`Bradium Technologies LLC - patent owner
`Microsoft Corporation - petitioner
`IPR2016-00448
`
`1
`
`
`
`2
`
`
`
`GEOGRAPHICAL INFORMATION SYSTEMS
`
`3
`
`
`
`VOLUME 1
`
`: PRINCIPLES
`
`GEOGRAPHICAL
`
`INFORMATION SYSTEMS
`
`PRINCIPLES AND APPLICATIONS
`
`EDITED BY
`
`DAVID J MAGUIRE,
`
`MICHAEL F GOODCHILD
`
`AND
`
`DAVID W RHIND
`
`Longrncm
`Scientific 8:
`—-v Technical
`Co-published in the United States and Canada with
`John Wiley 8: Sons, |nc., New York
`
`4
`
`
`
`Lungmnn Scientific and Technical,
`Longman Group UK Ltd
`Longman House, Burnt Mill. Harlow,
`Essex CMZCI 211-3. England
`and Associated Companies throughout the world.
`
`capable".-rlzed in the Um'ren' Stores and Canada with
`John Wiley di .S‘ort.r. lira, 605 Third A venue. New York,
`N Y 10158
`
`© Longman Group UK Limited 1991
`
`All rights reserved; no part of this publication may be
`reproduccd. stored in a retrieval system. or transmitted in
`any form or by any means, electronic, mechanical.
`photocopying. recording, or otherwise without either the
`prior written permission of the Publishers or it licence
`
`permitting restricted copying in the United Kingdom
`issued by the Copyright Licensing Agency Ltd, 90
`Tottcnharn Court Road, London WIP 9HE.
`
`Trademarks
`
`Throughout this book trademarked names are used.
`Rather than put a trademark symbol in every occurrence
`of a trademarked name, we state that we are using the
`names only in an editorial fashion and to the benefit of the
`
`trademark owner with no intention of infringement of the
`trademark.
`
`First published I 991
`
`British Library Cataloguing in Publication Data
`Maguire. David J.
`
`Geographical information systems: Principles tutti
`applications
`1. Title
`II". Goodchild, Michael F.
`Ill. Rhind, David W.
`9lU.9lll
`
`ISBN Dr-582-05661-6
`
`Library of Congress Cataloging-in-Publication Data
`Maguirc. D. J. (David J.)
`Geographical
`information ssyyyystems I by D. J.
`Maguire.
`Michael F. Goodchild, and David ‘W. Rhind.
`cln.
`
`Includes bibliographical references and index.
`Contents: v. 1. PrineipIes—v. 2. Applications.
`ISBN U-4'r'tJ-21789-8
`
`1. Geographical information systems.
`I. Goodchild, Michael F.
`1!. Rhind.
`David.
`lll. Title.
`G'7(l.2.M'.354
`1991
`
`910’ .285 - dc2fJ
`
`91-3724
`
`5
`
`
`
`Dedicated to the memory of
`
`DAVID S SIMONETT
`
`19264-J0
`
`Duuid Simanett was born in Austruito in 1926. After earttittg a Doctorate at the
`University of Sydney, he became a tending pioneer in the fietti of Remote Sensing,
`hoidirtgfacuity positions at the University of Kansas, the University Ofsydney arid
`the Ut'tilv‘€t'.S'Ety offlattfornia, Stmta Barbara. He was director of tamt use apptications
`at Ettrth .'S't2tL’iiitt.= Corp from 1972 to 1975.
`
`A3 Cimir at Santa Barbara from 1975. he was abie to btttid one ofti-tcftrrentost
`
`Geography pmgmmg in the US, cuin:t'natt‘ng in 1988 with the estabitsittnent oftht’
`Ncttiomrtt Centerfor Geographic Information mid Amttysis. The Stmta Barbara site
`of the Center was renamed the David Stmonett Center for SptttialAt1aiy.s't'.r in I990 in
`re'cogrtt'tion of his rote in its creation‘ He received the Honours A word from the
`A.9.coct'utton ofnrnericun Geogrupttertt and the Victoria Medal from the Roynt
`Geographical Society.
`
`David Simontttt fast a courageous fight against cancer on Decetttber 22. 1990 in
`
`the course of the prepttmtiott of his contribution to this book. The eciitortt t:t'edic'atc'
`tin".-s book to his memory and to the outstanding rote tut has played in the
`deueiopmertt of f.i‘t£’fit.‘i(i ofGeogrupiticui inforrnatiott SystEttt.S.
