Long Term Monitoring of Forest Fires in Silent Valley National Park, Western Ghats, India Using Remote Sensing Data

Abstract

The main objective of this study is to integrate sat- ellite derived forest burnt areas over a multi-decadal scale to understand fire frequencies in Silent Valley National Park, Kerala, India. There are nine vegetation types distinguished i.e. wet evergreen forest, shola, semi evergreen forest, moist deciduous forest, dry deciduous forest, savannah, reed brakes, scrub and grasslands using Resoursesat-2 LISS III data. The spatial extent of burnt area found to be highly varying and has been estimated as 7.95 km 2 in 1973, 17.19 km 2 in 1985, 5.51 km 2 in 1989, 9.55 km 2 in 1990, 4.55 km 2 in 1994, 6.89 km 2 in 1996 and 5.80 km 2 in 1999. Analysis of burnt area coverage for recent decade based on 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013 and 2014 reveals fire affected area as 8.10, 4.20, 12.21, 5.52, 3.19, 0.62, 1.81, 2.03, 3.00 and 1.36 km 2 area respectively. Spatial analysis indicates, an area of 47.51 km 2 (19.5 % of total vegetation cover) affected by varying frequencies of fires from 1973 to 2014. High frequencies of fires have been observed in dry deciduous forests and grasslands. Buffer zone of national park hasundergonefrequentfires which may degrade climaxforest vegetation and affects the regeneration and adaptability of indigenous species. Spatial tracking reveals 58.1 % grid cells were affectedby forest fires.Overall, there is a decliningtrend of forest fires indicates management effectiveness in Silent Valley National Park.