Twenty Years of Progress in GIScience Michael F. Goodchild University of California Santa Barbara

  • Published on
    13-Jan-2016

  • View
    219

  • Download
    0

Embed Size (px)

Transcript

<p>Twenty Years of Progress: GIScience in 2010</p> <p>Twenty Years of Progress in GIScienceMichael F. GoodchildUniversity of CaliforniaSanta Barbara</p> <p>Michael F. Goodchild1OutlineThe beginnings: GISThe beginnings: GIScienceMajor accomplishmentsresearchinstitutionalThe future</p> <p>--2A short history of GISMaps in computersfor decision-makingeach map representing one dimension of a decisionfor managing dataaggregating census returns to reporting zonesmanaging the multiple data types of transportation planningto support map-makingeditingprojection change</p> <p>3A model for landscape architectureIan McHargs school at the University of Pennsylvania</p> <p>Ian McHarg 1920-2001MeteorologyGeologyHydrologyPlant ecologyAnimal ecologyLimnologyComputationRemote sensingIan McHarg4</p> <p>5For the first time, a department of landscape architecture could recruit a faculty of distinguished natural scientists sharing the ecological view and determined to integrate their perceptions into a holistic discipline applied to the solution of contemporary problems.I.L. McHarg, A Quest for Life (Wiley, 1996, p. 192) Integration of science into action Frequently emulated as a model for environmental science But with a weaker intervention component The social context is missing Computation and remote sensing do not fit the modelI.L. McHargWiley, 1996, p. 1926The Canada Geographic Information SystemRoger TomlinsonIBM contracts 1964-687 layers of land characteristicssoil capability for agriculturerecreation capabilitycurrent land use.To assess the current use of Canadian landto measure area, plan new usesRoger TomlinsonIBM1964-19687</p> <p>Technical aspects of CGISManuscript maps at 1:50,0007 per tileHand-scribing of boundariesAn optical scanner creating a raster of boundariesVectorizationMerging with area attributesThe common boundary between two areas as the basic unitCGIS1500007</p> <p>9Flat-file options (tape)By face/polygondouble recording of internal boundariesspurious differencesBy edge/archalf the data volumecompute area in O(vertices)simplify overlayattributes of adjacent polygonsno polygon records / / O</p> <p>10</p> <p>Technical aspectsStorage on magnetic tapevariable-length recordsleftpolyID, rightpolyID, #points, (x1,y1),Indexing in Morton ordera quad-tree indexNumerical output onlytabulations of areano visual displayMainframe technologylater leased land lines at 300 bpsIDID#(x1,y1),Morton300bps14The quadtreeRecursive subdivisionvariable depth depending on local detail303132331023- 15Other types of mapsTransportation linkslinear featuresnetworksU.S. Bureau of the Censusblocks = 2-cellsstreet segments = 1-cellsintersections = 0-cells</p> <p> = 2 = 1 = 016</p> <p>Topological data structures1977 conferencesponsored by Harvard UniversityA unifying structure across many application areasall three of: decision-making, managing data, editing mapsThe birth of Esri1977Esri18The relational modelThe map as a collection of arcs, nodes, and facesF-A+N = 2Stored in tables with keysGIS built on RDBMSINFOVertices left outa hybrid solutionARC/INFOthe ARC data structure still proprietaryF-A+N = 2RDBMSGISINFO/INFO19Square pegs in round holesCul-de-sacsallow 1-nodesProperties of parts of edgesdynamic segmentationlinear referencingNon-planarityoverpasses and underpassesturntables20A 1990s house of cardsStill no vertices in the RDBMSPointscoordinates stored in tablesno topological relationships with other featuresDoes it have to be this hard?simple CAD data modelpoints, lines, and areas in an empty spacepotentially overlappingno topological relationshipscompute on the fly1990RDBMSCAD21</p> <p>Object-oriented data modelingAll features are instances of classesClasses inherit properties from more general classesFeatures can be aggregates of other featuresFeatures can be composed of other featuresFeatures can be associated</p> <p>24</p> <p>25</p> <p>26</p> <p>ArcGIS27</p> <p>AddressAgricultureArchivingAtmosphericBasemapBiodiversityCensus-Administrative BoundariesDefense-IntelEnergy UtilitiesEnergy Utilities - MultiSpeak TMEnvironmental Regulated FacilitiesForestryGeologyGroundwaterHealthHistoric Preservation and ArchaeologyHydroInternational Hydrographic Organization (IHO) S-57 for ENCLand ParcelsLocal GovernmentMarine PetroleumPipelineRaster TelecommunicationsTransportationWater