Inferring areal geographies for non-bounded places in Connecticut

Presentation at the NKOS Workshop, JCDL 2002, on Digital gazetteers: integration into distributed digital library services, July 18, 2002

Patrick McGlamery, Map and Geographic Information Center, Homer Babbidge Library, University of Connecticut

This presentation will discuss strategies for inferring areal geographies for non-bounded places in Connecticut using GIS analysis of raster datasets and results from experimentation with imputing the 'populated place' feature type from Land Use/Land Classification data.  These places lack administrative boundaries, but their populated attribution allows for inferring boundaries from analyzing classed remotely sensed imagery.  Similarly, areal geographies of hills can be inferred from analyzing Digital Elevation Models (DEM) and comparing them with the gazetteer entries with the 'summit' feature type.