Using Polsar to Assess Temporal Trends of Water Hyacinth Infestation in Lake Victoria

Water hyacinth (WH) is a dreadful invasive species in Lake Victoria which spreads and forms vast floating mats on the lake. In-situ monitoring of its spread is costly and time consuming compared to remote sensing-based techniques. In this research, the temporal variability of WH in Winam Gulf of Lake Victoria is assessed using Polarimetric SAR data by applying the Optimization of Power Difference (OPDiff) change detection technique. Sentinel-1 IW SLC data acquired between 2017 and 2022 were analyzed using their covariance matrices generated from the VV, VH and VV∗VH. The detector separated WH from clear water using the minimum eigenvalues which enable rejection of false alarm from polarized scattering (e.g., water waves and wind) while selecting contribution from unpolarized scattering. The average precision and accuracy were 92.6% & 89.9% depending on WH density. The algorithms allowed quantification of temporal monthly and annual trends which established that highest WH amounts were recorded in 2018 within the last 4 months of the year.