30 YEARS SINCE Z_GIS FOUNDING: PROVIDING SOLUTIONS FOR SOCIETY AND BUSINESS

Transcription

1 30 YEARS SINCE Z_GIS FOUNDING: PROVIDING SOLUTIONS FOR SOCIETY AND BUSINESS Michael F Goodchild Emeritus Professor of Geography University of California, Santa Barbara Keynote GIScience Symposium, Salzburg, October 25, 2018 GISCIENCE AND GEOGRAPHY: AN EVOLVING RELATIONSHIP' Abstract Geographers were quick to recognize the usefulness of GIS when the technology first began to appear in the 1970s, and geography rapidly established itself as the academic home of GIS. By the 1990s, GIS was being taught and used across academia, and a strong critique of GIS had developed within the discipline. Two responses led to a second quite different relationship: geography as the leading home of theory, in the form of GIScience, and geography as the home of critical perspectives on GIS. Today the rapid growth of data science, the consumerization of GIS, and the continued spread of GIS and cartographic tools across the disciplines from engineering to the humanities is prompting a third debate about the relationship between GIS and geography. The presentation ends with a sketch of the form that new relationship might take.

2 GIScience and Geography: An Evolving Relationship Michael F. Goodchild University of California Santa Barbara Dr Roger Tomlinson,

3 The Canada Land Inventory Land in 1955 was underutilized How much land is being used for x and might be used for y? 7 map layers current land use land capability for agriculture, forestry, recreation, 50,000 map sheets How to measure area on combinations of map sheets? 2

4 Put the maps in a computer Where simple algorithms can overlay layers and measure areas precisely and output the required statistics In 1965? with computers that have less power than your smart phone where all data must be stored on magnetic tape when scanners to input map sheets did not exist where algorithms had to be invented from scratch 3

5 The CGIS vision Multiple flat layers of data represented as boundary networks forming nonoverlapping polygons no allowances for uncertainty Overlay to combine layers All output statistical Enlarging the vision Meetings of the IGU Commission on Data Sensing and Processing 1970, 1972 A system capable of capturing, storing, analyzing, visualizing, archiving, sharing any kind of geographic information any information about what is where By 1980 many commercial companies engaged in producing and selling GIS software engaged in teaching and researching GIS 4

6 What of geography? Roger Tomlinson: PhD Geography UCL Many early proponents of GIS were geographers seeing it as immensely valuable to geographic research as a way of implementing geographic techniques as a way of sharing geographic knowledge as a language for the discipline as a stimulus for the spatial analytic tradition as a student path to employment Yet early GIS was far from a holistic way of representing geographic information The layer Data integrated into horizontal layers maintaining the conceptual link to maps we could slice the 3D cube (or 4D hypercube) in many other ways 5

7 Maps or globes? Maps: easy to print, store, ship, any scale Globes: fixed scale, hard to print, store, ship but digital globes are easy to store and ship, and readily zoomed and digital globes such as Google Earth were not even imagined in 1965 Tomlinson s choice was the right one at the time but is it still the right one? pyxisinnovation.com / globalgridsystems.com 6

8 The critique (circa 1990) Brian Harley maps are social constructions, what do they reveal about their makers? Denis Wood, Whose agenda is in your glove compartment? Neil Smith GIS ignores its strong roots in military and intelligence technology Should GIS researchers be held responsible for its misuse? GIS empowers only the powerful GIS data models impose unwarranted simplicity and thus bias on the world Should GIS be taught in universities? It s just a tool GIS is non-intellectual expertise We don t teach Microsoft Office so why GIS? It s an expensive use of the limited resources of geography departments Can t they just take Esri s training courses? 7

9 GIScience (1992) A theoretical framework for GIS responding to some of the critiques and acknowledging that others were well grounded A set of principles implemented in GIS some theoretical, some empirical all useful why spatial is special Re-asserting that geography has a unique relationship to GIS 8

10 9

11 Some principles The results of any spatial analysis change when the units of analysis change cancer and counties the Modifiable Areal Unit Problem Generalization from one geographic area to another is always problematic spatial heterogeneity 10

