A Methodological Framework for Flood Susceptibility Assessment Based on GIS, Multi-Criteria Decision Analysis and Artificial Intelligence – Case Study Kosovo

Flood susceptibility assessment has become a critical component of spatial planning and disaster susceptibility reduction, particularly under increasing climate variability and rapid urbanization. This study presents a methodological framework for flood susceptibility assessment based on the integration of Geographic Information Systems (GIS), Multi-Criteria Decision Analysis (MCDA), and Artificial Intelligence (AI), using Kosovo as a case study. The framework integrates topographic, hydrological, climatic, pedological, and anthropogenic factors to delineate flood-prone areas at high spatial resolution. A 5 m Digital Elevation Model was used to derive slope, flow accumulation, drainage networks, distance from rivers, and the Topographic Wetness Index (TWI). Extreme rainfall conditions were incorporated using percentile-based scenarios (P95 and P99 for 24-hour events) derived from ERA5 precipitation data. Multi-criteria decision analysis techniques, such as the weighted linear combination method, are widely applied in environmental susceptibility assessment due to their flexibility and transparency (Malczewski, 2006). Artificial Intelligence was employed in an AI-assisted framework to support parameter weighting and interpretation of flood patterns. The final susceptibility map classified Kosovo into five susceptibility levels and demonstrated strong agreement with observed flood events from 2025–2026. The proposed framework provides a transferable decision-support tool for flood susceptibility management, spatial planning, and climate adaptation strategies in practice.