基于Sentinel-1A的2020年鄱阳湖流域洪水灾害遥感监测  

Remote Sensing Monitoring of Poyang Lake Flood Disaster in 2020 Based on Sentinel-1A

作　　者：王磊 连增增[2] WANG Lei;LIAN Zengzeng(The First Geological Team of Henan Provincial Non-Ferrous Metal Geological and Mineral Resources Bureau,Zhengzhou 450000,China;School of Surveying and Land Information Engineering of Henan Polytechnic University,Jiaozuo 454000,China)

机构地区：[1]河南省有色金属地质矿产局第一地质大队,河南郑州450000 [2]河南理工大学测绘与国土信息工程学院,河南焦作454000

出　　处：《地理空间信息》2022年第6期43-46,共4页Geospatial Information

基　　金：河南省自然科学基金资助项目(202300410180);河南省高校基本科研业务费专项资金资助项目(NSFRF170807)。

摘　　要：基于Sentinel-1ASAR影像在洪水灾害监测方面的优越性,分别采用监督分类(SVM法)和非监督分类(IsoData法)2种方法对鄱阳湖流域开展灾前、灾后洪涝范围提取,以期为Sentinel-1数据在洪涝灾害监测及灾后评估方面的应用提供参考。结果表明,基于SVM与IsoData方法进行水体提取的总体精度均达到90%以上,Kappa系数均大于0.8,满足洪水变化监测需求,但SVM表现出更为稳定的水体信息提取能力。研究区灾后的洪水面积大约增加1 432 km^(2),受灾较为严重的区域位于鄱阳湖主体水面西南部及地势相对平坦的低洼区域。利用合成孔径雷达SAR影像能够有效获取地表水体变化信息,可用于洪水易发区的洪涝灾害灾情分析。Based on Sentinel-1A SAR image, we carried out pre-disaster and post-disaster flood inundation monitoring in Poyang Lake region.SVM and IsoData methods were used to extract water information of Poyang Lake, and the accuracy of the extracted results was evaluated. Finally, spatial analysis was made on the changes of Poyang Lake water before and after the disaster. The results show that the water body extraction based on SVM and IsoData method can achieve high accuracy, the overall classification accuracy is more than 90%, and the Kappa coefficient is greater than 0.8, which meets flood change monitoring requirements. The Poyang Lake area water structure extracted from SAR data is very clear. Compared with May 9, 2020, the water area increased by 1 432 km^(2). The submerged areas were mainly concentrated in the southwest of Poyang Lake, and most of the beaches around the lake were submerged. Sentinel-1A series data has great application potential in flood disaster monitoring.

关 键 词：洪水监测 Sentinel-1A 鄱阳湖 SVM ISODATA 

分 类 号：P237[天文地球—摄影测量与遥感]