基于Sentinel-1号雷达影像的2020年9月18日湖北黄冈龙感湖洪涝灾害灾中洪水分布范围数据集

及时、准确、科学地确定洪涝淹没范围,是防汛应急监测的重要内容。在灾前灾区正常水体的分布的基础上，通过对洪涝灾害发生过程中洪水分布范围的动态监测，可为洪涝灾害的演进和可能造成的损失提供科学的信息支撑。遥感数据作为洪涝灾害洪水淹没范围动态监测重要信息来源, 能够直接用于提取洪涝发生过程中的水体的范围信息。湖北黄冈龙感湖洪涝灾害灾中（2020.9.18）洪水分布范围主要利用从USGS下载的哨兵1号雷达数据，通过欧空局提供的SNAP 7.0软件对雷达数据进行预处理，具体预处理步骤包括轨道校正、热噪声去除、辐射定标、相干斑滤波、地形校正、分贝化、裁剪等处理，利用UNet深度学习网络快速提取灾中（2020年9月18）卫星影像中洪涝水体分布的结果。

Timely, accurate and scientific determination of flood inundation extents is a critical component of flood control emergency monitoring. Based on the distribution of normal water bodies in the pre-disaster affected area, dynamic monitoring of the flood distribution extent during flood disasters can provide scientific information support for flood disaster evolution and potential resulting losses. Remote sensing data, as an important information source for dynamic monitoring of flood inundation extents in flood disasters, can be directly used to extract water body range information during flood events. For the flood distribution extent of the Longganhu Lake flood disaster in Huanggang, Hubei on September 18, 2020 during the disaster, Sentinel-1 radar data downloaded from USGS was primarily employed. The radar data was preprocessed using SNAP 7.0 software provided by the European Space Agency (ESA), with specific preprocessing steps including orbital correction, thermal noise removal, radiometric calibration, speckle filtering, topographic correction, decibel conversion, and clipping. The UNet deep learning network was utilized to rapidly extract the flood water body distribution results from the satellite image captured on September 18, 2020 during the disaster.

搜集汇总

背景与挑战

背景概述

该数据集基于Sentinel-1雷达影像，利用UNet深度学习网络提取了2020年9月18日湖北黄冈龙感湖洪涝灾害期间的洪水分布范围，旨在为灾害应急监测和损失评估提供科学支持。数据以10米分辨率的矢量格式提供，经过预处理和质量评估，提取精度超过95%，具有高时效性和准确性。数据集来源于国家重点研发计划项目，由中国科学院空天信息创新研究院制作，适用于地理信息系统分析和遥感应用。