首页 >  , Vol. , Issue () : -

摘要

全文摘要次数: 274 全文下载次数: 335
引用本文:

DOI:

10.11834/jrs.20210547

收稿日期:

2020-12-03

修改日期:

2021-04-26

PDF Free   EndNote   BibTeX
时序Sentinel-1A数据支持的长江中下游汛情动态监测
郭山川1, 杜培军1, 蒙亚平1, 王欣1, 唐鹏飞1, 林聪1, 夏俊士2
1.南京大学 地理与海洋科学学院;2.日本理化学研究所 革新智能研究所
摘要:

合成孔径雷达(SAR)因其对地观测全天候、全天时优势,成为多云多雨天气限制下洪水动态监测中不可或缺的数据来源之一。由于Google Earth Engine (GEE)云计算平台的兴起和短重访Sentinel-1数据的可获取性,洪水监测与灾害评估目前正面向动态化、广域化快速发展。顾及洪水淹没区土地覆盖变化的复杂性和发生时间的不确定性,以时序Sentinel-1A数据处理与分析为基础,提出了针对大尺度范围、连续长期的汛情动态监测方法。该方法首先利用图像二值化分割时序SAR数据实现水体时空分布粗制图,逐像素计算时间序列中被识别为水体候选点的频率。利用Sentinel-2光学影像对精度较粗的初期SAR水体提取结果进行校正,得到精细的水体分布图。针对不同频率区间的淹没特点,采用差异化的时序异常检测策略识别淹没范围:对低频覆水区利用欧氏距离检测时序断点,以提取扰动强度大、淹没时间短的洪涝灾害区;对高频覆水区利用标准分数(Z-Score)检测时序断点,以提取持续淹没区。在GEE平台上利用该方法,实现了2020年5月–10月长江中下游地区全域洪水淹没范围时空信息的自动、快速、有效监测,揭示了不同区域汛情发展模式的差异性。本文提出的洪水快速监测方法对大尺度下的汛情动态监测、灾害定量评估和快速预警响应具有重要的现实意义。

Dynamic monitoring on flooding situation in the Middle and Lower Reaches of the Yangtze River Region using Sentinel-1A time series
Abstract:

Objective: Synthetic Aperture Radar (SAR) is one of the indispensable data sources for the dynamic monitoring of flood events due to the capacity for all-weather and all-time sensing. Owing to the emergence of cloud computing platforms like the Google Earth Engine (GEE) and accessibility of short-revisit Sentinel-1A radar data, flooding mapping and damaging assessment presently are developing rapidly towards dynamic monitoring and large scale. Considering the complexity of land cover changes within the flooded areas and the temporal uncertainty of flood events, the objective of this study is to propose a novel framework for continuously monitoring the floods at a large scale based on the data processing and analysis from Sentinel-1A time series. Method: Firstly, the coarse maps of spatiotemporal distribution of water body were generated by the binary segmentation of SAR time series. Accordingly, the frequency of candidate water pixels was computed from these coarse maps. The first fine water map in the temporal sequence was corrected by the Sentinel-2 optical image. In view of the distinctive inundation characteristics of different frequency area, differentiated sequential anomaly detection strategy was adopted to identify inundated area. Euclidean distance was used to detect the sequential breakpoint in low frequency water-covered area, for identifying the flooding disaster area with high disturbance intensity and short flooding time. Temporal Z-score was introduced to detect the breakpoint in high frequency water-covered area, for identifying the continuous inundating area. Result: Using the proposed framework on the GEE, we automatically, rapidly and effectively detected the flood events in the Middle and Lower Reaches of the Yangtze River Region between May and October in 2020. The results demonstrated the reliability and accuracy of water mapping retrieved by our method. The spatiotemporal information of inundation areas was collected accurately and promptly, and the differences of flooding patterns in different regions were also revealed. Conclusion: The rapid and robust flood monitoring method proposed in this study is of great practical significance to the dynamic monitoring of flood situation, quantitative assessment of flood disaster and rapid early warning response.

本文暂时没有被引用!

欢迎关注学报微信

遥感学报交流群 分享按钮