A framework for flood risk zoning and prioritization combining maximum entropy and game theory.

Floods, as one of the most destructive natural disasters, impose extensive human and economic losses on communities annually. This study pursues two primary objectives by introducing an innovative hybrid framework: (1) identifying flood-prone areas in Iran's Kashkan River Basin using the Maximum Entropy (MaxEnt) model, and (2) prioritizing critical sub-basins based on the Borda method in game theory. Variable selection was performed using the Random Forest algorithm, resulting in the identification of nine key factors influencing flood occurrence. Important variables affecting flooding include aspects, slope, distance from stream, drainage density, lithology, land use, precipitation, soil texture, and topographic wetness index. The MaxEnt model subsequently predicted high-risk areas with exceptional accuracy (AUC = 0.945 for training; 0.906 for validation), while the Borda method ranked the sub-watersheds through parameter weighting. According to the findings, flood vulnerability was most influenced by distance from streams-30.9%-then by slope at 23.2%. The most important parameter found, based on Borda method results from the game theory model, was maximum 24-h rainfall with a 25-year return time. Following this were parameters of agricultural land usage and the average slope %. Sub-basin code 2221 ranked highest in choosing and prioritizing important sub-watersheds depending on flood susceptibility inside the Kashkan basin. The unprecedented integration of MaxEnt and the Borda method provides a quantitative-qualitative strategy for flood assessment that overcomes the limitations of single-model approaches. Proposed solutions include the construction of sedimentation basins in Sub-basin 2221, the reinforcement of the channel walls in Sub-basin 2222, and the implementation of flood spreading projects in Sub-basin 2223. The integration of the MaxEnt model with game theory represents a strategic innovation in risk analysis and complex decision-making. This approach combines quantitative risk assessment data with competitive strategies and collective decision-making processes, enabling managers and policymakers to adopt optimized, coordinated strategies against natural threats such as floods, based on scientific evidence.