Hotspots of Land Use Land Cover Change: World’s Largest Riverine Island Scenario

Objectives: The study gives priority to the Land Use Land Cover (LULC) changes and changing hotspot locations of Majuli, which is the largest riverine island in Assam. Methods: Satellite imageries for the years 2002, 2011, and 2022 have been utilized to observe spatiotemporal dynamics using the iso cluster unsupervised classification technique. The Iso cluster unsupervised classification method can identify clusters of an image automatically and classify it based on its clusters. The analysis of change has been taken into account by assigning six land use classes, viz. cropland, forest, waterbody, built-up, grassland, and sandbar. Hotspot identification has been done using Getis-OrdGi* statistical methods. This statistical method can distinguish between areas with strong and low spatial associations and detect a tendency for favorable spatial clustering. Findings: The results showed that waterbody, forest, built-up, and grassland have significantly increased by 9.58%, 1.01%, 6.22%, and 9.88% respectively. In contrast, the cropland and sandbars decreased by 25.22% and 1.46% respectively. The kappa accuracy for 2002 is 90.85%, 87.88% for 2011 and 90.22% for 2022. In hotspot analysis, a total of 5 high-range hotspot patches and 11 moderate hotspot patches are identified for the district. The main factors responsible for changing LULC in the study area are the occurrence of floods every year, the combined effects of river erosion, and human interventions. Novelty: Novelty lies in the implications of geospatial techniques in changing LULC in a spatio-temporal context. Thus far, no previous research has been conducted on the identification of LULC hotspots within the study area. This study is significant for managing and conserving land resources and preventing recurring shifts in land cover. It will also offer a fresh perspective on extensively examining the underlying reasons for such frequent changes. Furthermore, identifying the hotspots of changes in the area over twenty years indicates that the district exhibits a volatile landscape. The results will certainly guide stakeholders in the region's sustainable resource management. Keywords: Change detection, LULC, Hotspots, Isocluster, Majuli