Revisiting the Spatial Analysis

We examined spatial patterns of crime incidents in national forests covering 112,396 km in the northwestern United States. In this study we analyzed a database containing 40,003 spatially referenced crime incidents representing felonies, infractions, and misdemeanors during 2 calendar years (2003–2004) at several geographic scales. We applied several geospatial analytical techniques including quadrat analysis, nearest neighbor analysis (NNA), and nearest neighbor hierarchical (NNH) clustering to investigate crime incident spatial patterning. These geospatial tools were beneficial in identifying crime incident relationships contained within a large, complex spatial database. NNH clustering identified 15 regional clusters with 16,138 crime incidents, focused in the central portion of Oregon’s national forests, specifically in the Deschutes, Mount Hood, and Willamette National Forests. Subsequent NNA tests confirmed spatial patterning in all three forests. Closer examination of a confirmed hot spot in one forest revealed a recreation corridor with adjacent recreation destination amenities and a large proximate metropolitan area, a combination of circumstances not apparent at the initial regional analysis scale. Other spatial...