`
`6
`
`
`
`Preface
`List of contributors
`Ackno wlezigcmcrus
`
`Section I Overview
`
`Introduction
`
`DJ Maguire, M F Good;-hiid and D W Rhind
`
`1. An overview and definition of GIS
`
`D J Maguirc
`
`2. The history of GIS
`J T Coppock and D W Rhind
`
`3. The technological setting of (315
`M FGOOdChii'd
`
`4. The commercial setting of G13
`1 Dangermcmd
`
`5. The government setting of GIS in the United Kingdom
`R Charley and R Buxrori
`
`6. The academic setting of (318
`DJ Unwin
`
`7. The Organizational home for GIS in the scientific
`professional community
`J L Morrixon
`
`91-100
`
`8. A critique of GIS
`R T Acmgeenbrug
`
`Section II Principles
`
`Introduction
`
`M F Goodchiid, D Wflhind and D J Maguire
`
`111-17
`
`7
`
`
`
`VOLUME‘!
`
`:
`
`PRINCIPLES
`
`ta)
`
`9.
`
`10.
`
`IL
`
`Nature of spatial data
`
`Concepts of space and geographical data
`A C Garret‘!
`
`Coordinate systems and map projections for (315
`D H Malirzg
`
`Language issues for (315
`A U Frank and D M Mark
`
`. The error component in spatial data
`N R Chrisman
`
`. Spatial data sources and data problems
`P FFi.9her
`
`. GIS and remote sensing
`F W Davis and D S Simonrtt
`
`(11)
`
`Digital representation
`
`. Computer systems and low-level data structures for GIS
`Wm R Franklin
`
`. High—level spatial data structures for (313
`M J Egenhofer and J R Herring
`
`. GIS data capture hardware and software
`M J Juckson and P A Woodxford
`
`119—34
`
`135-46
`
`l4‘7~63
`
`165-74
`
`175-39
`
`191-213
`
`215-25
`
`227-37
`
`239-49
`
`8
`
`
`
`299-317
`
`. Database management systems
`R G Henley
`
`_ Digital terrain modelling
`R Weibei and M Heller
`
`-20.
`
`Threemjimensional G15
`
`J F Raper and B Kelk
`
`(c)
`
`21.
`
`22.
`
`Functional issues
`
`The functionality of GIS
`D J Maguire and J Dangermond
`
`Information integration and (318
`1 D H Shepherd
`
`. Cartographic modelling
`C D Tomlin
`
`24.
`
`Spatial data integration
`R Flowerdew
`
`. Developing appropriate spatial analysis methods for (315
`S Opiznshuw
`
`26.
`
`27,
`
`Spatial decision support systems
`P J Denrham
`
`Knowledge-based approaches in (315
`T R Smith and Je Yiarig
`
`9
`
`
`
`VOLUIVIE1
`
`:
`
`PRINCIPLES
`
`(cl) Display issues
`
`28. Visualization
`B P Burtenfield and WA Mackaness
`
`29. Computer name placement
`H Freeman
`
`30. Generalization of spatial databases
`J-C Muller
`
`(e) Operational issues
`
`31. C115 specification, evaluation and implementation
`A L Clarke
`
`32. Legal aspects of (315
`E F Epstein
`
`33. Managing an operational GIS: the UK National Dn—Linc:
`Manpower Information System (NOMIS)
`M J Blakemore
`
`34. Spatial data exchange and standardization
`3 C Gupta’!
`
`C0ri.solic1nted bibliography
`List of acronyms
`A uthor index
`Subject index
`
`427-43
`
`445-515
`
`457-75
`
`477-433
`
`489-502
`
`503-13
`
`515-30
`
`531-591
`5 9.3‘-598
`399-613
`615-649
`
`10
`
`
`
`Preface
`List of contributors
`Acknowledgements
`
`Section III Applications
`
`introduction
`
`D W Rhind, D J Mnguire and M F Gootiehiid
`
`(a) National and international GIS programmes
`
`35. A USGS perspective on G13
`L E Starr and K :3.‘ Anderson
`
`36. Development of GIS-related activities at the
`Ordnance Survey
`M Sowtori
`
`37. National (315 programmes in Sweden
`L Ortoson and B Rysredr
`
`38. The development of GIS in Japan
`.5‘ Kubo
`
`39. Land and Geographical Information Systems in Australia
`J F 0'Callaghan and B J Gamer
`
`40.
