Application of Multiple Linear Regression as Downscaling Methodology for Lower Godavari Basin

Abstract

This paper focused on future precipitation scenarios adopting statistical downscaling approach, namely, Multiple linear regression (MLR) for Lower Godavari basin, India. Global Climate Model (GCM), namely, GFDL-CM3 simulations, are used for downscaling purpose. Five grid points of Lower Godavari basin are considered. Reanalysis data from National Centre for Environmental Prediction (NCEP) of the study area from 1969 to 2005 is used for analysis. Precipitation is chosen as predictand. Representative Concentration Pathways (RCPs) scenarios, 4.5 and 6.0 are used for the study. Projected precipitation from 2006 to 2100 is obtained by the developed MLR model. Downscaled precipitation predictions show that there is increase in precipitation in the future.