Comprehensive analysis framework reveals fire risk factors in urban areas, suggesting equitable prevention strategies.
Key Points
This research aims to develop a framework integrating the PSR model and environmental criminology to analyze urban fire risks and improve prevention strategies.
Analyzed data from Shenzhen's digital urban management system (2019) related to four fire risk types.
Applied spatial error models to assess variance across fire risk event types, achieving 82–89% explained variance.
Utilized spatial 5-fold cross-validation to evaluate model performance without overfitting.
Elderly proportion positively associated with water heater misuse (coefficients=2.64, p<0.01) and unauthorized power wiring (coefficients=3.06, p<0.01).
Restaurant density showed positive associations with all risk types (coefficients=0.24–0.60, p<0.01).
Functional diversity negatively affected visible violations but positively influenced concealed behaviors like electric bike charging violations.