`
`(318 and developing nations
`D R F Tayior
`
`(b) Socio-economic applications
`
`41. Land information systems
`P F’ Date
`
`42. G15 and utilities
`
`R P Mahoney
`
`43. Car navigation systems
`M White
`
`44. Counting the people: the role of (318
`D W Rhimi
`
`45. C315 and market analysis
`J R Beaumont
`
`11
`
`11
`
`
`
`VOLUME 2
`
`: APPLICATIONS
`
`{cl Environmental applications
`
`46. Soil information systems
`P A Burrough
`
`4?. Integration of geoseientifie data using GIS
`G F Bonhrtm-Carter
`
`153-69
`
`171-34
`
`48. Multisouree, multinational environmental GIS: lessons
`learnt from CURINE
`
`185-200
`
`H M Moimsey.
`
`49. Environmental databases and (315
`J R G Townshend
`
`50. Global databases and their implications for (318
`D M Ciaric, D A Hasn'ng.s' cmn'JJ Kinsman
`
`(d) Management applications
`
`51.
`
`(313 and public policy "
`H W Caikins
`
`52. Urban G13 applications
`R Parrot! and F P Sim:
`
`53. Land resource information systems
`K C Siderelis
`
`54. Land management applications of GIS in the state of
`Minnesota
`A Robin.-ztre
`
`55. GIS in island resource planning: a case study in map
`analysis
`J’ K Berry
`
`56. Integrated planning information systems
`D J Cowen and W L Shirley
`
`Section IV Epilogue
`
`Epilogue
`D W Rhino‘, M FGoodchi!d and D J Maguire
`
`Consolidated bibiiography
`
`201-16
`
`217-31
`
`233-45
`
`247-60
`
`261-73
`
`275-83
`
`285-515
`
`297-311]
`
`313-27
`
`329-389
`
`12
`
`
`
`THE HISTORY OF GIS
`
`J T COPPOCK AND D W RHIND
`
`Contputer-based CH5 have been used since at least the late lEit‘i0s: their mamta!
`]J|'t“t?cit.?C(.’.§'SCtr'S were in use perhaps .100 years earlier. Ackn.owiedgt'ng the paucity of
`well-documented t.’vident.'e, this chapter describes the background to the development
`ofsuch systems, stresting the (‘Oil-felt! in which such development took place, the role
`of organizations and indivicluais where this can be ascertained, and the applications
`which the systems were intended to meet. A broad definition is taken of (its so as
`not to excittde any significant developments,‘ computer mapping s_y.s*tent.y of all types
`(inchtding those with line-printer graphics, the forerunners of contemporary raster
`sytttettttu‘) are included.
`it is demonstrated that most, but by no l'm='tl-"£3 all, of the early developments
`originated in North America. The roles of key organizations such as the US Bureau
`of the Census, the US Geological Survey, the Harvard Laboratory for Comptttcr
`Graphics and the Experimental Cartography Unit are described and the activities of
`the commercial sector are wcempitfled by a case study of En vironmental Systems
`Research tnrtitttte. Reasonls are suggested for significant international differences in
`the ticvetopment ofGt.S. such as the ttttitudetr to owrtership of data and the perceived
`role oflhc tttttlc. It is concluded that several stages of evolution of GIS can be
`defined. These overlap in time and occur at different momertrs in tit'fj“erent parts of
`title world. The first, or pioneering age, extended from the early 1960; to about 1975,-
`in this. individttol personalities were of critical importance in determining what Wd.i'
`achieved. The second phase, approximately from I973 until the em-Iy 1980:, saw a
`regularization ofexperiment and practice within and fostered by national agencies,-
`local cxperirrtertt and action continued ttnlrammelled and duplictztiort of effort was
`common. The third phase, running from about I982 until the late 1980.9, was that of
`Cornmerciai dominance The fbttrth (and current) phase is one of user dominance,
`facilitated by competition among vendors. embryonic standardization on open
`tystettts and increasing agreement on the tt.\'er'.r perception of what :1 G15‘ should do
`and look like.
