Laboratory. In T.W. Foresman (Ed.), The history ofgeographic information systems: Perspectives from the pioneers (pp. 231-63). Upper Saddle River, NJ: Prentice-Hall. Natoli, J.G ., Pelgrin, W.H., Oswald, B. , & Montie, K. (200 1). Geographic Information Systems: The wave
Puger Sound LiDAR Consortium (PSLC). (2007). Puget
Sound Lidar Consortium: Public-domain high-resolution topography for Western Washingron. Retrieved April 23, 2007, from http://pugetsoundlidar.ess.washington. edu/ index.htm. Wing, M.G., & Bettinger, P. (2003). GIS: An updated prime r on a powerful management roo l. jou rnal of
Forestry, 101(4),4-8. Wing, M.G., & Sessions, J. (2007). Geospatial technology education. journal ofForestry 105(4), 173-8.
of the fmure for information analysis. Public Works,
May, 22-9.
255
Chapter 17
Certification and Licensing of GIS Users Objectives The progressio n o f GIS use in natural reso urce managemem has been evolving from development and imple-
mentation of systems. to (he distribution of analytical capabilities to field offices, to porting spatial technology into the resource setting. The evo lu tio n of GIS can be viewed from (he perspective of a single organization or from [he perspective of nationa l o r worldwide GIS communities. From a global perspective. GIS is facing one of its greatest challenges: mat of implementing certification and licensing processes to define and recognize 'professio nal' GIS users. In recem yea rs, co nce rn s have been voiced from the land surveying and engineering commu-
niry regarding the definition of surveying activities and
2. what organ izations might be relevant in certification and lice nsing discussions. an d 3. how cenification and licensing issues might affect the typical GIS user in a natural resource organizacion .
During the last 10 years, one of the primary goals of the GIS comm unity has been to educate other profession-
als and the general public about the power and usefulness of GIS beyond its map production capabilities. Many people in natural resource o rganizatio ns (as wel l as academia)
view GIS only as a map-making tool and may have li mited understanding of its analytical power. The American
Society for Pho togrammetry and Remote Sensi ng (ASPRS) and the University Consortium for Geographic Information Science (UCGIS) are perhaps the most active
whether GIS practitioners impede upon traditional surveying accivities when colleccin g o r map ping spatial data
groups in educating those who use GIS, as well as the pub-
(Gibson, 1999). These concerns have foStered debate
Nacional Convemion in Tampa, o ne AS PRS member was
among surveyors. engineers. and GIS users regarding the types of measuremem and analytical activi cies that migh t be required to compete ntly perfo rm spec ific accivit ies in co njunction with the analysis. Thus, in an effon to gain credibi lity and recognicion among other professions. the GIS commu ni ty has pondered the iss ue of cenificatio n and licensing.
heard to remark ' If you claim that GIS technology is vital
After reading th is chapter and exploring the questions posed in the applications sect ion. students sho uld have an awareness o f: 1. why certification and licensing of GIS users is being
debated,
lic, about the ca pabi lities of GIS. At the ASPRS 2007
to society yo u should also promote the need for certifica-
tion and licensing among GIS users.' The UCG IS cond ucts nadonal meetings rwice a year to identify research and other act ivi ties that wi ll idencify and promote rhe use of GIS as a problem-solving tool for society. At a narionai
meeting in June 2000, a member of the UCGIS asked the other delegates. 'Now that everyone seems co know about GIS , what are we going to do about it ?' These seemingly innocent observations speak positively about society's growing awareness of GIS while also indicati ng a potent ial pitfall within natural resou rce managemenr: GIS has been embraced by natural reso urce organizations in a way th at 256
246
P art 3 Contemporal'( Contemporary Issues in GIS Part
faci li[a[es open use by any professional who might be facilitates inreresred rhe <echnology. technology. Only recently has [he the GIS GIS interested in the commu community nity begun [Q co discuss in depth whether professional standards musr must puc put in place (0 to ensure professional professionaJ analysis. competency for data development. anal ysis. and other (asks. tasks. Given Gi ven that ocher o ther professions profess ions initiated these discus-
me
sio ns, leaders and champions of o f rhe the lise of GIS in natsions, ural uraJ resource management managemenr have found themselves at rimes times
The more esrablished established and rigo rigorous rous of the me oprions options is that made by [he the American Sociery Society for Photogrammetty Phorogrammerry & Remote Sensing (2006). The ASPRS has created Remore crea<ed a Mapping Scientis Scientistt certification for GIS GIS users and has also certification ifi catio n programs for remote sensing and created cen photogrammetry. photogram merry. The M Mapping apping Scientisr Scientist cerrification certification requi res app licants to develop a sratement statement of accomplishrequires ments, mems, which wh ich are peer-reviewed. peer- reviewed, and to (Q pass a written