12 Tobler s First Law of Geography Nearby things are more similar than distant things the principle behind spatial interpolation, contour mapping, regions Anomalies are interesting hot spots, cold spots high surrounded by low, low surrounded by high Geographic information will always lead us to question the basic assumptions of inferential statistics 11

13 Validity Nearby things are less similar than distant things negative spatial autocorrelation possible at certain scales the checkerboard retailing but negative a/c at one scale requires positive a/c at other scales In the geographic world smoothing processes dominate sharpening processes One more principle The Earth is not flat flatten the Earth to make maps and databases and ignore this principle at your peril! 12

14 Pixels at the Equator 50km by 50km Pixels at 85 degrees North 50km north/south by 4.4km east-west 13

15 ArcMap 10.0, Plate Carrée projection The changing landscape Big data finer spatial and temporal resolution new interest in dynamics new variables, methods faster computation new questions Consumerization crowdsourcing neogeography places over spaces 14

16 Big data volumes in GIScience? Landsat since m pixel for TM, 8 bits or 1 byte per band 500 x sq m of Earth surface 500 x 10 9 pixels 4 bands 2 x bytes per Earth coverage 1 coverage every 19 days 20 coverages per year = 4x10 13 bytes per year for 40 years 1 petabyte from Landsat TM 30 15

17 How has science dealt with massive volumes of geospatial data? Analysis of Landsat download, analyze each scene separately hope that your area of interest lies entirely within one scene hope that interactions of interest don t extend across scene boundaries Divide and conquer a classic approach to dealing with data volume Downside of divide and conquer Missing long-distance effects that span scenes How to generalize from analysis of one scene? spatial heterogeneity 16

18 Transitory, for one user only right now, delivered in under a second Synthesized from numerous inputs including crowdsourcing Point-of-interest, street centerline databases Traditional language mapping David Rumsey Map Collection, 17

19 Language mapping circa 2018 UCL Consumer Data Research Center The Fourth Paradigm Let the data speak for themselves data-driven science But geospatial data are never geography it is impossible to measure location perfectly all geospatial data are subject to a myriad of uncertainties, some are not replicable (e.g., CGIS) generalized, abstracted, synthesized detail is always truncated When the data speak for themselves they do not always speak for geography 18

20 Johannes Vermeer, The Geographer, Städelsches Kunstinstitut museum, Frankfurt The map is not the territory Alfred Korzybski,

21 Geography as ground truth It is not the geographer who goes out to count the towns, the rivers, the mountains, the seas, the oceans, and the deserts. The geographer is much too important to go loafing about. He does not leave his desk. But he receives the explorers in his study. He asks them questions, and he notes down what they recall of their travels. Antoine St-Exupéry, The Little Prince The user of GIS cannot know what is missing from the data without explicit information about uncertainty without access to reality (ground truth) without the wisdom of geographic knowledge Where does that leave (spatial) data science? Analysis of data without knowledge of ground truth are the results correct, or are they artifacts of the uncertain data? is it possible to do GIS without geography? without appreciating that spatial is special? 20

22 Courtesy Richard Church GIS: more than making a map Providing access to context horizontal: what else is in the neighborhood? vertical: what else is known about that location? Spatial analysis using techniques to reveal what is not otherwise visible Spatial modeling simulating the action of landscape-modifying processes Spatial prediction forecasting where as well as when 21

23 GIS and Geography today Geospatial technology is everywhere the sciences, humanities, the general public The acronym GIS is no longer all-encompassing many geospatial experts may not recognize it the 4 th Edition of Longley et al. focuses on GI not GIS Yet Geography as a discipline has a vital role to play the principles that make spatial special the value of a spatial perspective the importance of integrating data the importance of ground truth A robust future for GIS in Geography Build bridges to programs in data science that will otherwise treat geospatial as just another application Focus on the principles of GIS which will still be true in 20 years which make the case that spatial is special Take advantage of the new opportunities new sensors, social media, real-time data spatial prediction new levels of spatial and temporal resolution high-performance computing 22