Assessment Of Forest Fragmentation Mapping And Predictive Modelling Using Landsat Data Of Goalpara District, Assam

Abstract

Forest land is primarily comprises with diversities of trees species and have significant forest canopy, which is crucial part of our environment and primary habitat area for many terrestrial ecosystem. But at time of rapid urbanization, industrialization, and globalization the forest land is highly effected on these natural ecosystem. As result the naturally forest cover area breaking down into small patches or disrupted from from core area. To understanding the issues in Goalpara district, where small fragmentation and isolated forest area are highly changing over time and well known habitat area for of Asian elephant the present studies were carried out with Landsat imagery (i.e.1993, 2003, 2013, and 2023) through landscape fragmentation tools (LFT) using GIS environment. The forest fragmentation predictive mapping were also performed to know future estimation (2033) of forest cover change using Artificial neural network (ANN), and Cellular automata (CA) simulation in QGIS software.

The significant outcomes of the study are in last 30 years the district loss forest cover 33.38 sq.km (i.e.1993-2023) with 1.09 sq.km per year and it will accelerate at 9.20 sq.km per year till 2033. Whereas the positive r square value (r² = 0.34) determine that the forest cover area will constantly decline at current trend in upcoming days. While the fragmentation mapping (1993-2033) depicted that out of six classes namely- Patch, Edge, Perforated and Core(Large<500, Medium 250-500, and small <250 acres), most of the transitional changes are found in the class between of edge to patch (25.13sq.km), large core to perforated (13.48 sq.km), medium to small core (4.77 sq. km), patch to edge (33.99 sq.km), perforated to edge (20.26 sq. km.), small core to edge (13.52 sq. km).