Flood identification in Jakarta (2018 and 2020): a combined InSAR coherence and rapid flood mapping

Abstract

Jakarta has experienced a land subsidence rate of 3–10 cm per year, intensifying existing environmental pressures that gradually shape long-termhydrological responses in the region. Additionally, low infiltration factors, soil compaction, poor drainage systems, and rising sea levels resulted in an increase in the quantity of flooding in Jakarta, affecting the area of inundation, height of the floodwaters, and duration of flooding in the region. Consequently, flooding continued to recur during the rainy period with increasing severity. Based on these facts, efforts are needed to minimize or address the areas affected by flooding in Jakarta, one of which involves flood identification through the utilization of active remote sensing technology. We utilized Sentinel-1A to generate coherence values based on the Interferometric Synthetic Aperture Radar (InSAR) coherence method. Additionally, we applied Rapid Flood Mapping based on backscatter values. Our results indicate that the image acquisition period significantly dictates the outcomes, with the image pairs of “dry_rain” proving more effective than “rain_rain” in Jakarta’s flood identification. Furthermore, integrating results from both methods improves flood identification, as they can validate each other. Specifically, the mapped flood extent was influenced by rainfall intensity and the spatial distribution of validation points within the presumed flood zones.