HYDROLOGICAL MODELING OF PEAK DISCHARGE REDUCTION THROUGH DISTRIBUTED STORAGE SYSTEMS IN IRRIGATED CATCHMENTS

Nurnawaty Nurnawaty (Universitas Muhammadiyah Makassar)

Nusantara Hasana Journal
Vol. 3 No. 8 (2024): Nusantara Hasana Journal, January 2024

Accurately simulating flood attenuation across heavily modified agricultural landscapes remains a major challenge in large-sample hydrology. This study evaluates a coupled distributed hydrological-hydrodynamic framework across 47 irrigated catchments to assess peak-flow reductions and temporal delays induced by distributed farm dams. Local model calibration via the L-BFGS-B algorithm yielded a mean Kling-Gupta Efficiency () of 0.1657, revealing systematic underestimation () and high runoff variability (). Scenario simulations show that antecedent storage capacity strongly controls flood mitigation; under passive operation, empty reservoirs achieved a mean Peak Discharge Reduction (PDR) of 26.227%, while full reservoirs achieved a mean PDR of 6.084%. Integrating predictive active management using Quantitative Precipitation Forecasts (QPF) maximizes flood retention, generating net temporal gains of up to +4.70 hours and extending lag times to 8.20 hours. Spatial analysis confirmed that maximum attenuation (>40%) clusters in lowland alluvial floodplains with dense reservoir networks. Accurately simulating flood attenuation across heavily modified agricultural landscapes remains a major challenge in large-sample hydrology. This study evaluates a coupled distributed hydrological-hydrodynamic framework across 47 irrigated catchments to assess peak-flow reductions and temporal delays induced by distributed farm dams. Local model calibration via the L-BFGS-B algorithm yielded a mean Kling-Gupta Efficiency () of 0.1657, revealing systematic underestimation () and high runoff variability (). Scenario simulations show that antecedent storage capacity strongly controls flood mitigation; under passive operation, empty reservoirs achieved a mean Peak Discharge Reduction (PDR) of 26.227%, while full reservoirs achieved a mean PDR of 6.084%. Integrating predictive active management using Quantitative Precipitation Forecasts (QPF) maximizes flood retention, generating net temporal gains of up to +4.70 hours and extending lag times to 8.20 hours. Spatial analysis confirmed that maximum attenuation (>40%) clusters in lowland alluvial floodplains with dense reservoir networks.