Wetland delineation simulation and prediction in deltaic landscape

Abstract

Abstract Present study attempts to delineate wetland in Indic part of mature Ganges delta with the help of satellite images applying convenient water body extraction indices. It is found that the existing methods are bringing noise in fulfilling the purpose and therefore a new index, Re-modified Normalized Difference Water Index (RmNDWI) is introduced. After proving the viability of the new index, it was used to simulate multiple land use changes and wetland area using Cellular Automata (CA) with Artificial Neural Network (ANN) along with other convenient indices. Results show a significant shrinkage of overall wetland area in both pre-monsoon and post-monsoon seasons, with complete disappearance of scattered wetland in pre-monsoon season in next ten to twenty years. Adaptive exponential smoothing method is used for simulating depth of wetland and results revealed that wetland depth may reduce by 70% in about >60% wetland areas in next 10 years. Some patches in the outskirts however exhibit increasing trend which is attributed by the local fishing activities.