Instead of viewing the typical hot spot analysis maps, a fun wrinkle is to generate space time cubes aggregated by points. Each cube contains a sum of the points that occurred within that bin at the specified time interval. The cubes can then be analyzed using the "emerging hot spot analysis" geoprocessing tool in #ArcGIS.
As you can see in the picture below, viewing historical crimes in downtown Denver is an arduous task. Viewing every crime with no context to location or time is a bit of a lost cause.
The first step to creating a better visualization is to derive hexagonal bins from the crime data. These bins will allow us to categorize the data by attributes such as type, date and neighborhood.
Once the bins are created, we can then categorize them based on time. To do this, I utilized a tool called space time cube analysis in ArcGIS Pro. The tool utilizes a time field to classify the bins. For the crime layer that I downloaded from the City of Denver GIS hub, I used the first occurrence date in the attribute table. Since it was such a large dataset, the interval was set at 1 month and the distance interval at 350 meters. The interval should be set at a distance comparable to the study area.
In the map below that analyzes crimes in downtown Denver, dark red hot spots have been significantly hot spots for 90% or more of the time slice. Persistent cold spots (dark blue) are bins where crime is becoming less and less prevalent. These are areas where crime is statistically, and persistently, less prevalent. The concerning areas are bins where crime is not only persistent but also increasing.
I also published the same data to ArcGIS Insights to create a 1 page #dataviz on crimes in downtown #Denver correlated with businesses and rental properties. The next step is to utilize the R-ArcGIS Bridge to add and identify attributes that influence crime.
