Spatial Point and Area Pattern Analysis

In this video, we explore different ways to analyze and visualize large point datasets using crime data from Manhattan as an example. With over 100,000 crime incidents in a single year, simply plotting points on a map doesn’t tell us much. That’s why we use specialized spatial analysis techniques in ArcGIS Pro to reveal meaningful patterns. You’ll see step-by-step how to: Create a Fishnet Grid: Break Manhattan into evenly sized cells and calculate crime density per square kilometer. Aggregate by Census Tracts: Compare crime densities by administrative boundaries. Apply Kernel Density Estimation: Generate smooth surfaces showing where crimes cluster, using different search radii (500m and 1km). Measure Spatial Autocorrelation (Moran’s I): Statistically test whether crime is clustered, random, or dispersed. Run Cluster and Outlier Analysis (Anselin Local Moran’s I): Identify crime hotspots and areas that stand out compared to their neighbors. By the end, you’ll understand the strengths and limitations of grid-based, census-based, and density-based analyses, as well as how statistical clustering tools add rigor to visual interpretation. Whether you’re studying geography, criminology, or just interested in how spatial data science uncovers hidden patterns in big datasets, this walkthrough offers a practical introduction.