Hot Spot Analysis of 911 Calls map is available

By Charlie Frye, Esri Chief Cartographer

Link to Hot Spot map

The Hot Spot Analysis poster shows the steps in the analysis of 911 Call data. The data were processed using the Hot Spot Analysis tool and the design of the poster is, we think, faithful to the underlying data.

I did the first edition of this poster almost three years ago, and since then it has been tacked up on the wall at the end of one the hallways here in Redlands. When Jack brought his tours through the Software Products & Development area, he’d often show this poster, extolling the analytical power of GIS. The original poster was a little overly flashy, but more importantly after we recently checked with our spatial statistics experts, we found it was symbolized in a way that was slightly misleading. Thus, we undertook an upgrade and presented it at the Users Conference in our ”Mapping the Results of Your Geographic Analysis” technical session.

As I was putting the finishing touches on the poster in order to publish it in the Maps section of Mapping Center; I found a data error that related to how I misconstrued some of the expert advice I had gotten about which value ranges were most important. So, the poster got another upgrade and what you see on Mapping Center now is a version that I am quite satisfied with. It just goes to show that making maps will keep you humble.

This entry was posted in Mapping and tagged , , . Bookmark the permalink.

Leave a Reply


  1. cpoynter says:

    How do you construct the models for this?

    Can they be added for download?

  2. cfrye says:


    We don’t have the models, only screenshots of them. These are the screen shots I made years ago when I made the first edition of this map.

    I consulted Dr. Lauren Scott, who created the models and she indicated that the model and she gave the following advice:

    First, the model is likely not sufficiently generic for general use. So, if you’ve already got incident data then either of the following are potential courses of action:

    1) Make a backup copy of their incident FC
    2) Run integrate to snap nearby features together (tolerance to 100, 200, or 300 feet, for example)
    3) Run the Collect Events with Rendering tool to created weighted points
    4) Run Hot Spot Analysis with Rendering on the ICOUNT field.


    1) Use Spatial Join to count the number of incidents falling within each census block.
    2) Run Hot Spot Analysis with Rendering on the count field.

    You will also need to determine what distance band to use for analysis. This topic and others relating to hot spot analysis are covered well in this link: