Sử dụng ảnh Sentinel-1A đa thời gian để phát hiện lũ quét, thử nghiệm tại tỉnh Lào Cai

In recent years, Vietnam has frequently experienced severe flash floods that cause significant damage to both people and property.These flash floods often occur suddenly, accompanied by mud and rapidly rising floodwaters, making it challenging to conduct field surveys and access disaster sites and carry out recovery efforts.Currently, microwave satellite image is being applied in practice to monitor and assess the impacts of flood and flash flood areas, yielding better results compared to traditional methods.This study presents a method for detecting flash flood areas using Sentinel-1A imagery.Accordingly, 4 images were collected in IW mode, level 1 at the times before and after the flash flood.The experimental study area is Lao Cai province, where flash floods often occur causing heavy damage.The proposed method for extracting flash flood areas from Sentinel-1A images involves the following steps: (1) Data collection: Gather Sentinel-1A imagery and relevant data; (2) Image preprocessing: Apply preprocessing steps to the Synthetic Aperture Radar (SAR) images and crop them based on the study area boundaries; (3) Image analysis and image classification: compare and analyze the images to extract flash flood regions.The accuracy of the results was verified against ground-truth data.The research findings serve as valuable input for developing flash flood warning models and providing information to relevant authorities for effective disaster response and mitigation.