Measuring nonlinear relationships and spatial heterogeneity of influencing factors on traffic crash density using...

Traffic crash density (or frequency) remains a critical public safety concern, posing significant challenges for transportation planning and risk mitigation, particularly in rapidly urbanizing regions. Crash occurrence is closely associated with socio-demographic and roadway determinants, which vary substantially across space. As a result, crash patterns often exhibit pronounced spatial heterogeneity because of socio-spatial disparities and differences in regional conditions, especially at finer spatial scales such as the census tract level. This study applies a Geospatial Explainable Artificial Intelligence (GeoXAI) framework through combining a high-performing machine learning model with GeoShapley to analyze the spatially heterogeneous and nonlinear determinants of traffic crash density in Florida census tract-level. Through comparing the results obtained from GeoShapley framework against other established methods (e.g., SHapley Additive exPlanations (SHAP) and Multiscale Geographically Weighted Regression (MGWR)), GeoXAI framework demonstrates its powerful ability to provide interpretable, tract-level insights into how roadway characteristics and socioeconomic factors contribute to crash risk from the perspectives of nonlinearity and spatial heterogeneity simultaneously. Key variables such as road density, intersection density, neighborhood compactness, and educational attainment exhibit complex nonlinear relationships with crashes. Extremely dense urban areas, such as Miami, show sharply elevated crash risk due to intensified pedestrian activities and roadway complexity. Other major metropolitan areas including Orlando, Tampa, and Jacksonville display significantly higher intrinsic crash contributions, while rural tracts generally have lower baseline risk. Based on these findings, the study proposes targeted, geography-sensitive policy recommendations, including traffic calming in compact neighborhoods, adaptive intersection design, speed management on high-volume corridors such as I-95 in Miami, and equity-focused safety interventions in disadvantaged rural areas of central and northern Florida.