`
` INTRODUCTION
`
`A variety of information indicates that the field of
`GIS has expanded rapidly in recent years (see
`Maguirt: 1991 in this volume). From where did all
`this business and the resulting jobs arise‘?
`Unhappily, we scarcely know. GIS is a field in
`which history is little more than anecdotal. To
`rectify this. a Search through the archives of
`
`government departments and agencies would
`certainly help. As yet. however. few organizations
`have given any thought to fotrnalizing the history of
`their involvement in GIS and at least one major
`player (Ordnance Survey; soc Finch 198?) has
`refused to let its detailed records be examined by
`external researchers. Less certain1y.tl1e records of
`computer hardware and software companies could
`also be a source of relevant information but no such
`
`13
`
`
`
`have a tradition of writing books or papers on their
`experience of an emerging technology. Research
`staff in government or private sector research
`organizations are exceptions to this rule but, even
`for them, writing papers for the benefit of the
`scientific community at large has a relatively low
`priority. As far as is known, the only official attempt
`anywhere to provide a broad overview of the field as
`a whole is that given by the Report of the
`Committee of Inquiry inl;o the Handling of
`Geographic information (Department of the
`Environment 1987; Rliind and Mounsey 1989}.
`The main source of information. with all the
`
`risks of partisan bias. remains researchers in the
`academic community. In reality, however, even the
`numbers of acatlemies working in this field were
`quite small until the expansion of the last decade.
`Moreover, as Chrisman (1988) and Rhind [1988]
`both testify. those active in universities in this field
`in the early stages of the development of GIS were
`often outside the formal academic career structure
`
`and were so heavily involved in project work that
`they had little time or inclination to write papers. In
`any case, at the beginning there were no obvious
`outlets for publication in a topic that was seen as
`marginal to a large number of interests; Rhind‘s
`(1976) report. for instance, may well be the first
`example of a record of GIS conference papers
`which were described as such in a mainstream
`
`academic publication. While the advent of specialist
`GIS conferences (often disguised by use of other
`titles such as AUTDCARTO) provided one
`publishing mechanism from 1974 onwards. the early
`conference proceedings were intermittent and were
`not easily accessible to those who had not attended
`the gatherings. We do not believe this postulated
`paucity of recorded history represents
`incompetence on our part: a correspondence
`prompted by the editor of Pltorogramtntrrric
`Engineering and Remote Sertsing. for example
`{Marble 1989; Tomlinson 1989}, generated great
`controversy and revealed a lack of documentation
`on the first use of GIS in the refereed literature.
`
`Finally and most crucially. the content of any
`history of GIS depends in large measure on the
`definition adopted. A strict definition. as a
`computer-based system for cu-ialyst'ng spatially
`referenced data. would greatly restrict the held
`
`more general interpretation. as any system for
`handling geographical data. would greatly widen the
`field and hence enlarge the number of contributors.
`Such ti definition would embrace, not only the
`whole held of automation in cartography (which
`was often the precursor to any involvement in GIS
`and provided. in terms of computervgenerated
`graphics. the most common form of output for most
`early systems). but also many general-purpose
`statistical and database packages capable of
`handling x,y.z point data. Formal definitions of GIS
`are not. therefore. of much help and relatively little
`reliance is placed on them in this book as a whole.
`In any event. the field evolved not from some ex
`cntfmdrn definition of the subject but through sets of
`interactions. The main backgrounds of those
`involved have been cartography. computer science.
`geography, surveying. remote sensing. commercial
`data processing, mathematics and statistics. The
`purposes to which the systems have been put
`include environmental protection. urban and
`regional planning, land management, property
`owncrsltip and taxation. resource management. the
`management of utilities, site location. military
`intelligence and tactics. and many others - as later
`chapters in this volume testify. The field has
`developed, then, from a melting pot of concepts.
`ideas. practice. terminology and prejudice brought
`together by people from many different
`backgrounds, interacting with each other often on a
`chance and bilateral basis in the early days and
`normally proceeding in blissful ignorance of what
`was going on elsewhere. The essence of GIS is thus
`its multidisciplinary character, with some at least of
`those involved in developing this technology having
`little: previous involvement, or even interest, in the
`handling of geographical data as such (see Maguire.
