Logo image
A Robust Geographically Optimal Zones-based heterogeneity model for analyzing the spatial determinants of national traffic accidents
Journal article   Peer reviewed

A Robust Geographically Optimal Zones-based heterogeneity model for analyzing the spatial determinants of national traffic accidents

Hongyi Luo, Peng Luo and Liqiu Meng
GIScience and remote sensing, Vol.62(1), 2448283
12/31/2025
DOI: 10.1080/15481603.2024.2448283
url
https://doi.org/10.1080/15481603.2024.2448283View
Published (Version of record) Open Access

Abstract

Revealing the factors associated with traffic accident risk across cities nationwide, including demographic and economic elements, is crucial for supporting traffic safety policy, urban planning, and insurance evaluation. Spatial stratified heterogeneity models, such as Geographically Optimized Zone-based Heterogeneity (GOZH) model, are widely used for analyzing the spatial association in large scale. However, as their spatial discretization process heavily depends on a manually set complexity parameter (cp), introducing significant uncertainty. To address this, we developed a Robust GOZH (RGOZH), by analyzing the inter-parameter relationships within GOZH and introducing the Q function as an optimization function to achieve precise and controlled geographic partitioning. By selecting the optimal cp, RGOZH produces the most reliable spatial partitioning results. Testing RGOZH on Germany's traffic accident dataset revealed strong geographic patterns, with RGOZH achieving superior spatial groupings while maintaining over 80% of explanatory power - a stark contrast to the less interpretable results from GOZH. RGOZH identified vehicle ownership, government employee proportion, and income level as primary factors shaping traffic accident risk. This study highlights the critical role of large-scale spatial pattern analysis in traffic management and establishes RGOZH as a robust framework for future interdisciplinary geospatial research. Furthermore, RGOZH provides a replicable method that can adapt to various regional datasets, enhancing its utility in international traffic safety studies. As a methodological advancement, RGOZH demonstrates the value of integrating optimized spatial parameters into predictive geospatial models.
GOZH spatial association Spatial stratified heterogeneity traffic accidents

Details

Metrics

Logo image