Hot Spot Crime Analysis

-

Hot Spot Analysis: Using Spatial Analysis Tools to Identify, Target, and Predict Criminal Activity

Hotspot analysis is a spatial analysis tool that identifies areas of unusual density. There are many methods for identifying these clusters but the most common are thematic choropleth maps, kernel density, and Moran's I. Hotspot analysis is of particular interest to the criminal justice field as police departments and political leaders are interested in targeting areas with high crime rates. As demonstrated in the two maps below, how crime hotspots are represented can dramatically alter interpretation and future policy decisions.  

Major Types of Hot Spot Analysis:

  • Thematic choropleth
  • Grid Overlay
  • Kernel Density
  • Moran's I


Thematic Choropleth Mapping


Grid Overlay Mapping

First, the grid cells feature class was joined with the total number of homicides (2017), which provided us a joint count of homicides per cell. I exported only the cells containing one or more homicides to a new feature class called “homicide_count” with the expression [join_count > 0]. After sorting from highest to lowest, the top 20% of grid cells were selected (rounded down to 62 total) and exported to a new feature class called “hom_top20per”. A new field was added and calculated for the integer one. Dissolve was performed on all remaining cells to create one new, single feature.



Kernel Density

Through the geoprocessing tools in ArcGIS Pro, a kernel density analysis was performed on the 2017 homicides class. The output cell size was set at 100 feet with a search radius of 2,630 feet (1 US mile). The symbology of the new raster layer was then adjusted to Natural Breaks (Jenks) with two classes. The classes were manually set at 3*mean (3*2.88=8.64) and more than 3*mean (>8.64 to 38.92). By using the reclassify tool, the raster was given two values (1 and 2). The file was then converted to a polygon layer. All grids labeled as two were exported to a new class (hom_kd), which was used as our final hotspot result.



Local Moran's I

Similar to grid overlay mapping, a spatial join of the census tracts and the 2017 homicide table was performed to produce homicide counts per tract. A new field was added to calculate the crime rate per 1,000 households (Crime_Rate = join_count / total households * 1000). The cluster and outlier analysis tool was used to create a new output feature class, which classified areas of high-high, high-low, low-high, and low-low crime. The high-high clusters were exported to their own feature class, which was then dissolved into a multipart feature.



Further Resources:


Comments

Post a Comment

Popular posts from this blog

Coastal Flooding & Storm Surge Analysis

-
Image
     As climate change continues to intensify storms and increase flooding, it is vital for communities and governments to prepare for inevitable storm damage. In this week's lab, we explored coastal flooding models and prepared our own storm surge analysis. Our study area was Mantoloking, New Jersey, a low-laying barrier island that suffered major damage from superstorm Sandy in 2012. More information about the storm can be found on New Jersey's Department of Environmental Protection . Superstorm Sandy     By comparing LiDAR data before and a few days after the storm, we can create a comparison analysis. The LiDAR data is converted into a DEM by first creating a TIN (triangulated irregulated network). As the TIN shows, the topography varies little with only very mild sloping.  LiDAR .las point cloud dataset is converted into a  TIN       By subtracting the two DEMS, we can see the changes in the environment. Further color manipulat...
Read more

Welcome to GIS 5007 Spring 2023

-
Image
My name is Kelsey Yoder-Ostroski and I am a new(er) student in the world of GIS. This is my third class in UWF's GIS Admin graduate program as a full-time student. At the moment I am a caregiver to our kids but I also do volunteer work with a nonprofit providing support to local moms. We are a military family; we have lived in the Navarre area for 3 years and are preparing to move to Germany this summer. I'm not ready for the cold! I am originally from Texas so I love the heat, but we can't wait to see everything Germany and its neighbors have to offer.  In my free time, I enjoy learning German, baking (mostly breads), or doing home improvement projects. I have a weakness for cookies and power tools.  I am pursuing a master's in GIS business administration because I love how applicable and versatile it is. It literally encompasses every topic on Earth from the land to its resources and to all of history. I'm not 100% sure what I will do after graduating but I am int...
Read more

Introduction to Python

-
Image
As an introduction to Python, which is a computing language that is embedded in ArcGIS Pro, we were tasked with creating a flow chart and pseudocode for converting radians into degrees. Flowcharts are a visual method of planning, which is an especially important process to complete before delving into complicated code. Flowchart created in iodraw.com The aim of this exercise was to introduce us to calling up IDLE and ArcGIS Notebooks. IDLE looks like the stereotypical coding interface with its intimating emptiness and blocky text; whereas Notebooks is nearly the opposite with its bright, white interface, menu options, and bubbly outlining. Both programs act as language interpreters for the Python coding language, converting user commands into binary which is then read and executed by the computer processing system. 
Read more