Lab 6: Applications in GIS

This week's lab focused on crime analysis.  The three types of to determine crime hotspots in Albuquerque was Grid-based thematic mapping, Kernel Density, and Local Moran's I. After performing all three analysis, the next step was to compare which hotspot analysis is best for predicting future crime.

To create the map below, I first performed the grid-based thematic mapping.  To find the hotspots by using grid-based thematic mapping, I did a spatial join of the grids with the 2007 burglaries.  I selected the grids that contains at least 1 burglary and made it into a new shapefile.  I then found the top 20%.  I then dissolved the polygons to make one single polygon.  I then added a field to calculate the square kilometers.

Next I used Kernel Density to determine the hotspots.  I set the environment to only show the grids.  I then used Kernel Density tool to calculate.  The parameters for the tool used output cell size of 100 and the search radius of 1320 feet.  I kept the area units to square miles.  Next I removed the areas with 0 density.  I found the mean and used that to determine the classifications.  Once that was complete, I converted the raster to polygons.

Finally, I used Local Moran’s I to determine hotspots.  I did a spatial join of block groups and 2007 burglaries.  I then found the crime rate of burglaries to housing units.  Next I used the Cluster and Outlier Analysis script and left the parameters to the defaults.  Next I used a query to create a shapefile of just the HH polygons.  I dissolved the polygons and then found the area by using calculate geometry.

Below is a map layout of all three analysis. This helps the Albuquerque police determine where to patrol more by comparing the 2007 hotspots to the 2008 burglaries.

Hotspot Analysis