exam . Currenrly, Currently, only abou aboutt 50 people are cerrified certified as Mapping Sciemisrs Scientists rhrough through rhe the ASPRS. The APSRS has
in a defensive position when addressing issues rdated related (0 CO the rhe call for licensing and certification cerrificarion of GIS users. Throughout the Throughour rhe US, some professional land surveyors
also creared crea red technologist-certifications for GIS, remote
and engineers have perceived th that at GIS users use rs we re vio lating
sensing, and phowgrammerry. photogrammetty. These GIS GIS technologist rechnologisr
scate state surveying laws when usi ng CPS GPS [0 to collecr coliect spacial spatial data, and when reponing reporting positional accuracies accu racies of coldata. lecred measurements. Some of these GPS data dara collection collecrion
certifications require less work experience than the
acdviries have actually acrually led to activities ro legal dispures, disputes, parricularly particularly when the [he collection collec rion and mapping of sparial spatial data dara refercoees land ownership locations. ences locado ns. In California, state legisaccivities involving spalation was developed to clarify the activities
tial rial data dara collection collecrion (Korte, 1999). These acriviries activities full fuJI inro into [wo caregories: (I) (wo categories: (1) those that thar constitute consrim<e 'land surveying', and hence require professional licensing, and (2) all other
Mapping Scienrisr Scientist and other full certifications (th (three ree years as opposed [0 co six). six) . There are currently currendy s ix certified
G GIS/LIS IS/LIS Technologists. The Urban and Regional Regional Informar Info rmation ion Systems Sysrems Association Associarion (U (URlSA) RISA) iniriared initiated a GIS professional certificarion tion program in 2004 2004.. The GIS Certification Institute Innirure (GISel, (G ISel, 2007) manages the rhe cerrification certification program, which resulrs in qual results qualified ified app ap plicants licanrs becoming recognized as a certified geographic informarion information sysrems systems specialisr spec ial ist (GISP).
activities, ich do not activ iti es. wh which nor require professional professional licensing. Most states stares have registration boards mat (hat license and regsurveyors engineers. ulate land surveyo rs and engi neers. These boards often have rhe the abili ability interpret exis existing ting starutes statutes and laws that ty to interprer
Three categories of experience must be ~ demonstrated in order o rder to qualify. The primary experience necessary in ga ining GISP cert ification is a documented work history gaining
rcpresemarion of spacial spatial data data,. govern the collection and representation and may also initia initiate. te, support. suppocc, or approve legislation legi slat ion regarding collection llection activities. In addition addition., regard ing spatial data co state-level professional land surveying sociecies societies are active modifying laws regarding spatial data dara in promoting or modifYing
ence category is an education background that (hat can be sat-
collection, and they may occasionally engage political lobcollecrion, byists to assist in influencing the legislative legislative: process. the la land nd surveying and engineering In contrast to rhe fields. the GIS community is not directly fields, direcdy re reppresented resented or controlled nally- or state-recognized licensing contro lled by a natio narionallyboard in most mosr cases. cases. There are GIS-related GIS-relared professional state oorr province level but these societies societies at the srate long. and are typically have nOt been in existence for very long, gene rally nor very acrive active in influencing legislation rdared related generally to spa spatial rial data co collection. llection. The main objective of stare-level state-level iona.l-Ievel GIS socieries societies has been to communicate and nat narional-Ievel information related to the collection. collection, maimenance. maintenance, and to interested inrerested users. analysis ana lysis of GIS databases (Q
Current Certification Programs In terms of nationally-recognized GIS certification cerr.ification prothe US. US, there are primar primarily il y twO (wo current options. grams in rhe
involving GIS GIS and orher other spa spatial rial tools. rools. T The he second experiisfied by arrending attending conferences and workshops, as well as completing formal education ed ucatio n programs or o r earning cenificertificates. The third category is desc ribed as 'contributions' 'conuiburio ns' GIS publications. pub lica tions , conference planning or and includes GIS presentations. presentations, and volunteer vo luntee r activities related to GIS.
