Cluster Analysis

I used ArcGIS Pro 3.0.1 to solve this problem. The main geoprocessing tools that I used include “Cluster and Outlier Analysis” and “Hot Spot Analysis”. There are MXD files provided by GIS520 class, which are maps of the calls for services from Battalion 2, Fort Worth, TX (calls for service data, census blocks data, and street network data), and census blocks with median household income from Dallas County.

(1) For exercise 1a, I imported the map layer of calls for services into ArcGIS Pro. Then I ran the “Cluster and Outlier Analysis” tool without specified distance. Then I changed the symbology of the result layer to make it stand out and easy to interpret. Then I turned on the income layer and compared it with the result layer to see if high-high clusters or low-low clusters are more likely to be located in certain income areas.

(2) For exercise 1b, I ran the “Cluster and Outlier Analysis” tool again with a specific distance band of 900 feet. Then I updated the symbology to improve the visibility of the result layer. Then I compared the result layer with the income layer to explore if income and clusters are correlated.

(3) For exercise 2, I first imported the map layer of Dallas County census blocks’ median household income. Then I ran the “Hot Spot Analysis” tool with a distance band of 5280 feet. Then I refined the map elements to make them easy to interpret. Then I visually examined where hot and cold spot occurs.

Local spatial pattern analysis is another great skill to have. In addition to examining if clustering is happening from the global spatial pattern assignment, the local spatial pattern analysis tools allow me to identify where the clustering occurs, which has very practical applications, such as prioritizing resources to specific neighborhoods. A possible scenario would be for an avian ecologist to explore the local spatial pattern of bird diversity.

Problem description: As a GIS specialist, I am asked to perform local spatial pattern analysis to determine if high bird diversity areas are clustered together and if the hot spots would correlate with median household income in the triangle area of NC.

Data needed: Tabular data of bird species richness from Triangle Bird Count and shape file of US Census tracts with median household income.

Analysis procedures: I would import the point layer of Triangle Bird Count with randomly distributed survey locations and data on bird species richness. Then I would run a “Hot Spot Analysis” and "Cluster and Outlier Analysis" on the bird data layer. Then I will also import the census block layer with median household income (graduated color symbology). Lastly, I will visually compare if hot spots of high bird diversity would be located in high-income census blocks based on the "Hot Spot Analysis" and if high-high clusters (high income and high bird diversity) could be identified from the "Cluster and Outlier Analysis"