Utilities </p> <p>-MultiSpeak TMIHOS-57,ENC</p> <p>31A paradigm shiftAway from the map metaphorgeoreferenced events, transactionsobjects with no georeferencesphenomena that were never mappedNeogeographycustomized mapsuser-centrictransitoryInteractions, flows</p> <p>32</p> <p>****0..10..10..20..20..1ORIGINAL USE CASE MODELSINTERACTIONMINARD NAPOLEON MAPKARST FLOW ROUTES36</p> <p>**0..10..2</p> <p>0..1 Generic Flow Model </p> <p>slide 19 / 22</p> <p>slide 15 / 22SDH 4, Zrich 1990Spatial information scienceWhat, after all, is spatial data handling? It may describe what we do, but it gives no sense of why we do it. It suggests that spatial data is (sic) somehow difficult to handle, but will that always be so? It suggests a level of detachment from the data themselves, as if the USGS were to send out tapes labeled with the generic warning handle with difficultyWe are concerned with much more than the mere handling and processing of data. We are more than the UPS of GIS.(Proceedings p.3)SDH 4, Zrich 1990USGSGISUPS P342GIScienceSecond European GIS Conference 1991fleshes out the research agendareference to geographic information scienceRapid progress was made on algorithms and data structures in the 1970s and 1980s, but many of the hard problems of data modeling, error modeling, integration of spatial analysis, and institutional and managerial issues remain. Some of these may be unsolvable for example, there may simply be no generalities to be discovered (about) the process of adoption of GIS by government agencies, however easy it may be to pose the research question.(Proceedings pp. 342-350)GIS1991- 19701980GIS pp342-35043G or S?A play on the acronymsystems, science, services, studiesWould discoveries about geographic space apply to all spaces?(1992) Geographical information science. International Journal of Geographical Information Systems 6(1): 3145.G S19926(1): 314544The agenda in 1990The content of GIScience (IJGIS 1992)Data collection and measurementData captureSpatial statisticsData modeling and theories of spatial dataData structures, algorithms and processesDisplayAnalytical toolsInstitutional, managerial, and ethical issues1990IJGIS 199245Other listsNCGIA research agenda (IJGIS 1989)UCGIS research agendas (from 1997)Tests of inclusionproblem is not yet solvedtruths remain to be discoveredproblem is genericproblem is hardproblem would be recognized by a scientist skilled in the artThe Varenius frameworkthe humanthe computersociety</p> <p>NCGIAIJGIS 1989UCGIS1997Varenius </p> <p>46Definitions of GIScienceDavid Mark in Foundations of Geographic Information Science, Taylor and Francis, 2003the development and use of theories, methods, and data for understanding geographic processes, relationships and patterns (UCGIS, 2002)the basic research field that seeks to redefine geographic concepts and their use in the context of geographic information systems (Mark, 2000)David Mark, Taylor and Francis, 2003- (UCGIS, 2002)- (Mark, 2000)</p> <p>47Major accomplishments in researchInternational Symposium on Geographic Information Science20th anniversary of the funding of NCGIASanta Barbara, Dec 11-12, 200849 participantshttp://ncgia.ucsb.edu/projects/isgis/Ten most significant discoveries?can GIScience be empirical?NCGIA20Santa Barbara, Dec 11-12, 200849http://ncgia.ucsb.edu/projects/isgis/10- 48Kate Beard, University of MaineSpecification of spatial data types: object and object-relational databasesSpecification of spatial relationsConditional simulationLocal spatial statistics: local autocorrelation, GWRUser interfaceGeographic brushingStandardization: common formats, specificationsDorling cartogramsGeneralization as constrained optimizationGoogle EarthKate BeardGWRDorlingGoogle49Marc Armstrong, University of IowaGIScience is transformationalfrom map to machinebiggest discovery is GIScience itselfAbstraction/theoryTransformationTopological conceptsHierarchical data structuresOntologiesGeocodingOverlay and other manipulationsLocal spatial analysis</p> <p>Marc Armstrong/50</p> <p>UZH51</p> <p>- </p> <p>52</p> <p>/9- /VGILBS,/CA53</p> <p>///54</p> <p>/55</p> <p>Classics from IJGIS ed. Fisher (2007)19 prominent articles based on citationall basic research, one per researchereven distribution over 20 years</p> <p>IJGIS ed. Fisher20071920</p> <p>57Openshaw, Charlton, Wymer, Craft: A Mark I Geographical Analysis Machine for the Automated Analysis of Point Data Sets (1987)Brassel and Weibel: A Review and Conceptual Framework of