`1991 in this volume for further discussion of the
`
`definition of G13).
`This review of the history of GIS is inevitably a
`consequence of the authors’ accidental exposure to
`early developments and their own set of value-
`judgements; different views certainly exist. such as
`that manifested in Cooke's portrayal of the
`genealogical structure of geoproccssing systems in
`general (Fig. 2.1]. In particular, it is suspected that
`the role of those who did not contribute to the
`
`formal literature has been unclcrplayed. especially
`
`14
`
`22
`
`14
`
`
`
`THE GRASS ROOTS EVOLUTION OF GIS
`
`What seems clear is that there were many
`initiatives, usually occurring independently and
`often in ignorance of each other, concerned with
`different facets of the field and frequently
`originating in the interests. often disparate. of
`particular individuals. Like the reality (as opposed
`to the reporting) of scientific research, there was no
`strictly logical progression towards the development
`and implementation of (318, but rather a mixture of
`failures. set-backs. diversions and successes.
`
`inevitably, more is known about the successes than
`about the failures which. according to both
`Dangermond and Smith (1988) and Tomlinson
`(1988). have been numerous and often attributable
`to bad advice, ignorance and a determination to go
`it alone. This is unfortunate because "failures are
`
`often as illuminating as successes, if not more so
`(Giles 1987). What also seems clear is that
`particular individuals and institutions played key
`roles. acting as examples or as sources of expertise,
`advice and often skilled personnel; since these
`contributions are now better recorded than is the
`
`generality of progress, this account will tend to
`emphasize them, particularly those of Howard
`Fisher in the Harvard Laboratory for Computer
`Graphics (LCG), Roger Tomlinson in the Canada
`Cieographic Information System (CGIS) and Jack
`Dangermond in the Environmental Systems
`Research Institute (ESRI) in North America. and
`David P. Bicltmore at the Experimental
`Cartography Unit (ECU) in the United Kingdom.
`Many others played significant parts (c.g. Tobler
`1959; Nordbeck 1962; Cook 1966; Hagerstrand
`1967: Diello, Kirk and Callander 1969 and Boyle
`(sec Rhind 1988)). but these four have been the
`subject of particular articles in a special and
`invaluable issue of The /lrrtericttrr Crtrtogrrtpher
`(Tomlinson and Petchenilt 1938). Fortunately, these
`individuals seem to typify the interests. attitudes
`and commitments of those working in the vintage
`era of GIS from the late 1950s to the end ofthe
`19703.
`
`The motivations for developing GIS or
`
`15
`
`new sources otdata or techniques, through the
`desire for greater speed or efficiency in the conduct
`of operations on spatially referenced data, to the
`realization that desirable tasks could be undertaken
`
`in no other way. The last was undoubtedly a
`powerful motive in two key developments which are
`discussed in more detail below — the Oxford System
`of Automated Cartography and the Canada
`Geographic information System. It was the
`experience of publishing the Atlas of Great Hrimin
`and Northern lrelnr-id (Bickmore and Shaw H63)
`and the criticisms this attracted of being out of date
`and unwieldy that convinced D. F. Bickmorc,
`probably in 1958 but certainly no later than 1960,
`that only the computer could provide a cost-
`cffcctive mechanism to check. edit and classify data,
`to model situations and to facilitate experiments in
`graphic display {Rhind 1988). Similarly. it was the
`impossibility of analysing maps of East Africa at an
`acceptable cost that first led R. Tomlinson (.1988) to
`think of a digital approach. A calculation made in
`1965 indicated the need for some $Can a million in
`
`1965 prices and a requirement for 556 technicians
`for three years in order to overlay the Z1 :5(Jll(lU scale
`maps of the Canada Land Inventory; this
`unacceptable level of resources acted as an
`incentive to develop a more automated approach.
`lt was, of course, the advent of the digital
`computer and the order-of-magnitude decrease in
`computing costs every six years over a 30-year
`period (Simonett 1988} that made such alternative
`digitally based approaches viable. It is interesting to
`note, however. that not all early work used the
`digital computer. Thus perhaps the earliest attempt
`to automate map production, the preparation of the
`Artrnr oftlre Britta}: Flora, employed a modified
`punch card tabulator to produce maps on pre-
`printed paper from cards on which had been
`punched the grid references of recorded
`occurrences (Perring and Walters 1962). Altltough
`this approach was not repeated and Perring (1964)
`later recognized that the analysis of voluminous
`data could more easily be undertaken by computer.