As Ai; of October 2007, there rhere were 1,709 certified GISPs, giving the GISel GISCI program visibility visibiliry beyond [he the certification rion programs of the rhe ASPRS. Airhough Although the rhe creation of the GISP certification is noteworthy. the experienced-based
portfolio approach to ro qualifYing as a GISP lends itSelf irself to ro criricism (Lo ngley et er aI. aI.,, 2005). It I[ remains to ro be seen criticism (Longley whether a certification approach rhar that does not include a writte n examinarion examination w will recognitio itio n written ill receive respect and recogn disc iplines. IIn profess ions and disciplines. n addition. addition, there from other professions is no clear merhod. method for addressing unprofessional unprofess io nal activicies activities
related to GIS, should [hey reJa<ed they occur. Given the emphasis on self-reporti ng of experience, ex perience, another issue of discussion is se lf-reporting status. whether any applicants have been denied GISP status. Many co lleges and universities now offer certification IS, as well as other spatial data degrees mar rhat are related ro to G GIS. co collecrion llectio n and analysi analysiss rechnologies. technologies. however no Stanstandards exist ex ist for what should be offered in those curricu257
Chapter 17 Certification and licensing of GIS Users
Besides the current course that you are taking (that
fishe ri es, wildlife, oceanography, forestry , soi ls,
hopefully uses this book), what other GIS courses
rangeland resources, and others. More than likely, GIS courses offered in departments other than geography will provide a different perspective on wha t is important (Q smdents pursu in g namral resource
does yo ur uni versity, college, or community coll ege offer? Although there are many educational instimtions that offer coursework or curricu lum related to GIS, these programs can look quite different from one instimtion to [he next. At most educacional instim[io ns, for exam ple, GIS courses are located within the
247
degrees. If a university or college does not offer courses related co GIS, students can still learn about
geogra phy department. H owever, special ized GIS
the capabilities of GIS through Internet courses, selfstudy of GIS texts, and volunteer wotk wit h local
cou rses may be found within departments such as
agenc ies or gove rnment offices.
lums. The National Center for Geograph ic In formation & Analysis (2000) has produced, and suggested for use, a core curriculum to serve as a foundation fo r studies in
The fi rst section of the Model Law clarifies the necessity for guidelines by stating that the practices ofland surveying and engineering are a matter of pubic interest. The
GIS. Typically, GIS users can earn a GIS certificate after
decisions made (or recommended) by people employed in these professions can potentially affect the life, health, and property of the general public. Sectio n 2 defines the tasks
tWO yea rs of part-time study. While this option does include organized coursework (and perhaps exams to evaluate competency), programs that offer cert ifi cat ion degrees lack a recognized set of certification gu idel ines
and are ge nerally not accredited by a professional GIS or remote sensing society. As mentioned in chapter 15, the GIS&T Body of Knowledge (DiBiase et aI., 2006) was recently published in o rder to define critical concepts and skills that relate to geographic information science and
that are associated with surveying and engineering and no longer refers directly to GIS , as it did in an earl ier version
of the Model Law. Section 2 does state that mapping involves the configuration of the Earth's feamres. the subdivision ofland, the location of survey control points, reference points, or property boundaries, and thus people performing 'mapping' are performing the 'Practice of
technology (G IST) . This document was created through
Surveying'. The Model Law suggests that these people
the joim efforts of many GIScience researchers and educators, and is an initial attempt to define the skills that you can use to describe geospatial com petency. A second edition is intended that will provide detail fo r instructional activities that suppOrt important geospatial concepts and
should be reg istered as professional surveyors before engaging in those act ivit ies. The most d irect pathway co becoming a professional surveyor is to fi rst graduate from an accredited fou r-year college program in engi neering or surveying. Then yo u must successfully pass an e ight-hour written exam covering survey in g fundamentals. Once the fundamentals
skills.
