GIS C ONCEPTS AND A GIS M RC ETHODS 3rd Edition, July 2007 David M. Theobald, Ph.D. Warner College of Natural Resources Colorado State University GIS Concepts and ArcGIS Methods Copyright Copyright © 2007 by David M. Theobald. All rights reserved. Trademarks ArcGIS, ArcMap, ArcCatalog, ArcToolbox, ArcView, ArcInfo, Spatial Analyst, Shapefile, Image Analysis, 3D Analyst, and Avenue are registered trademarks of Environmental Systems Research Institute, Inc. Publisher Conservation Planning Technologies, 1113 West Olive Street, Fort Collins, Colorado, 80521, USA. Phone: 970.980.1183. For book inquiries, please visit the following website: http://www.consplan.com Published in the United States of America ISBN 0-9679208-4-1 (paper) ii GIS Concepts and ArcGIS Methods GIS defined: what is where and why the landscape? Third, data can be transformed into information that fits a user’s context through a rich set of analytical tools. The relationship between different attributes can be examined to investigate, for example, the spatial arrangement of aspen clones with respect to soil type, aspect, or time since disturbance. Fourth, geographic data in a GIS do not suffer, as they do in paper maps, from the fundamental trade-off between spatial detail and geographic coverage (although because most geographic data are derived from paper maps, they typically can only represent features to a certain resolution). This is because the scale of a map can change depending on user needs, yet the underlying data remains the same. Clearly, GIS utilize maps as the primary means of graphic display, however, it is critical to understand that it is the underlying digital data that is associated with the map display that gives GIS its analytical power. FIGURE 1.1. A map of land cover types in Colorado, USA. Patches of quaking aspen are displayed in bright yellow, and county boundaries (black lines) are shown for reference.2 (Download color figures from www.consplan.com). 2. Data source: the Colorado Gap Analysis Project, 2000. GIS Concepts and ArcGIS Methods 3 GIS defined: what is where and why FIGURE 1.2. Scientists using GIS to develop a conservation plan for the Southern Rockies Ecoregional Plan. 1.1.3 Spatial analysis Another characteristic commonly used to describe GIS and differentiate it from other spatial technologies is that it allows users to conduct advanced spatial analysis. Spatial analysis is a general term to encompass the manipulation of spatial data to examine the location, attributes, and relationships of geographic features to gain information. There are three types of spatial relations3: • proximity, • directional, and • topological. 3. A classic paper on spatial relations is: Freeman, J. 1975. The modeling of spatial relations. Computer Graph- ics and Image Processing 4: 156-171. GIS Concepts and ArcGIS Methods 5 GIS defined: what is where and why FIGURE 1.3. Examples of topologic relationships: equivalent (top left, two polygons of Lake Tahoe, CA on top of one another); partial equivalent (top right, Lake Tahoe overlaps parts of 5 counties); contained (bottom left, two islands are wholly inside Lake Mono and Lake Mono is wholly within Mono County); adjacent (bottom right, Placer County is adjacent to Washoe, Carson City, Douglas, and El Dorado counties); separate (bottom right, Washoe County is disjoint from El Dorado County). Spatial analyses range from simple to advanced and typically rely on some combination of spatial relations. For example, location analysis enables you to query “What is here?”, which uses a combination of equivalent (Is this city at the same location as the user- defined location?) and containment (Which polygon is the user-defined location within?) relations. Nearest-neighbor analysis utilizes adjacency relations. It is used, for example, when all the counties adjacent to a disease outbreak need to be determined. Proximity analysis uses the concept of a buffer around an object to determine what features are within a certain distance of another feature. This assists us in answering questions like: which houses are further away than a three-mile radius of a fire station? What areas are sensitive to encroachment in parks and protected areas? (see Figure 1.4). Note that the relationships most closely associated with advanced GIS analyses—connectivity, containment, and contiguity—are all topological relationships. GIS Concepts and ArcGIS Methods 7 Introduction to GIS and ArcGIS FIGURE 1.4. Spatial analysis can be used to examine how parks and protected areas may be effected by non-compatible land uses in adjacent areas. Jardin Botanico outside San Miguel de Allende, Mexico. 