Mapping Hidden Risk Landscapes: Integrating Spatial Factor Analysis with Risk Terrain Modeling to Identify Latent Crime Environments

Research in environmental criminology shows that crime concentrates at places because opportunity is structured by the built environment. Risk Terrain Modeling advances this insight by estimating how spatial features elevate risk, but RTM is typically undertaken offense-by-offense, leaving unclear whether crime-specific “risk terrains” converge into a smaller number of shared, latent risk environments. In response, we integrate RTM with Spatial Factor Analysis to identify environmental risk structures across multiple crime types, using police incident data from Detroit. Factor retention was guided by parallel analysis and model fit, with oblique rotation used to accommodate correlated latent dimensions. The analysis yielded a 3-factor solution differentiating a broad General Risk Intensity dimension, a Public-Space Convergence dimension capturing contact-crime risk in commercial and transit-adjacent settings, and a narrower Vehicle-Target dimension specific to auto theft. Spatial structure in factor scores was evaluated using global Moran’s I and Local Indicators of Spatial Association to identify statistically coherent hot and cold regions. The current study demonstrates how RTM outputs can be synthesized to produce interpretable, citywide typologies of environmental vulnerability and provides a methodological template for moving from offense-specific crime risk estimation to multi-harm place-based diagnosis. Policy implications for place-based prevention and strategic resource allocation are discussed.