High-frequency Monitoring of Inland Lakes Water Extent Using Time-Series Sentinel-1 SAR Data

Abstract

Poyang Lake and Dongting Lake are the largest freshwater lakes in China. They are also important for regulating and storing flood in the middle and lower basin of the Yangtze River. In this paper, high-frequency water extraction and inundation mapping using time-series of Sentinel-1 synthetic aperture radar (SAR) images on the Google Earth Engine cloud computing platform is proposed. A monthly analysis of water area variation in the two lakes from June 2015 to March 2019 is analyzed. The results show that: (1) The area variation is seasonal. The minimum monthly water area of Poyang Lake during the monitoring period is 2785 km2 in February 2017, and the maximum monthly water area is 4989 km2 in July 2016. For Dongting Lake, the minimum monthly water area during the monitoring period is 1870 km2 in December 2015, and the maximum is 3447 km2 in July 2016. (2) The variation trend of water area was consistent between the two lakes. (3) The inundation frequency of two lakes in recent years, especially the long-term inundation of water area remains stable. This means the lake degradation has been curbed effectively.