Geospatial Analysis of Land Use Land Cover Change Modeling at Phewa Lake Watershed of Nepal by using Cellular Automata Markov Model

Abstract

Improper practices of land use/ land cover (LULC) are deteriorating watershed conditions. Remote sensing and GIS tools were used to study LULC dynamics using Cellular Automata (CA)–Markov model and predict the future LULC scenario for years 2015 and 2020, in terms of magnitude and direction, based on past trend in Phewa Lake watershed, Kaski district, Nepal. The analysis of LULC pattern during 1995, 2000, 2005 and 2010 using satellite-derived maps has shown that the biophysical and socio-economic drivers including slope, road network and settlements proximity have influenced the spatial pattern of the watershed LULC. These lead to an accretive linear growth of Medium to Fairly Dense Forest, Open Forest, Bush/Scrub, Waste Land and Built-up Land but decrease in other LULC classes. Annual rates of increase from 1995 to 2010 in Medium to Fairly Dense Forest, Open Forest, Bush/Scrub, Waste Land and Built-up land were 9.16, 8.14, 20.66, 15.27 and 27.77 ha/year respectively, while the rates decrease in Dense Forest, Terrace Agriculture, Valley Agriculture and Grass Land were 39.17, 10.30, 23.32 and 3.78 ha/year respectively. Subwatershed-wise LULC change showed decrease by 130.5 ha and 65.4 ha of Dense Forest, and increase in Medium to Fairly Dense Forest by 7.2 ha, 68.2 ha and Open Forest by 75.0 ha and 0.0 ha in mid and North flowing Subwatershed respectively from 1995 to 2010. Medium to Fairly Dense Forest is predicted to increase by 51.3 ha and 113.4 ha and Dense Forest is predicted to decrease by 44.9 ha and 136.5 ha from 2010 to projected 2015 and 2020 in Harpan Subwatershed. The predicted LULC scenario for 2015 and 2020, with reasonably good accuracy would provide useful inputs to the LULC planners for effective management of the watershed.