Exploring Google Earth Engine for Flood Detection (A Case Study in Bandung City)

Climate change has increased the frequency and severity of urban flooding worldwide. This research utilizes Google Earth Engine (GEE) and Sentinel-1 synthetic aperture radar (SAR) imagery to detect, map, and analyze flood inundation in Bandung City, Indonesia, from 2014 to 2023. Our workflow combines radiometric calibration, Lee speckle filtering, and Otsu thresholding implemented through GEE's JavaScript API to delineate flooded areas in all weather conditions at 30 m resolution. The results show clear spatial and temporal fluctuations in inundation levels, with peaks in flood coverage in 2016, 2020 and 2021. This approach identifies recurring inundation points and supports targeted disaster management interventions such as prioritized drainage improvements, flood forecasting systems, and nature-based solutions to improve flood resilience in Bandung City by rapidly processing a decade of data in the cloud.