The NCEES Model Law
exam has been successfully passed, four years of surveying
The Nationa l Council of Examiners for Engineering and Surveying (N e EES) has developed a set of guidelines
experience under the supervision of a licensed surveyor must be accumulated before admittance is allowed to an eight-hour written comprehensive exam covering surveyin g principles and practice:. Once this comprehensive
described in a Model Law document
to
help states with
licensing issues related co land surveying and engineering (National Council of Examiners for Engineering and
Surveyi ng, 2006). The Model Law contai ns reference to GIS activities assoc iated with spacial data collection and use. The Model Law contains 29 sections that are
exam has been successfully passed, and all other state-level requirements are satisfied, you are qualified to become a professiona l land surveyor. Those who grad uate from four-yea r surveying cu rriculums that are not accredited must spend an addirional rwo to four years working in
designed to help state boards and other legislative bodies
the land survey profession before they can be admitted
create or amend laws for land surveying and engineering.
to the fundamentals exam. T his process of attai ning
258
248
Part 3 Contemporary Issues in GIS
licensing is daunting, and requires a long-term commitcommi tmenr ment on [he rhe part pare of potential poten tial su surveyors rveyors or others who w wish ish to to comply with the Model Law. For many cu current rrent act ive engagenatural resource professionals who use GIS. active ment in a career combined with family and communiry comm un iry commitments offer few realisdc realistic opportuniries opportunities [0 to pursue the guidance of a an engineering degree or to work under rhe land surveyor. Earl Earlier ier versions of the Model Law definition of surveying thar that contained comained direct mention of CIS GIS were criticized as being [00 (00 strin srringenc ge nt and expansive in its description descriptio n of afG GIS IS applications within thin surveying acrivities. activities. Several promiap plica tions wi nent organizations related [0 (0 surveyin survey ing g and an d G1S G IS coauthored a repon repan that suggested modifications [0 the Model Law (Ame (American rican Congress on Surveying and Mapping [ACSM] et aI., aI. , 200 2001). I). The report urged lhe the NeEES to co drop dro p exp explicit licit reference [Q to GIS as a data maniprefine the definit definition ion of sura nd mapping tool. to refme ulation and veying it was less broad, and [0 ro specifically include veyi ng so that i[ and exclude certain activities in its defini definition tion certai n GIS-related accivities of surveying. The or T he NCEES appears to ro have acted a[ a( leas leaS[t in pan recommendations. mmendatio ns. Previous to the Model parr to these reco Law changes, changes. the broader definitions of surveying su rveyi ng created creared difficulties states that char tried to incorporate the difficulries for some stares into their own ow n state Model Law's definitions of surveying inco starutes regulations ations (Thurow & Frank, 2001). statutes and regul
me
The Need for GIS Certification and Licensing Chief among amo ng the t he argumenr argum enr for GIS G IS certification certifi ca ti o n is an assessment ment of how GIS activities might impact society's sociery's assess
welfare and safery safety (Gibson, 1999). The surveying and engineering engineerin g professions professio ns have long been involved in deterdete rmining how best to accllrarcly accurately and precisely collect and ana lyze Earth Earrh aand nd structural strucwral measurements. Since Si nce land va lues in North Ame America rica w will ili likely likdy concinue conrinue to increase as values population also increases. increases, land areas that are the human popularion resulr in large monetary losses or inco rrectly measured incorrectly measmoo can result
gai ns for land ow gains owners. ners. Knowing the reliabi reliabiliry lity of land measurements (exp ressed through uncertainty unce rtai nty estimates) es tima tes) ro make better bener decisions. While will allow land managers to you may argue that a distinct parr of natural natu ral resource data quantifying collection and analytical processes involves quamifying and a nd expressin expressing g the th e uncertainty that is associated with those measurements, measurements. this quanrification quantification is rarely used to rate the reliab ility reli ab ili ty of measurements measu rements collected. collected. Thus the surveying and enginee ring professions may be better bette r posi-
tion tioned ed to provide informa information tion regarding r~garding the inherent inherem uncertainty uncertainry in land meas measurements. urements. Land surveyors are also charged with locating estab~ Jocaring or estab-
