Heterogeneous landscape assessment using multi-temporal LISS4 and hybrid approach: A case study in Barak Valley, India

Remotely sensed data of landscapes characterized by diverse interacting patches or ecosystems displays greater degree of mixed pixels making it strenuous to accurately delineate the different land components. The present study aims to analyse the spatial characteristics of such a landscape (Barak valley) using multi-temporal IRS-R2 LISS4 mosaic data and a hybrid classification approach. Agriculture was the predominant land use type with an aerial extent of 2457.3 km² (35.5%). Riverbank farming land is a significant land use type in the region that is spatially assessed and mapped for the first time in the study. This land use type covered a total area of 16.27km².Anoverallmapclassificationaccuracy of 91.1% is achieved. The study approach adopted may be applied for mapping of similar landscapes consisting of small land use land cover components. The final map would serve as vital baseline data for management, conservation, research activities and monitoring of different land resources of Barak valley.