Geographic Information Systems: Theory and …

1 Geographic Information Systems: Theory and ApplicationTim SchoenharlNotre Dame CSEA pril 11, 2007 Tim Schoenharl (Notre Dame CSE) Geographic Information Systems: Theory and ApplicationApril 11, 20071 / 16 Overview1 Motivation2 Ellipsoids, Geoids, Datum, Coordinates and Projections3 GIS Software Systems4 GIS and Relational Databases5 Spatial Analysis6 ReferencesTim Schoenharl (Notre Dame CSE) Geographic Information Systems: Theory and ApplicationApril 11, 20072 / 16 MotivationFigure:GIS Image from aWIPER SimulationGIS Uses: Visualizing tower locations,relationship to urban areas, etc Analyzing call patterns, activityin relationship to income levels,etc Simulations: agent and towerlocations initialized from data,agents can interact with realworld geographyTim Schoenharl (Notre Dame CSE) Geographic Information Systems: Theory and ApplicationApril 11, 20073 / 16 Sphere, Ellipsoid, GeoidRepresentations ofthe Surface of the EarthSphereSimple solid, uniform of rotation formed froma two-dimensional referred to as The surface of the Earth sgravitational field.

2 Figure:3D plot of taken Schoenharl (Notre Dame CSE) Geographic Information Systems: Theory and ApplicationApril 11, 20074 / 16 DatumFigure:WGS 84 and NAD27 datums compared tothe shape of the Earth. A Datum is a representation of the surfaceof the Earth, an ellipsoid that has localcorrections to account for the irregularnature of the Earth s surface. Datum are important because they relatelocal measurements (I am here) to a globalreference (Here = lat, long ( , )). The most common datum and mostaccepted global datum is WGS 84. Whenyou take a measurement with a GPS unit,that coordinate is likely in the WGS taken Schoenharl (Notre Dame CSE) Geographic Information Systems: Theory and ApplicationApril 11, 20075 / 16 ProjectionsFigure:The Mercator Projection3In order to visualize geographical data, either on a monitor or on paper,the data must be transformed from its spherical/geoidal reference to a 2 Dplane.

3 This is the act of Schoenharl (Notre Dame CSE) Geographic Information Systems: Theory and ApplicationApril 11, 20076 / 16 SoftwareSoftware Titles Proprietary: ESRI ArcMap, ArcGIS, GoogleEarth (almost GIS) Open Source: GRASS GIS, OpenMap,Quantum GIS Open Source Libraries/Tools: GDAL, OGR,PROJ4, GEOS, JTS, GeoToolsTim Schoenharl (Notre Dame CSE) Geographic Information Systems: Theory and ApplicationApril 11, 20077 / 16 GIS and Relational DatabasesPostGIS Allows data to be stored in db tables Stores points, lines, polygons, etc in WKT (Well-Known Text) orWKB (Well-Known Binary) format Can be used as a data store for GIS systems Most importantly, allows users to conduct spatial analysis with SQLqueries When using PostGIS, it is important to understand the SRID: SpatialReferencing system IdentifierTim Schoenharl (Notre Dame CSE) Geographic Information Systems.

4 Theory and ApplicationApril 11, 20078 / 16 SRID sridauthnameauthsridsrtextproj4text4326 EPSG4326 GEOGCS[ WGS84 ,DATUM[ WGS1984 ,SPHEROID[ WGS 84 ,6378137, ,AUTHORITY[ EPSG , 7030 ]],AUTHORITY[ EPSG , 6326 ]], PRIMEM[ Greenwich , 0, AUTHORITY[ EPSG , 8901 ]], UNIT[ de-gree , ,AUTHORITY[ EPSG , 9122 ]],AUTHORITY[ EPSG , 4326 ]]+proj=longlat +ellps=WGS84+datum=WGS84 +nodefsTable:SRID table entry SRID is referenced by an integer, [2000, 32766] SRID contains Information on the agency/group that defined it andthe characteristics (datum, projection, ellipsoid) PostGIS uses PROJ4 to convert data between different SRIDs PostGIS recognizes 2671 different SRIDSTim Schoenharl (Notre Dame CSE) Geographic Information Systems: Theory and ApplicationApril 11, 20079 / 16 SQL and Spatial AnalysisThe advantages of using SQL for spatial analysis are enormous.

5 Considerthe following example: In order to determine the quality of the towerinformation, we wanted to quantify the error in the tower locations. Tower location Information contains a tower ID, street address, postalcode, and latitude/longitude coordinates In a separate file we have polygons representing the shapes of all ofthe postal codes We want to find all the towers that are outside their given postalcodeTim Schoenharl (Notre Dame CSE) Geographic Information Systems: Theory and ApplicationApril 11, 200710 / 16 SQL and Spatial AnalysisThis can be accomplished with a one-line SQL query, joining the tables on thepostal code:SELECT towernum, postalcode, id FROM towerdata INNER JOIN postalcodesON = WHERE NOTcontains( , ).

6 We can improve this analysis by getting a list of the towers, sorted by increasingdistance from the postal code, to allow for discrepancies in coordinates towernum, postalcode, id, distance( , ) FROM towerdata INNER JOIN postalcodesON = WHERE NOTcontains( , ) ORDER BYdistance( , )Tim Schoenharl (Notre Dame CSE) Geographic Information Systems: Theory and ApplicationApril 11, 200711 / 16 SQL and Spatial AnalysisGotchasBe ware of several issues when using PostGIS for spatial analysis: All data must be in the same SRID for spatial queries to be valid Data that is loaded using shp2pgsql without an SRID will come inwith SRID = -1 Spatial queries will return results inDEGREES. For results in meters,you will need to first reproject the coordinates into an SRID that hasa datum in meters, then convert your query to meters.

7 When choosing an SRID, best choices are ones with +ellps=WGS84and +units=m. Find these by querying the spatialrefsys table and searching in the proj4text stringTim Schoenharl (Notre Dame CSE) Geographic Information Systems: Theory and ApplicationApril 11, 200712 / 16 Spatial Analysis with GRASSW orking with GRASS When starting GRASS, it will ask for a bounding box around theregion you are viewing. GIS files come in two types, raster and vector. These define thecommands you will use for data manipulation in GRASS. For vector files: use , For raster files: use , Convert from vector to raster using Build voronoi cells by importing the list of points, then Associate data with the cells using input=inputvectoroutput=outputraster use=attr column=attribute Images can then be exportedTim Schoenharl (Notre Dame CSE) Geographic Information Systems: Theory and ApplicationApril 11, 200713 / 16 Sample Images from the WIPER ProjectFigure:Cell phone activity overlaidon a satellite :3D View of Tower ActivityTim Schoenharl (Notre Dame CSE) Geographic Information Systems: Theory and ApplicationApril 11, 200714 / 16 ReferencesFor More Information : Lavender Wiki GIS page: PostGIS Manual.

8 Grass Homepage: OpenMap: GeoTools: Schoenharl (Notre Dame CSE) Geographic Information Systems: Theory and ApplicationApril 11, 200715 / 16 QuestionsTim Schoenharl (Notre Dame CSE) Geographic Information Systems: Theory and ApplicationApril 11, 200716 / 16