lishing roads and other utilities such as power lines and fire hydrants. hydranrs. Surveyors have argued that [hat it is inappropriate for un licensed surveyo surveyors GPS equipment for unlicensed rs to operate CPS this pu purpose rpose since locational errors can potentially potenriaIly affect addition, rveyo rs public safety. safery. In add ition, since professional su surveyors must absorb rhe th e on-going costs associated wi with th maintainma i ntain~ ing licensing and liability liabiliry insurance, they are a re likely likel y to charge more mo re fo rr thei theirr services than unlicensed GPS CPS operath at it is unfair unfai r for unl unliitors. Thus, surveyors have stated that censed o perators rs to compere compete with professional land ce nsed GPS operaro g these types typ es of data d ata collection surveyo rs in offerin su rveyors offering servICes. servIces. Licensing varies [he world and Licensin g va ries throughout throu ghout the a nd may also scate o r province. In addition, rs that the vary by stare add itio n. it appea appears numbe r of licensed professions is in transition transicion . In number Ca nada. professions mat that require requ_ire licenses are referred to as Canada, regulated regula red occ occuparions. upations. There are approxima[e1y approximately 50 di diffferenr ferent regu lated professions professio ns in Canada (Government (Govern ment in Canada, Ca nada, 2007). Within [he the US, the th e number of profesrequ ire a lice license nse fo forr participation participario n has been bee n sions thar require gradually increasing. reasing. Doyle (2007) found that lha[ about abo u( graduall y inc 20 per cem cent of all professions in the rhe US requi require re licensing, up from abo abour ut 5 per ce cent nt duri during ng the early 1960s. Abou Aboutl 50 professions have a regist registration ration process (hat that is recogstares. Some people criticize (he the licensin licensingg nized in all 50 srates. thal it resulrs results in higher prices for servprocess and claim thar a n equivalent gain gai n in the qua quality lity of the rhe services without an ice or good provided. In additio addition, n. some peo people ple see enrry entry into the profession as being bein g prohibitively prohibi tively limited once licensing is in place. While Kleine Kle inerr (2000) found (hat thar incomes from licensed occu occupations pations were we re higher for those training. occupations that [hat required more education and training, growrh was evidenced in many faste e mpl oy ment growth fasterr employment licensed professions, such as engineering and law. law, when co mpared to no compared non-licensed n-licensed professions. professions. Kleiner Kle iner (2000) also reports that empirical empirica l evidence address addressing ing whethe whetherr licensing results in greater soc societa ieta l good s. s, such suc h as increased inc reased safety, safety. is currently lacking. nforcement are additional additiona l Iss ues of conrrol and eenforcement stares have aspecrs of licensing and certification . Most states aspects clearr defi definitio ns of what constitutes acceptab le surveying clea n itions pracrices. rules les typically rypica lly cover what clients should practices. These ru expect ional land surveyor's services, services. in ex pecr from a profess professional ethical considerations. terms of products. producrs. as well as erhical consideratio ns. Ethical co conside nsiderations rarions nor nOt only on ly address add ress the surveyor259
Chapter 17 Certification Certffication and Licensing of GIS Users Users
client relationsh relationship ip but also offer advice on the professional relationships cha thatt should exist berween becween surveyors. Most Sca tes also have developed [0 fac facilitate il itate (he the subsu bstates develo ped a process [0 mirral of complaincs minal complaints against land surveyors. surveyors. All state to revoke a surveyi surveying ng licensing boards have the power ro disciplinary inary infractions occur as a resulr result of comlicense if discipl plaines. plainrs. This Th is power power encourages most land surveyors [0 to become familiar with board rules for professiona1 professional conduct they engage in survey act iviand [0 (Q adhere to the rules as [hey activinor exist in general general GIS use in natural namral ties. Such rules do not resource management. Criticism icism has also aJso been weighed agai aga inst nst un unlicensed licensed Crit GIS GIS and GPS users due to [he the lack of an acc redited redi ted educadonal tional curriculum. curriculum . Most professions, including forestry and wildlife. wildlife, have h.ve identified an education. educa[ionalI and professional backgro background und [hat that is necessary for accred acc rediitat ta tion ion or within their fields. The accred accreditation itation process usulicensing IlJithin suggestS me the coursework, minimum competency sranstanally suggests dards, professional standards, and imegration with other disciplines [hat that should be provided to students pursuing degrees in [hose those fields (Huxhold. (Huxhold, 2002). An accreditation process for GIS G IS could be developed (and perhaps is currently under developmenr). development). However. However, me the dilemma is rhat thac srudy, [he the students in [hose those if you view GIS as a field of study. curricula cu rricula will graduate with a geography degree, degree. and be considered professional profess ion al geogra geographers. phers. Most "amrai natural foresters, soils sciresource organizations hire biologists, foresters. entists, or other associated professionals. professionals, as well as geogranatura l resources. phers, to assis assistt in the management of natural orga nizations increasingly expect all of thei theirr perThese organizations sonnel ucili1.e ize GIS, GlS. not nor just JUSt those who have obtained a so nnel to util degree from an accredited accredired geography program.
