Detecting change in hot spot analysis

Question

I have completed an optimized hot spot analysis for (geocoded) crime incidents for the same three-month period in two, different years.  I would like to understand the changes between the two hot spot analyses.  Specifically, I am looking at the census tracts (aggregating by polygons in analysis) to see where the increase/decrease in crime has been between the two, different years.

What are some ways to do this efficiently?

I've done change detection and map algebra with raster data but not with layers of vector data symbology.  A series of maps (small multiples) would not work for the project I'm trying to do.

I have looked at the emerging hot spot analysis method.  However, I didn't see how to work the time slices to fit the problem.  I tried to create the space time cube by aggregating points.

JenniferGrove · Answer

Thank you so much for sharing this!  I'm excited to try the steps you listed.

For my project, I ended up doing a crime surface density using the point density tool for each year.  Then, I used map algebra to find the difference and created a crime surface density map for the difference.

LaurenGriffin · Answer

If your goal is to show where crime is increasing and where it is decreasing, you could run Hot Spot Analysis on the differences. This is how I would do it:

1) Run Optimized Hot Spot Analysis on the incident points to get my fishnet grid (squares or hexagons).

2) Use Summarize Within to get the count of crimes for time period 1 and the count of crimes for time period 2 within each grid cell.

3) Use Calculate Field to compute the difference between crimes in time period 1 and crimes in time period 2 (most current time period crime count minus older time period crime count).

3) Run Optimized Hot Spot Analysis setting the Analysis Field to the differences. (If the distance the tool recommends is too large or too small, you can over ride it with a distance that makes the most sense).

The output will show you where crime is increasing (any hot spots) and where crime is decreasing (any cold spots).

It's worth a try, anyway.

Best wishes!

Lauren