Using Dasymetric Mapping for Spatially Aggregated Crime Data

With availability of crime data to the public via sources such as the Uniform Crime Reports, and increasing geographic information system (GIS) capabilities for mapping crime, macro-level studies of crime have advanced knowledge of how crime is distributed over large areas. Choropleth mapping, commonly used in macro-level studies, visually displays data by assigning the number of crimes or crime rate to the corresponding spatial unit and using different shades or textures for each value or classified values creating a thematic map. However, crime incidents or crime rates are not dispersed evenly within spatial units, and choropleth mapping masks the underlying nuances of the distribution. Artificial boundaries, along with variations in the size of the unit of analysis, can further distort the true distribution of crime. Dasymetric mapping provides a methodology for refining the distribution of crime within a spatial unit. It does so by using additional data, such as land use and census data, to provide a realistic estimate of how crime may be distributed within the units of analysis. Dasymetric mapping is also useful in creating density maps to reveal clusters of crime normally masked with choropleth maps. This paper will show how dasymetric mapping can estimate the spatial distribution of aggregate level residential burglary within political boundaries in Massachusetts based on land use and housing data.