<![CDATA[Flooding in the Bekasi River Basin, West Java, Indonesia, is now more frequent and destructive. This is due to rapid urbanisation, land‑cover change, and intensified rainfall. Conventional early warning systems rely on sparse rain‑gauge networks and manual reporting. This leads to limited coverage and delayed responses, such as during the March 3–4, 2025 flood event. To address these issues, this study presents a real‑time, satellite‑enabled flood forecasting framework. It integrates precipitation estimates from GSMaP with the Universitas Pertamina Rainfall–Runoff Model. Near‑real‑time GSMaP_NOW and GSMaP_NRT products, at 0.1° spatial resolution and hourly intervals, are assimilated into a semi‑distributed hydrological model. This model runs continuously to generate 24‑hour river discharge forecasts at multiple monitoring locations. When applied to the March 2025 flood event, the simulated discharge captured the timing of flood onset and escalation. Alert Level‑1 thresholds were identified about one hour in advance at Cileungsi station. At the watershed scale, upstream rainfall detection gave a five to six-hour lead time before downstream flooding. These results show that satellite‑based rainfall monitoring combined with real‑time hydrological modelling can support operational flood warning in data‑sparse urban watersheds.]]>
Copyrights © 2026
