Flood Hazard Mapping in a Data-Scarce Basing: Case Study of the Lakitan River, Indonesia

Abstract

Flood hazard maps are a basic instrument of disaster risk management, yet their preparation normally assumes that streamflow, water level and inundation records are available for model calibration. This study therefore aims to establish a flood hazard mapping framework for data-scarce basins, and to examine how far the resulting classification depends on the assessment framework adopted. GPM satellite rainfall, DEMNAS topography, HWSD soil and MoEF land cover were combined with hydrological (HEC-HMS) and two-dimensional hydraulic (HEC-RAS) modeling to simulate the 2-, 5-, 10-, 25- and 50-year return periods. The GPM rainfall was bias-corrected against the BMKG Tugumulyo station and two hydrological configurations were run in parallel in place of calibration, with the simulated inundation verified against observed flood extents. Hazard was then classified under the BNPB, DEFRA and Australian Government frameworks. For the 25-year return period DEFRA produced the largest extreme hazard zone (467.74 ha), followed by BNPB (384.39 ha) and the Australian Government model (323.87 ha). All three-concentrate extreme hazard along the riverbanks, but their areas are not directly comparable because of differences in parameters and thresholds. Credible flood hazard mapping is therefore attainable without in-situ calibration data, while the BNPB framework still requires clearer definitions of the inundation threshold and of how the flood area and frequency indices are applied spatially.