Deriving wetland-cover types (WCTs) from integration of multispectral indices based on Earth observation data

The wetland cover is defined as the spatially homogenous region of a wetland attributed to the underlying biophysical conditions such as vegetation, turbidity, hydric soil, and the amount of water. Here, we present a novel method to derive the wetland-cover types (WCTs) combining three commonly used multispectral indices, NDVI, MNDWI, and NDTI, in three large Ramsar wetlands located in different geomorphic and climatic settings across India. These wetlands include the Kaabar Tal, a floodplain wetland in east Ganga Plains, Chilika Lagoon, a coastal wetland in eastern India, and Nal Sarovar in semi-arid western India. The novelty of our approach is that the derived WCTs are stable in space and time, and therefore, a given WCT across different wetlands or within different zones of a large wetland will imply similar underlying biophysical attributes. The WCTs can therefore provide a novel tool for monitoring and change detection of wetland cover types. We have automated the proposed WCT algorithm using the Google Earth Engine (GEE) environment and by developing ArcGIS tools. The method can be implemented on any wetland and using any multispectral imagery dataset with visible and NIR bands. The proposed methodology is simple yet robust and easy to implement and, therefore, holds significant importance in wetland monitoring and management.