A multidimensional approach to cyclone risk assessment: integrating GIS and Fuzzy-AHP data in coastal Odisha, India

The eastern coast of India faces an increasing risk from tropical cyclones, which will cause estimated annual economic losses of USD 60 billion by 2100. This research develops and implements a high-resolution, multidimensional cyclone risk assessment for six Odisha coastal districts through the integration of remote sensing data with Geographic Information Systems (GIS) and fuzzy Analytic Hierarchy Process (Fuzzy-AHP) and machine learning-based–based sensitivity analysis. The fourteen spatial criteria were converted into normalized membership grades through linear fuzzy membership functions before assigning weights from expert pairwise comparisons. The hazard, vulnerability, and mitigation capacity indices were combined to generate a composite cyclone risk map. The XGBoost sensitivity analysis showed that socio-economic vulnerability (0.09 importance) exceeds both physical hazard metrics (0.025) and mitigation capacity (0.075). The High–Very and High-risk areas cover 37.6% of the study area, primarily in Balasore, Bhadrak, and Kendrapara districts, because of increased surge penetration and population exposure. The preparedness levels of interior zones reach below 55% as indicated by the mitigation capacity mapping results. The findings highlight the need for targeted investments in embankments, shelters, and early warning systems in high-risk areas. This framework offers evidence-based disaster planning and resilience building in cyclone-prone coastal regions.