GIS Community Response to Certification and Licensing As you might expect, expect. some members of the GIS commuto the sugges suggestion tion that thar certa cerrain in nity have voiced opposition co ies be included in the lis t of funcrions functions only co to GIS activit activi ries be performed by land surveyors. C Criticism riticism has been directed toward the sometimes-broad definitions of land surveying, surveying. and whether GIS databases deve. developed loped and distributed by public agencies should require management by a licensed surveyo Uoffe, 200 1). In addition, surveyorr GofFe. addition. some rtifica tionn and lilicens ing proposals for ce certificatio censin g have also been co uld viewed d prevenr natural \'iewed as exclusionary. and coul resource professionals from performing ~rforming the GIS GIS activities acr iviries thatt they tha rhey historically performed. performed. Thus the opposition opposirion to
249
certification cert ifi cation and licensing from the natural resource GIS GI communiry, community. largely composed of natural narural resource manage rs. bio logists. foresters, others. is to be expected . agers, foreseers, and others, Defining those areas of spatial data co llection and collection
mapping that porentially potentially afFecr .ffect public welfare welfure and safery safety is easy in some cases, cases. and challenging in others. orhers. Clearly, Clearly. GIS databases used for acdviries activ ities such as navigatio n, n. locatwhe n excavating. excavating, or ensuring ing underground facilities Facilities when that property boundaries have been accurately located and used in calculating calcularing land areas, areas. should fall full under [he the purview of a professio professional nal lan land d surveyor o r enginee engineer. r. Acriviries ities (hat that involve displaying display in g data for illustrative purActiv activities ies [hat that may only poses, recreational recreatio nal gu ides, or in activit .ffect perso n{s) or agency responsible for the associaffect the person{s) ated ared decision, decision. could perhaps fall ourside outside [he the purview of a surveyor or engineer. engineer. Although these professional land surveyor illusrra te distinctions di stinct ions between becween acti activvici it ies es [hat chat examples illustrate clearly affect afFecr public welfare and safety safery and those rhose that [har do not, nor, there are many orher other examples rhat that are less clear and will require furthe lUrrherr discussion before agreements between the rhe surveying and GIS commun ities are reached. In cases where mapping products mighr might affect public welfare and safety, severa severall scenarios fo r making map consumers aware of pocenriallimitations poremial limitations have been suggested. Suggestions include requiring requ iring that (hat maps explicicly explicitly iden identify ci fy me the source so urce documents, associated assoc iared meradara, metadata, appropriate approp riate use of map content, and any positional comenr. positional adj adjustments usrmenrs Goffe. Uoffe, 2001). Examples Examp les of these caveats caveatS and disclaimers were intro intro-duced in chapter chapee r 4. Whether these caveats will continue to ro fall shon short of requiring requirin g a certified or licensed GIS G IS professional sional to ro develop the mapping products wi ll likely be an area of discussion. Finally. Finally, and inreresringly. interestingly, some people have argued that the surveying profession has traditionally not required training tra ining in GIS in order to obtain professional surveyi ng cert ification (American (American Congress on Survey Surveying in g surveying and Mapping, Mapping. 1998). For th [h is reason. reason, they argue. argue, it i[ may not nor be app appropriate ropriate fo forr land surveyors su_rveyors to manage rhe the developmenr. maintenance. maimenance. and use of GIS databases. development. syllabi the national However, the current exam syll abi for fOr rhe narional land surveying exams (borh (both the rhe fundamental fundamemal an andd professional principles include I.nd informainfo rmapri nciples and practice) do incl ude GIS and land tion systems as potential poremial exam topics. topics.
MAPPS Lawsuit In June 2006, rhe the Managemenr Management Associarion for Private Phorogrammerric Photogrammetric Surveyors (MAPPS) and rhree three other 260
250
Part 3 Contemporary Issues in GIS
engineering lawsu it professional enginee rin g associations filed a lawsuit against government. The lawsuit was fi led on aga inst the US governmenr. W