The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Publications Copernicus
Articles | Volume XL-8
28 Nov 2014
 | 28 Nov 2014

Study of temporal variation of vegetation indices and phenology of tropical deciduous broadleaf forest in eastern India

U. Gupta

Keywords: Vegetation, Forestry, Identification, Monitoring, Satellite, Multitemporal

Abstract. In the recent years study of vegetation variation has taken significance as a quintessential part of a bigger research agenda of climate change. This study has attempted to (a) Analyse the inter-annual and seasonal variation of vegetation (b) Observe the variation in phenological transition dates of tropical deciduous broadleaf forest (DBF) in Eastern India from 2003 to 2012. The study was conducted on 1500 sq. km area of dense forest in the districts of West Singhbhum and Sundargarh. MODIS EVI data of 250 meter resolution was used. Land cover mask was used to study only the DBF pixels. Least squares convolution method proposed by Savitzky and Golay was used for noise reduction and smoothing. For the determination of phenological parameters iterative Savitzky-Golay method followed by threshold method was used in TIMESAT software. The results showed that (i) EVI varied between 0.41 in early April and 0.71 in mid October (ii) The overall trend was decreasing with a slope of 0.0022, representing a degradation trend (iii) EVI of summer season was found to be more stable than of rainy and winter season (iv) Summer and rainy months showed a decreasing trend whereas the winter months showed an overall increase (v) Day of start of season varied between 2nd May and 20th June whereas the day of end of season varied between 1st February and 7th March (vi) Length of season was longest for 2007 at 302 days and shortest for 2010 at 235 days and showed an overall decreasing trend.