`it anticipated the widespread mapping in the late
`1960s byline printer. it is also interesting to note
`that Perring was a botanist. with no training in
`cartography. who was faced with the taslt of
`providing Ztlllil maps from data that had been
`
`15
`
`
`
`
`
`
`Addrefis
`§3L?.;"n““(‘l€i§)
`Z|P+4, etc.
`
`?
`
`MAPMASTER
`
`of Oregon
`
` University
`
`Mal’1'§'§’E'3%de'S
`
`Polygon
`Intersection &
`Overlay System
`
`PC Arc—|nfo
`
`Synercom
`
`
`
`AFC-W0
`1930
`
`Morehouse
`
` Automap
`
`1970
`
`Dangermond
`
`
`
`Group
`International
`1981
`
`
`
`Carter
`
`ESRI
`1959
`
`
`
`lntergraph
`1969
`
`Sinton
`
`Odyssey
`Mid—70's
`
`OMB,
`
`
`
`
`Arithmicon
`1973
`
`ETAK
`Map Engine
`
`Schweitzer
`
`
`
`
`National
`
`Decision
`Systems
`
`Sandiego
`council
`of Govern-
`ments
`
`University 0
`Washington
`Seattle
`
`51 1
`13a"3—“‘4"
`
`n orma 1
`. tam
`Display System
`(DIDS) 1982
`
`,
`.::3s?:,:::,
`System (DIDS)
`1978
`
`Whitehouse
`
`
`
`Decision
`Resources
`1962
`
`
`
`MAP-INFO
`1987
`
`Hand
`
`Rand Map
`1985
`
`Videodisk
`Product
`
`PC-MAP
`
`Johnson
`
`Fig. 2.1 An individual perception of the genealogy of geoprocessing in the United States (Pers. Comm.
`Don Cooke, 1990). Circles are ‘places’, i.e. companies, government agencies, universities, etc.; rectangles
`are ideas or concepts, often embodied in a software package or database; directed lines show direct or
`indirect migration or influence in a number of different ways. Examples of flows or lack of expected ones
`include:
`
`16
`
`24
`
`16
`
`
`
`Roadshow
`1937
`
`Matchmaker
`1935
`
`Street Address
`Matchin
`
`System 1 B9
` DPMAP
`
` rban Data
`
`Processing
`Inc. 1963
`
`
`
`New
`Haven
`Census use
`
`
`
`Transport
`-‘Survey
`Coding Guide
`
`
`Census
`Ti'l-atate
`SCRIS
`Bureau
`Trnnaporl
`'69-'71
`
`Unirnatch
`1970
`
`
`MAPCJI Gri Spit
`
`1955
`
`1971
`
`Tape
`
`?
`
`e Donald Cooke was
`
`DI
`Line
`
`ital
`raph
`
`
`
`
`
`
`U.S.
`Geological
`3“"""V
`
`0 Harvard Labs influence on (315 vendors (Morehnuse tn ESRI, Simon to Intergraph; Odyssey tn
`Synereom)
`
`0 DIME was independent from the SACS (Small Area Census Studies)
`
`9 the diagram suggests that the USGS and the US Postal Service had very little inlluence on most
`developments.
`
`17
`
`17
`
`
`
`driven by an urgent need ofthe users. that such a
`task would have to take advantage of the best
`available technology - whatever its limitations —
`rather than await the ideal solution; it was also
`similar to many later applications in that it was a
`‘one-off‘ development which, having served its
`purpose. was not taken any further. Slightly later
`work (around 1967) by Berlin in Paris involved the
`modification of IBM ‘golfball‘ typewriters driven
`directly by punch card readers to produce
`proportional symbol maps.
`It is also clear that it was in North America that
`
`most of the significant early developments in. and
`applications of, GIS and related technology were
`made. By the early 1980s, Tornlinson (1985)
`estimated that there were probably more than l(ll.'ll'l
`systems in North America, :1 figure that must have
`represented a very high proportion of the systems
`then existing in the world as a whole. The bulk of
`this account will accordingly focus on North
`America, with later references to the United
`Kingdom and other European countries and to
`developments elsewhere in the developed world. It
`is only in the late 1980:: that any significant
`developments have occurred in developing
`countries and then often through the aid and
`encouragement of developed countries {see Taylor
`1991 in this volume).
