25. Night-Time Light Remote Sensing and Google Earth Engine for flood impact mapping: A case study of the Cat Ba island

Abstract

Natural hazards such as typhoons, floods, and other extreme weather events often cause severe damage to infrastructure and human livelihoods, necessitating rapid and accurate assessment methods to support disaster response and recovery. In this context, Night-Time Light (NTL) data have emerged as an effective tool for monitoring the impacts of natural hazards over large areas. This study examines the case of Typhoon Yagi in 2024, which caused widespread power outages across Cat Ba island. Using daily NTL data from the VIIRS Black Marble product (VNP46A2) on the Google Earth Engine (GEE) cloud platform, the Normalized Change Ratio (NCR) was computed at the pixel level for the pre-event (01 - 10 August 2024) and post-event (20 - 30 September 2024) periods, and validated against flood extent data derived from Sentinel-1 SAR imagery. Results indicate that the area classified as very low illumination increased by 26.5 %, while the “high” and “very high” categories decreased by over 60 %. The most severely affected areas were Cat Ba, Phu Long, and Viet Hai. Notably, Western Phu Long recorded the lowest NCR value (-0.83 nW·cm⁻²·sr⁻¹), coinciding with the deepest inundation zones. These findings highlight the potential of integrating NTL data with GEE for rapid post-disaster impact assessment, enabling the prioritization of response efforts in coastal and island regions.