Automated Map Generalization (1988)Heuvelink, Burrough, Stein: Propagation of Errors in Spatial Modelling with GIS (1989)Skidmore: A Comparison of Techniques for Calculating Gradient and Aspect from a Gridded Digital Elevation Model (1989)Worboys, Hearnshaw, Maguire: Object-Oriented Data Modelling for Spatial Databases (1990)Egenhofer and Franzosa: Point-Set Topological Spatial Relations (1991)Miller: Modelling Accessibility Using Space-Time Prism Concepts within Geographical Information Systems (1991)Goodchild: Geographical Information Science (1992)Fisher: Algorithm and Implementation Uncertainty in Viewshed Analysis (1993)Raper and Livingstone: Development of a Geomorphological Spatial Model Using Object-Oriented Design (1995)Jankowski: Integrating Geographical Information Systems and Multiple Criteria Decision-Making Methods (1995)Fotheringham, Charlton, Brunsdon: The Geography of Parameter Space: An Investigation of Spatial Non-Stationarity (1996)Frank: Qualitative Spatial Reasoning: Cardinal Directions as an Example (1996)Kiiveri: Assessing, Representing, and Transmitting Positional Uncertainty in Maps (1997)Clarke and Gaydos: Loose-Coupling a Cellular Automaton Model and GIS: Long-Term Urban Growth Prediction for San Francisco and Washington/Baltimore (1998)Bishr: Overcoming the Semantic and Other Barriers to GIS Interoperability (1998)Andrienko and Andrienko: Interactive Maps for Visual Data Exploration (1999)Smith and Mark: Geographical Categories: An Ontological Investigation (2001)Llobera: Extending GIS-Based Visual Analysis: The Concept of Visualscapes (2003)</p> <p>Andre Skupin, San Diego State UniversityMy own (highly subjective) listTheories of representationdiscrete objects and continuous fieldsobject fields, metamapsunificationModels of uncertaintyerror propagationdownscalingPrinciples of spatial cognitionTheories of the geographic worldspatial dependence, spatial heterogeneitymetamaps- </p> <p>60Institutional accomplishmentsJournals and articlesIJGIS, CaGIS, J GISciencesConferences and societiesBooks and curricula4 members of the US National Academy of Sciences4 members of the Royal SocietyIJGIS, CaGIS, J GISciencesUS61What of the future?Max Egenhofers list for 2010:Spatial cognition about geographic space and systemsSpatial semantics for information systemsA general theory of geographic space and timeSpatial communicationSocietal issues of spatial information and spatial systemsMax Egenhofer 201062Future prospectsKnowing where everything is (at all times)every mobile phoneevery vehicleevery farm animalevery item in a storeevery construction beamevery asset for emergency responseevery victim of a disaster63The role of the citizenPlacenames, streets, social characteristicsEarly notification of changeEarly reports of damage from a disasterBoth producer and consumer of geographic informationThe local expert64A technology of dynamicsReal-time, continuous monitoringThe state of the world at all timesthe state of the transportation networkthe state of human healththe state of the environmentSensor networksstaticcarried on moving objectshumans as sensors---</p> <p>65The third (and fourth) dimensionThe third spatial dimensiondata acquisitionpositioningThe conceptual problemthe map as metaphorThe attribute dimension</p> <p>66Location as common keyThe stack of layers</p> <p>67 But in realitySpatial databases are organized as layershorizontal integration not verticalproperty z about all places rather than all properties about location xtell me everything about location xoverlay must be invoked explicitlygraphical overlay or topological overlaymany mashups are merely graphical overlaya visual spatial joinzX68</p> <p>The spatial joinUsing location as a common key to link tablesAll location references are subject to uncertaintymeasurement errorvagueness in feature identificationindeterminate limitsThe probabilistic join73The challenge of educationAs the technology becomes easier to useas everyone utilizes geospatial technologyWhat does everyone need to know?Critical spatial thinkingan understanding of the fundamental concepts behind the technologyWhat characterizes a spatial thinker?74A call for participationTechnology will continue to advance and pose interesting and challenging research questionsThe domain of geographic information science is well defined and boundedComparisons with other spaces will stimulate creative thinking in GIScienceMuch has been accomplished, but much more remains to be discovered and developed75</p>