| 引用本文: | 李健锋,张宗科,魏显虎,等.基于Google Earth Engine和Sentinel-1/2的洪水信息快速提取模型研究[J].灌溉排水学报,2020,(Supp.2):-. |
| lijianfeng,Zhang Zongke,Wei Xianhu,et al.基于Google Earth Engine和Sentinel-1/2的洪水信息快速提取模型研究[J].灌溉排水学报,2020,(Supp.2):-. |
|
| |
|
|
| 本文已被：浏览 2777次 下载 0次 |
|
|
|
| 基于Google Earth Engine和Sentinel-1/2的洪水信息快速提取模型研究 |
|
李健锋,张宗科,魏显虎,等
|
|
1.陕西地建土地工程技术研究院有限责任公司;2.中国科学院空天信息创新研究院;3.中国科学院中国-斯里兰卡水技术研究与示范联合中心;4.陕西省土地工程建设集团有限责任公司;5.中国科学院地理科学与资源研究所资源与环境信息系统国家重点实验室
|
|
| 摘要: |
| 洪水灾害是威胁人类生存和经济发展的十大自然灾害之一,快速准确地获取洪水信息对洪水灾情评估、救援以及资源调配有重要的现实意义。本文基于Google Earth Engine云平台,以斯里兰卡中东部为研究区,从目视判读与定量分析两个角度对比分析了大津法(OTSU)-阈值模型提取Sentinel-1影像水体和OTSU-归一化水体指数(NDWI)模型提取Sentinel-2影像水体的精度,发现有云区域Sentinel-1影像的提取精度明显高于Sentinel-2影像,能有效地避免云和云阴影对水体提取结果的影响；无云区域Sentinel-2影像的提取结果较好,能比较准确地提取出水体的边界；两种影像均存在山体阴影误分为水体的现象。以此为基础,结合地形建模后的山体阴影掩膜数据,构建了基于Google Earth Engine和Sentinel-1/2影像的洪水信息快速提取模型,并应用于2017年5月斯里兰卡马塔拉区和2020年7月鄱阳湖区的特大洪灾分析。研究结果表明该模型效率高、可行性强、零成本,有效地避免了耗时的影像下载以及预处理操作,能近实时的进行洪水淹没范围测绘,可广泛应用于洪灾研究中。 |
| 关键词: Google Earth Engine;洪水灾害;Sentinel-1/2;快速提取;OTSU |
| DOI： |
| 分类号:TP79 |
| 基金项目:中国科学院战略性先导科技专项（XDA2003030201）；天津科技计划项目智能制造专项(自主控制无人机智能组网观测与应用技术)；中国科学院中国-斯里兰卡水技术研究与示范联合中心项目。 |
|
| Research on Quick Extraction Model of Flood Information based onGoogle Earth Engine and Sentinel-1/2 |
|
lijianfeng1,2,3,4, Zhang Zongke2,5, Wei Xianhu2,5, Yang Liangyan1,4, Ye Huping6,7
|
|
1.Institute of Land Engineering and Technology, Shaanxi Provincial Land Engineering Construction Group Co., Ltd.;2.Aerospace Information Research Institute, Chinese Academy of Sciences;3.China-Sri Lanka Joint Research and Demonstration Center for Water Technology, Chinese Academy of Sciences, Beijing 100085, China;4.Shaanxi Provincial Land Engineering Construction Group Co., Ltd.;5.China-Sri Lanka Joint Research and Demonstration Center for Water Technology, Chinese Academy of Sciences.;6.State Key Laboratory of Resources and Environmental Information System, Institute of Geographic Sciences and Natural Resources Research, Chinese Academy of Sciences;7.China-Sri Lanka Joint Research and Demonstration Center for Water Technology, Chinese Academy of Sciences
|
| Abstract: |
| Flood disaster is one of the ten natural disasters threatening human survival and economic development. Obtaining flood information quickly and accurately has important practical significance for flood disaster assessment, rescue and resource allocation. Taking the east-central Sri Lanka as the study area, this paper based on the Google Earth Engine, the accuracy of the Sentinel-1 image water body extracted by OTSU-threshold model and the Sentinel-2 image water body extracted by OTSU-NDWI model were compared from visual interpretation and quantitative analysis. Through the analysis, it was found that the extraction accuracy of Sentinel-1 image was significantly higher than that in Sentinel-2 image in cloudy area, which can effectively avoid the influence of cloud and cloud shadow on the water extraction result; the extraction result of Sentinel-2 image was better in cloudless area, and the boundary of the water can be extracted relatively accurately; both images mistakenly divided some mountain shadows into water bodies. On this basis, this paper combined with the mountain shadow mask data obtained by terrain modeling, a quick flood information extraction model based on Google Earth Engine and Sentinel-1/2 images was constructed. The model was successfully applied to the analysis of severe flood in Matala District of Sri Lanka in May 2017 and Poyang Lake in July 2020. The results show that the model has high efficiency, strong feasibility, and zero cost, and effectively avoids time-consuming image download and pre-processing. It can extract the flooded area in near real time, which can be widely used in flood disaster research. |
| Key words: Google Earth Engine; flood disaster; Sentinel-1/2; quick extraction; OTSU |
|
|
|
|
