Lecture Materials by Austin Troy except where noted© 2008
Part 3. Regions Topology and advanced vector queries
------Using GIS--Fundamentals of GIS
Lecture Materials by Austin Troy except where noted© 2008
Regions Topology Allows for • Nested representation: allows
for hierarchical nesting of certain polygon types within larger polygon types
• Example: a regions layer may contain block groups within tracts within counties, within states, each level with their own behaviors and each with their own membership; The former Soviet Union through time
Lecture Materials by Austin Troy except where noted© 2008
Regions Topology Allows for
• Associations: dispersed or connected polygons can be given membership in the same feature; deals with “void area”
• Example: Hawaii or Florida; each comprised of many sub-features, but computer recognizes them as single entity
Lecture Materials by Austin Troy except where noted© 2008
Regions Topology Allows for
• Overlapping geographic features: features that are not spatially mutually exclusive, but overlap can be coded for
• Example: fire perimeters, overlapping soils or geologic data at different layers
From ESRI ‘s ArcINFO help
Lecture Materials by Austin Troy except where noted© 2008
Fire perimeters over time
1880-1900
Lecture Materials by Austin Troy except where noted© 2008
Fire perimeters over time
1900-1915
Lecture Materials by Austin Troy except where noted© 2008
Fire perimeters over time
1915-1930
Lecture Materials by Austin Troy except where noted© 2008
Fire perimeters over time
1930-1945
Lecture Materials by Austin Troy except where noted© 2008
Fire perimeters over time
1945-1960
Lecture Materials by Austin Troy except where noted© 2008
Fire perimeters over time
1960-1975
Lecture Materials by Austin Troy except where noted© 2008
Fire perimeters over time
1975-1990
Lecture Materials by Austin Troy except where noted© 2008
Fire perimeters over time
1990-2001
Lecture Materials by Austin Troy except where noted© 2008
Regions and Subclasses
• In Arc/Info a grouping of polygons that share some logical grouping are called a subclass
• Each hierarchical level is a subclass in this case
From ESRI ‘s ArcINFO help
Lecture Materials by Austin Troy except where noted© 2008
Regions and Subclasses
• This allows attribute data to be stored much more efficiently. attributes can be associated with them which are distinct from the attributes associated with the individual polygons
• Region attributes are stored in subclass-level attribute tables that are named cover.PATregionsubclass (e.g. Census.PATcounty).
Lecture Materials by Austin Troy except where noted© 2008
Regions and Subclasses
• In the upper one, the three islands are managed as a subclass
• In the lower, 103, 104 and 105 refer to areas of fire damage for three given years, each managed as a subclass
From ESRI ‘s ArcINFO help
Lecture Materials by Austin Troy except where noted© 2008
Methods for Making Regions• AREAQUERY a dynamic polygon overlay tool that allows you to integrate, query, and
aggregate polygon and region layers from multiple coverages in a single operation.• POLYREGION converts a polygon coverage to a region subclass. All polygons in the
in_cover become a region of the output subclass.• REGIONBUFFER creates buffer regions around specified input coverage features.• REGIONCLASS creates preliminary regions from arcs by region subclass, or appends
preliminary regions to existing regions for the subclass.• REGIONCLEAN merges adjacent polygons that belong to the same regions.• REGIONDISSOLVE constructs new region subclasses by aggregating polygons or regions with the
same value for a specified item.• REGIONJOIN creates new regions by joining a related table to a region subclass attribute table.• REGIONQUERY creates new regions from existing regions or polygons based on attribute values
in multiple subclasses and specified output items.• REGIONXTAB an ATOOL enabling you to do region cross tabulation.• In ARCEDIT• MAKEREGION creates regions from selected polygons.
From ESRI ‘s ArcINFO help
Lecture Materials by Austin Troy except where noted© 2008
Example: Areaquery
• This is an extremely helpful tool for doing multi-layer queries with a minimum of steps
• Utilizes the regions approach for analysis, querying and manipulation of overlapping coverages
• Areaquery’s power lies in its ability to integrate many coverages (up to 32) into one “super-coverage” so all the coverages can be analyzed
• Can do queries for highly specific criteria, and on numerous attributes
• Similar to doing multiple unions, but simpler and more data efficient
Lecture Materials by Austin Troy except where noted© 2008
Areaquery• Say you want to do a specific query for the following:
– landtype is brushland, soil is suitable for development, site within 300 meters of sewer line, site is 20 meters beyond existing streams and the site must contain a contiguous area of 8000 square meters
• First, you specify each coverage to be analyzed; in the process, each polygon coverage is turned into a regions cover
• Then you specify a logical query across data sets
Lecture Materials by Austin Troy except where noted© 2008
Areaquery example• Say we’re doing a site selection where:• preferred landuse is brushland• Soil types should be suitable for development• Site must be within 300 meters of existing sewer lines• Site must be beyond 20 meters of existing streams• Site must contain an area of at least 8000 meters square• These codes are specific to the coverages in our database:
– LANDUSE (attribute LU-CODE = 300 - brushland)– SOILS (attribute SUIT >= 2 - suitable for development), – SEWERBUF (a 300 meter buffer around sewerlines)– STREAMBUF (a 20 meter buffer around streams).
From ESRI ‘s ArcINFO help
Lecture Materials by Austin Troy except where noted© 2008
Areaquery
From ESRI ‘s ArcINFO help
Lecture Materials by Austin Troy except where noted© 2008
Areaquery example
• Now let’s say we wanted to try this on our habitat site selection example we did in lab
• We would start by turning all our buffer feature classes into coverages in Arc Catalog
Lecture Materials by Austin Troy except where noted© 2008
Areaquery example
• Then input expression
• Followed by the coverage names and subclass names
Lecture Materials by Austin Troy except where noted© 2008
Areaquery example
• Followed by your logical expression, which here is just presence/absense
Lecture Materials by Austin Troy except where noted© 2008
Areaquery results
• You get a coverage with a bunch of subclass, one for your desired conditions, and one for each constituent subclass
• Hence, this becomes like an Arc/Info version of a geodatabase
Lecture Materials by Austin Troy except where noted© 2008
Areaquery results
Here’s land meeting all our criteria
Lecture Materials by Austin Troy except where noted© 2008
One more example
• Now let’s say we’re looking for vacant land (LUcode=3100) that is in a floodplain (not $FLOOD) and not on public land (not $open), so that the government can target areas to buy to prevent future floodplain development
Lecture Materials by Austin Troy except where noted© 2008
Areaquery example 2
• Again we specify our function and then our subclasses
• Then our criteria
Lecture Materials by Austin Troy except where noted© 2008
Areaquery example 2
• Then we get the following result