Integration hierarchical analysis process (AHP) and geographic information systems to map flood risk levels in the province of Utcubamba, Peru

Abstract

Floods are one of the most common and destructive natural disasters globally, causing significant economic losses and loss of life. In Peru, this problem is particularly critical, affecting nearly 1.51 million people living in flood-prone areas. This study aims to identify and map flood-prone areas in the province of Utcubamba, Amazonas, Peru, using a semi-quantitative approach that integrates the Analytic Hierarchy Process (AHP) with Geographic Information System (GIS) tools and Weighted Linear Combination (WLC). Fifteen conditioning variables grouped into five factors—topographic, geomorphological, hydrological, climatic, and anthropogenic—were selected, weighted, and spatially combined. The AHP consistency ratio (CR = 0.076) confirmed the methodological validity of the model. Results indicate that approximately 401.39 km 2 (10.09%) of the provincial territory presents very high flood risk, concentrated primarily in Bagua Grande, the most populated city in Amazonas, while 824.25 km 2 (22.47%) correspond to high risk. Model validation was conducted using two complementary approaches: (1) the Relative Mean Error (RME) and Root Mean Square Error (RMSE) method applied to 585 georeferenced flood occurrence records from the INDECI/SINPAD national emergency database, yielding RME = 12.82% (good performance) and RMSE = 0.7845; and (2) Receiver Operating Characteristic (ROC) curve analysis, which produced an Area Under the Curve (AUC) of 0.816 (95% CI: 0.721–0.895), confirming good discriminatory capacity of the model. These results underscore the urgent need for territorial planning and disaster prevention measures in the region, while demonstrating the robustness and replicability of the AHP-GIS methodology for flood risk assessment in data-scarce tropical regions.