Utilization of GIS And HEC-HMS Hydrological Model to Forecasting of Peak Discharge of the Air Bengkulu Sub-Catchments

Abstract

Introduction: The increasing frequency and intensity of floods in Bengkulu's watersheds are driven by land-use changes and climate variability, prompting the need for accurate flood discharge estimation, especially in ungauged basins.

Objectives: This study aims to estimate peak discharge using the integration of the HEC-HMS hydrological model and the SCS-CN method, supported by GIS-based spatial data analysis, in the Bengkulu Hilir sub-watershed.

Methods: The research employs a quantitative modeling approach with five stages: data collection and preprocessing, rainfall distribution analysis (via PSA 007 and Log Pearson Type III), watershed parameterization using ArcMap, model calibration with NSE and RMSE metrics, and simulation using return periods (2-100 years). Synthetic Unit Hydrograph (SUH) models were evaluated to support hydrograph development under limited streamflow data conditions.

Results: This study integrates the HEC-HMS hydrological model with the Soil Conservation Service Curve Number (SCS-CN) method and GIS-based spatial analysis to forecast peak discharge in the Bengkulu Hilir sub-watershed. Sixteen years of rainfall data (2007–2022) were analyzed using frequency distribution techniques, selecting Log Pearson Type III as the optimal model. The calibrated HEC-HMS model demonstrated superior performance, achieving Nash-Sutcliffe Efficiency (NSE) values up to 0.98 and low Root Mean Squared Error (RMSE). Simulations revealed a substantial increase in peak discharge, from 29.27 m³/s for the 2-year return period to 118.87 m³/s for the 100-year event, with peak discharge consistently occurring at the 13th hour. GIS spatial analysis delineated over 1,800 hectares of flood-prone areas in urban and lowland zones of Bengkulu, emphasizing the urgency for targeted flood mitigation and adaptive planning.

Conclusions: This study demonstrates that integrating HEC-HMS, the SCS-CN method, and GIS spatial analysis provides a reliable approach for estimating peak discharge in the Bengkulu Hilir sub-watershed. The model showed strong performance (NSE up to 0.98), with peak discharges rising from 29.27 m³/s (2-year) to 118.87 m³/s (100-year). The identification of over 1,800 hectares of flood-prone areas highlights the urgency for focused flood risk management. This framework offers a robust and adaptable tool for flood forecasting in data-limited tropical catchments.