1.2 A brief history of GIS Geographic Information Systems (GIS) have emerged as a key technology to manipulate and analyze geographic data. Although there are many roots of GIS (and they are often intertwined), there are a few that are worth mentioning here. The term GIS was first coined in the early 1960s by Roger Tomlinson during his work with the Canada Land Inventory (CLI). At that time, a system was needed to analyze the data collected by the CLI to support the development of land management plans for rural areas of Canada. In the United States, the Bureau of the Census developed the DIME-GBF data structure and 8 GIS Concepts and ArcGIS Methods Vector extension to create and analyze these data (though you can view an existing TIN in ArcMap). FIGURE 2.6. Example of a TIN data structure representing the elevation surface near Yellowstone National. Park, Wyoming. 2.2.2.4 Other structures In addition to shapefiles, coverages, and geodatabases, a number of other spatial data formats are directly supported by ArcGIS. That is, these do not need to be imported or converted somehow to be usable. Table 2.11 summarizes the vector data structures that are supported in ArcGIS. TABLE 2.11. Supported vector data formats. Name Description ArcIMS Feature Service, ArcIMS Feature Service streams features over the Inter- ArcIMS Map Service net, so a Feature Service layer works the same as any other feature layer. Coverages Both Workstation and PC ArcINFO coverages. DGN MicroStation design file, supported to v8. DWG AutoCAD drawing file. Note that a CAD layer can be geo- referenced by using the Transformation tab in the Proper- ties dialog (through v2004). DXF CAD interchange files. Note that ASCII, binary, and par- tial drawing interchange files that comply with DXF are supported. GIS Concepts and ArcGIS Methods 53 Geographic data FIGURE 2.7. Three patches of aspen embedded in sagebrush illustrate the difference between raster and feature attribute tables, and zones and polygons. Type Map Attribute Table Polygon ID Type 1 Sage 2 Aspen 3 Aspen 4 Aspen raster (zone) Value Type 1 Sage 2 Aspen raster (after Value Type region grouping) 1 Sage 2 Aspen 3 Aspen 4 Aspen 58 GIS Concepts and ArcGIS Methods Raster FIGURE 2.9. Raster data represented as a flat ASCII file (upper left), run-length encoded (upper-right), and graphic (lower). NoData values are value -9 (and shown in black). Flat ASCII Run-length -9 -9 1 1 0 1 0 1 -9 -9 2 -9 2 1 1 0 1 1 1 0 1 1 2 -9 -9 1 1 2 2 2 1 1 -9 -9 -9 1 1 1 2 2 2 2 3 3 2 -9 2 1 3 2 2 1 1 -9 1 1 1 1 2 2 2 2 3 3 4 1 4 2 2 3 0 1 1 1 1 2 3 3 3 3 4 1 4 2 2 3 0 0 0 1 1 1 1 1 1 1 1 0 4 1 1 2 4 3 0 2 2 3 3 3 -9 4 4 4 3 0 7 1 0 2 2 3 3 3 3 4 4 4 1 0 2 2 3 3 1 -9 3 4 0 2 2 2 4 4 4 4 4 4 1 0 2 2 4 3 3 4 1 0 3 2 6 4 2.3.1 GRID A GRID is an implementation of a raster data structure developed by ESRI. GRIDs are implementations of the single-layer, multiple value method. GRIDs have m columns and n rows of cells, and are ordered starting from the origin in the upper-left corner. There are two types of GRIDs: integer GRIDs and real GRIDs. Integer values are used to represent nominal, ordinal, and ratio data and cells that have the same integer values have the same attributes. Integer GRIDs can also be used to represent continuous phenomena. Note that the Spatial Analyst extension is needed to conduct analyses on GRIDs, though GRIDs can be displayed and queried in ArcMap without Spatial Analyst. Run-length encoding is used to compress GRIDs. For example, a GRID comprised of 1,000 columns and 1,000 rows containing all 1s requires about 30KB, but a checkerboard of GIS Concepts and ArcGIS Methods 61 Attribute (non-spatial) data FIGURE 2.10. A remotely-sensed image taken mid-morning on April 1, 2004 of the Picnic Rock fire west of Fort Collins, Colorado (image from NASA Earth Observatory website). 2.4 Attribute (non-spatial) data Recall that one of the key features of GIS is the explicit linkage between a geographic feature at a place and its attributes, or information about that feature at that location. A collection of GIS Concepts and ArcGIS Methods 63