Modeling Anthropogenic Urban Fire Risk and Its Environmental Pollution Implications Using LR–WLC Approaches: A Case Study of Kashan, Iran

Urban fires represent one of the most significant human-induced hazards in modern cities, often linked to increased pollution emissions and environmental degradation. This study aims to evaluate fire risk in Kashan using a multi-criteria evaluation (MCE) method based on fuzzy logic, logistic regression, and the analytic hierarchy process (AHP). Initially, criteria and indicators were identified using the Delphi method and normalized using fuzzy logic as capacity and vulnerability factors. The weight of these factors was determined using the AHP method. Subsequently, all layers were combined using the weighted linear combination (WLC) technique. To evaluate the effectiveness of the methods, the outcomes were contrasted with the fire risk map produced via logistic regression. The ROC index was also used to validate the WLC model and logistic regression results. The results reveal that 0.820% of the total studied area (788.96 hectares) and 5% of the area identified by the WLC method (30,483 hectares) are at high risk of urban fires. The ROC index validation results, with values of 0.95 and 0.74 for the logistic regression and WLC methods, respectively, confirmed the superior predictive performance of the LR model. The validated fire risk maps not only support emergency response planning but also highlight spatial correlations between high-risk zones and areas of potential pollutant release or poor air dispersion. These findings demonstrate that mapping human-induced fire risk can provide a valuable spatial framework for air quality management, the mitigation of anthropogenic pollution risk and for advancing sustainable urban environmental management toward cleaner and safer cities.