Long-term river extent dynamics and transition detection using remote sensing: Case studies of Mekong and Ganga River

Mohammad Adil Aman,Hone-Jay Chu

DOI: https://doi.org/10.1016/j.scitotenv.2023.162774

2023-06-10

Abstract:River dynamics are currently comprehensively studied at either a bankline or reach-scale level. Monitoring large-scale and long-term river extent dynamics provides fundamental insights relevant to the impact of climatic factors and anthropogenic activities on fluvial geomorphology. This study analyzed the two most populous rivers, Ganga and Mekong, to understand the river extent dynamics using 32 years of Landsat satellite data (1990-2022) in a cloud computing platform. This study categorizes river dynamics and transitions using the combination of pixel-wise water frequency and temporal trends. This approach can demarcate the river channel stability, areas affected by erosion and sedimentation, and the seasonal transitions in the river. The results illustrate that the Ganga river channel is found to be relatively unstable and very prone to meandering and migration as almost 40 % of the river channel has been altered in the past 32 years. The seasonal transitions, such as lost seasonal and seasonal to permanent changes are more prominent in the Ganga river, and the dominance of meandering and sedimentation in the lower course is also illustrated. In contrast, the Mekong river has a more stable course with erosion and sedimentation observed at sparse locations in the lower course. However, the lost seasonal and seasonal to permanent changes are also dominant in the Mekong river. Since 1990, Ganga and Mekong rivers have lost approximately 13.3 % and 4.7 % of their seasonal water respectively, as compared to the other transitions and categories. Factors such as climate change, floods, and man-made reservoirs could all be critical in triggering these morphological changes.