Assessment and Prediction of River Bank Shifting Using Automated GIS and Remote Sensing Approaches: A Case Study of the Jamuna River in Bangladesh.

Abstract

Being the biggest delta in the world, Bangladesh is home to the mighty Ganges-Brahmaputra-Meghna (GBM) Basin. The dynamic morphological changes in the basin shape the geography of the country. Jamuna, one of the largest sand-bed braided rivers globally, is the most dynamically evolving river in Bangladesh. With its incessant morphological changes and perpetual rise and disappearance of bars and dunes in the channel, Jamuna takes a new form every few years. Hence, understanding the river bank shifts along with erosion-accretion patters of Jamuna is salient in assessing the river’s impact on its floodplain, surrounding landscape and population in the context of aggravating climate change scenarios.  This study aims to assess the spatio-temporal changes of the Jamuna river banks for last four decades from December 1984 to January 2024 and predict the changes in 2034 and 2044. The research employed Google Earth Pro for manual digitization of bank lines with five-years intervals and the Digital Shoreline Analysis System (DSAS) in ArcGIS for trend analysis and future prediction. The study encompasses the whole length of Jamuna river situated in Bangladesh and divided the zonal changes of erosion and accretion in district levels for analyses. 2500 and 2700 transects were taken for left and right banks respectively with a spacing interval of 100m. The rate of changes was analyzed based on Linear Regression Rate (LRR), Weighted Linear Regression Rate (WLR) and End Point Rate (EPR) and, distance measurements were derived from Net Shoreline Movement (NSM) and Shoreline Change Envelope (SCE). Subsequent predictions were obtained using simple extrapolation of the analyses data using Kalman Filter Model. Root Mean Square Error (RMSE), t-test and R2 were calculated to assess prediction accuracy for 2024 before utilizing the resultant data in forecasting analysis for next two decades. Additionally, net erosion and accretion area were determined for the study area relying on polygon-based analysis in ArcGIS. The results demonstrate that the rate of erosion and accretion fluctuates for the whole length of the river. Along the left bank, the average erosion rate is observed to be -42.93m/year and accretion rate is around 38.89m/year. For right bank, the average erosion rate is -59.20m/year with accretion rate being 55.61m/year. Both the banks have been shifting westward continuously and the right bank has been shifting more compared to the left bank. Bankline fluctuations are more prominent in the downstream closer to the conjunction of Jamuna and Ganges rivers in the central part of Bangladesh. The findings of the forecasting analysis showed that the alleviating rates of erosion will continue until 2041 for both banks. The quantitative nature and results of this study can be utilized to assess the extend of erosion control measures needed for the river. The historical trends and contemporary predictions of the study can be useful in further studies on the impacts that drastic river bank shifting can have on natural and anthropogenic factors surrounding the river channels.