Utilization of Multi-Temporal Sentinel-1 Satellite Imagery for Detecting Aquatic Vegetation Change in Lake Rawapening, Central Java, Indonesia

Abstract

Lake Rawapening has high ecological, historical, and economic value​​, such as in terms of agricultural irrigation, fisheries, hydropower generation, and tourism. However, a major problem faced by the lake is the uncontrolled growth of aquatic vegetation. Monitoring of the dynamics of the condition of the vegetation cover on the waters needs to be conducted periodically as one of the lake management efforts. We propose an Automatic Aquatic Vegetation Extraction method to monitor the dynamics of the condition of the lake's vegetation using multitemporal C-band radar satellite imagery. The method utilizes data from the SAR satellite imagery to ascertain the maximum water boundary, which can be used to distinguish between aquatic and terrestrial vegetation. The Otsu algorithm approach was used to determine the boundary between land and water areas. The method was applied on several dates, employing VV and VH polarizations. The results show that the proposed method could rapidly monitor lakes and their aquatic vegetation from year to year. The overall accuracy of the study varied from 79.48 to 88.46 percent, with an average of 84.4 percent.