Modeling flood plain wetland transformation in consequences of flow alteration in Punarbhaba river in India and Bangladesh

Ecologically precious and economically remunerative wetland resources in the riparian flood plain of Punarbhaba river basin are under massive transformation in post hydrological alteration phase triggered mainly by damming. It affects the people who are directly and indirectly depends on these resources. The present work intends to measure the degree and direction of wetland transformation in reference to the river flow change using multi-date satellite images. For addressing this issue, modeling of river flow change, wetland area, wetland depth, wetland water appearance consistency within wetland and landscape transformations have been done in pre and post dam periods (dam over Punarbhaba river is installed on 1992) and future state is also predicted. Modeling wetland transformation using advance geo-spatial and machine learning techniques like combined cellular automata-markov chain, water presence frequency approach, spatial nonlinear regression (adaptive spline) is quite new in such data sparse condition. NDWI intensity transformation modeling is a novel approach and it is successfully applied in this work. Few major findings of this works are (1) river flow is declined in post dam phase and further predicted to be declined in coming days (2) reduction of peak flow has reduced wetland area by 34.52% in post dam period and predicted to be decreased by 22.66% by 2030 (3) NDWI intensity of wetland is also declined by 37.87% and it is caused for drying out 40.26 km(2) shallower wetland. Simulated trend shows further decline of wetland depth in predicted model. (4) 10.56% highly consistent wetland is transformed into inconsistent wetland in terms of water presence (5) encroachment of agriculture land and infrastructural development have enhanced the degree of wetland fragmentation. (C) 2020 Elsevier Ltd. All rights reserved.