`
`THE NORTH AMERICAN SCENE
`
`Aangecnbrug (pers. comm. 1990) has argued that
`the earliest antecedents of GIS in the United States
`
`can be traced back to the University of Washington.
`In the 19505, both geographers (notably Garrison)
`and transportation engineers (notably Horwood)
`developed quantitative methods in transportation
`studies. Garrison‘s colleagues and students included
`Berry. Tobler and Marble; Horwoud’s included
`Barb and Dueker (see Di1elter’s important 1974
`papcr). Much of the original leadership of the
`Urban and Regional Information Systems
`Association (founded in H63) and that of other key
`bodies was derived from or directly influenced by
`this group.
`By the early 1960s. at least in North America.
`
`and a memory 32 times as great as its predecessor.
`the IBM 1401 (Tomlinson 1985}. These machines
`were employed primarily for one of two very
`different purposes: for routine administrative and
`data management tasks in business and government
`(such as pay—roll. stock control and record keeping
`of various kinds) and for scientific applications
`inyolving extensive computations, notably in
`chemistry, mathematics and physics. There was
`inevitably a good deal of discussion in government
`departments and agencies about the possibility of
`applying computer technology to handle numerical
`data. especially where these were already in
`machine-readable form. as with many censuses.
`where punch-card technology was widely used. In
`1965 the US Bureau ofthe Budget compiled an
`inventory of automatic data processing in the
`Federal Government. in which it noted the
`
`significant use of computers to handle land use and
`land title data (Cook and Kennedy 1966). The
`following year, a conference on a comprehensive
`unified land system at the University of Cincinnati
`was advised that a system must be designed such
`that it obtained the maximum benefit from
`
`electronic data processing equipment (Cook 1966).
`The conference also heard that the District of
`
`Columbia already had a property data bank. which
`could be searched, updated and retrieved, and that
`Nassau County in New York would be the lirst to
`provide fully-automated access to records of land
`ownership.
`The significance of the developments at the US
`Bureau of the Census. stemming directly from its
`need for automated address matching. is difticult to
`overemphasize. This need arose from the
`predominantly mail outfmail back nature of the US
`census and the requirement to produce area based
`tabulations from records whose only geographical
`reference was the postal address. An early advisory
`committee on small area data included Garrison
`
`(see above). who urged in development project to
`test automated data linkage procedures, A director
`hired to run the test. Caby Smith. recruited a team
`which included Corbett, Cooke. Maxfielcl. White.
`Farnsworlh, Jaro. Broome and others who appear
`elsewhere in these pages. The first demonstrations
`of address matching, computer mapping and small
`area data analysis were provided through the 1967
`
`18
`
`26
`
`18
`
`
`
`development and widespread distribution of
`ADMATCH (address matching software) all had
`major impacts upon government and academia in
`the United States. Indeed, the Census Use Study
`also sponsored the First International DIME
`Colloquium in 1972, leading to the creation of the
`Segment (later re-named as the Spatially}
`Oi-ientatcd Referencing Systems Association (or
`SORSA), an organization which still holds
`international conferences.
`
`increasing availability of computers in
`universities was undoubtedly instrumental in the
`development of the quantitative revolution in
`academic geography in the early 19605 (James and
`Martin 1973; Hudson 1979), particularly in the field
`of spatial analysis (a term which was in general use
`by the late 1960s — sec Berry and Marble 1968), with
`its emphasis on the statistical treatment of
`geographical data and on modelling. However,
`these applications, despite their potential relevance
`to handling geographical data had little interaction
`with computer mapping, primarily because the
`statistical methodology was largely aspatial. One
`exception is a paper in an edited collection on
`computers in geography which related modelling to
`a crude cartography using the line printer (Rushton
`1969). it is only in the middle and late 1980s that
`successful attempts have been made to develop
`closely coupled spatial statistics and ‘geographical‘
`displays.
`Computers in the 1960s had, in general. no
`explicitly graphical facilities. usually operated in
`batch mode and were very expensive by today's
`standards. Despite this. Tohler (1959) had early
`recognized their potential for automating
`cartography, as had Nordbeck (1962) in Sweden.
