Natural resource mapping using hybrid classification approach: Case study of Cooch Behar District, West Bengal

An ever increasing population associated with various other factors has put tremendous pressure on the environment and its respective resources. Hence it has become necessary to study the human interactions with its surrounding environment to facilitate sustainable developmental plans. Land use and land cover is the most dominant factor to study these interactions. Remote Sensing technology has become an indispensable tool to evaluate the environmental processes and provide adequate information of available resources and to design strategies for the sustainable use of these resources. In the current study an attempt has been made to study the land use and land cover in the Cooch Behar district by mapping the natural resources of the area. The study has been made by integrating Geographical Information System (GIS) approach with Hybrid classification using District Planning Map and multispectral Indian Remote Sensing-P6 - LISS_III data for 2008 and 2009 based on an emerging digital classification technique. The land use land cover classification was accomplished on the basis of prior knowledge of the study area instead of the traditional classification schemes and the satellite imageries. A hybrid supervised/unsupervised classification approach coupled with GIS analyses was employed to generate land use/cover maps with eight classes; upper terrace, single crop, double crop, forest, degraded forest, rivers and water bodies, sandy area and settlements. The study highlights the usage of a non-conventional classification scheme purely based on researcher's in-depth knowledge of study area and application of a hybrid classification technique for far more accurate land use mapping.