Flood Risk Analysis Based on the Integration of Multi-Dimensional Hazard and Vulnerability Levels in Sitiarjo Village

Abstract

Sitiarjo Village, Malang Regency, has previously only been assessed for risk at a macro-resolution, thus spatial heterogeneity down to the neighborhood unit (RT) level has not been captured. This study maps micro-scale flood risk at the neighborhood unit level using a Spatial Decision Support System (SDSS) instrument. Inundation hazards from the Frequency Ratio-Logistic Regression model (AUC = 0.926) were combined with vulnerabilities derived from five socio-demographic and physical variables weighted by PCA, then classified using Jenk's Natural Break into five risk levels. Of the 65 neighborhood unit (RT) polygons, hazards were concentrated along the riverbanks of Dusun Krajan, while vulnerability was more evenly distributed and driven by socio-physical factors in Dusun Rowotrate. RT 24 RW 05 recorded the highest composite risk score of 0.9728. High and Very High risk zones covered only 5.83% of the study area, but coincided with densely populated residential cores. Based on the dominant variables of FR-LR and PCA, this study proposes short to long-term mitigation, including vertical evacuation sites, riparian restoration, and community-based evacuation support, as replicable SDSS instruments for flood risk management.