`There were, indeed, developments in autotnatirtg
`cartography in several national agencies concerned
`with mapping and in military establishments which
`could afford equipment that was prohibitively
`expensive to others. The US National Ocean Survey
`was creating charts on a Gerber plotter for the
`production of ‘figure fields‘ or matrices of depth
`values and such organizations as the Aeronautical
`Charting and Information Center at St Louis, the
`Rome Air Development Center and the Central
`Intelligence Agency were active in aspects of this
`field (Diello, Kirk and Cal1enderl968;Ton‘t|inSott
`
`Hydrographic Survey had automated display
`facilities in operation and Surveys and Mapping had
`embarked on a programme to apply automated
`cartography to the l :5[)tl0tl series in Canada. In the
`main, however, the aim in computer applications in
`national mapping agencies was to mimic manual
`methods of production and so to product: maps that
`were virtually indistinguishable from their manual
`counterparts. Little information appears to be
`available on the extent to which these methods were
`
`Cost effective, although Tonilinson (1935) suggests
`that the high cost of hardware placed them at a
`disadvantage in competition with manual systems:
`continuing evaluations of costs by the Ordnance
`Survey in Britain, for example, did not find
`automated approaches to map production as a
`whole to be cost effective until the ltiltills. Unlike the
`
`situation in Britain, where at digitizing production
`line was in operation from 1973, the Topographic
`Division of the United States Geological Survey did
`not implement plans to automate the production of
`topographic maps until the start of the 19805 — a
`severe handicap to the development of rnany
`gcographit:ally—hased information systems in the
`United States.
`
`An entirely different approach to the
`automation of cartography was adopted elsewhere,
`notably in the universities, using the standard line
`printer as a mapping device. In cartographic terms,
`the results were crude, but this was not the point:
`the aim was to produce maps quickly and cheaply so
`as to display the characteristics of the data
`(especially statistical data for census tracts and the
`like) and to undertake simple analyses of such data
`by relating different parameters. It was here that
`Howard Fisher made a sig'ni.l"it:an1 contribution and
`this approach found ready applications in landscape
`design, in urban and regional planning and, to a
`lesser extent, in resource man agcmcnt.
`
`The Harvard Laboratory for Computer Graphics
`
`Fisher was not a cattographer but trained and
`practisctl as an architect, He had begun work on a
`computer mapping system in 1963 while at the
`North Western Technical Institute (Schmidt and
`
`19
`
`19
`
`
`
`approaches to Chicago and Northwestern
`Universities (both strongholds of non-spatial
`computer applications to the analysis of
`geographical data), established the Laboratory for
`Computer Graphics (a title subsequently
`lengthened by the addition of ‘anti Spatial
`Analysis‘) in 1965 in the Graduate School of Design
`at Harvard University — from which he himself had
`gratluatetl. ‘There he built up a team of
`programmers and others to create a mapping
`package (SYMAP) which used the line printer as a
`mapping device and was capable of producing
`isoline, choroplcth and proximal (Thiessen polygon
`or Dirichlet tessellation) maps. The package was
`easy to use by the standards of the day, particularly
`in relation to data for census tracts, incorporated
`default options when nothing was speci lied by users
`and was widely distributed. In addition to many
`pirated copies. over 500 institutions acquired
`SYMAP [Schmidt and Zatft l975:Chrisn1an I988):
`half of these were in universities, with the
`
`remainder equally divided between government
`agencies and private institutions. Copies were
`acquired not only in North America but also in
`Europe and elsewhere and the manual was
`translated into several languages, including
`Japanese. A subsequent program. CALFORM,
`which produced higher quality choropleth maps by
`pen plotter and rellected the increasing (if still
`sparse) availability of these plotters, seems to have
`had less success although it too was a pioneering
`effort. SYMAP was important as the first widely
`distributed computer package for handling
`geographical data. It introduced large numbers of
`users to the possibilities of computer mapping: it
`was the precursor. and ppssibly the progenitor. of a
`large number of other programs using the line
`printer; and it found a wide range of applications
`particularly through the connection between the
`I-larvard Laboratory and landscape architects in the
`Graduate School of Design. notably C. Steinitz and
`his associates — one of whom, D. Sintori, produced a
`cell-based program (GRID) which permitted
`multiple o