Sahyadri Conservation Series - 53 ENVIS Technical Report: 100,  April 2016

Time-series MODIS NDVI based Vegetation Change Analysis with Land

Surface Temperature and Rainfall in Western Ghats, India


Ramachandra TV           Uttam Kumar          Anindita Dasgupta

Abstract

Land cover (LC) changes are one of the important factors altering the landscape structure affecting ecosystem goods and services. Landscape status assessments through well-established spatial metrics help in monitoring ecosystem trends. Incorporating two-date or sometime lower-frequency-multi-date images based change detection using Landsat or IRS LISS-III/IV data have limited performance for applications in ecologically rich and biologically complex systems. In this context, MODIS based temporal data have wide applicability in providing cost-effective means to develop LC details over large geographic regions. This study explored the use of 250 m multi-temporal (2003–2012) MODIS NDVI 16-day composite data, MODIS LST (land surface temperature) data and monthly rainfall data obtained from Department of Economics and Statistics (Government of Karnataka), Indian Water Portal and NOAA to study vegetation changes and corresponding impact on local temperature along with the rainfall pattern analysis in one of the hottest biodiversity hotspots of the world – the Western Ghats of India.

The study revealed that dense forest area (intact contiguous forests) in the northern, central and southern Western Ghats has considerably decreased by 2.84%, 4.38% and 5.77%, and agricultural/grassland has increased by 2.23%, 4.32% and 5.85%. There is a marginal increase in settlement/open area with the decrease in the spatial extent of water bodies in all the three studied regions. The coefficient of correlation of the different season NDVI of dense vegetation revealed various responses to changing LST and precipitation. NDVI trends were found correlated with the regional climatic variables in 10 years. Monthly NDVI exhibited significant correlations with monthly mean temperature and monthly precipitation during the study period. The trends in rainfall time-series data were analyzed using statistical methods and modelled using autoregressive integrated moving average (ARIMA) which indicated a decreasing trend in the rainfall pattern over forest and agricultural/grassland areas from 2013 towards 2020 in northern, central and southern Western Ghats, revealing the likely grave situation threatening water and food security in peninsular India with the increasing trends of deforestation in the ecologically vital Western Ghats.


Keywords: Vegetation, MODIS, climate change, NDVI, LST, rainfall, modelling, Western Ghats, India