IV-7-Downscaling of Climate Variables using Support Vector Machine.

Abstract

World's climate is showing the changes in a number of components of the hydrological cycle and hydrological systems. Thus it is very important that scientist try to predict the future climate so that we can prepare strategies as part of mitigation and adaptation. Global Climate Models (GCMs) are the best tool to predict future climate but have resolution of hundreds of kilometer. However, many impact applications require the local scale climate variations. Statistical downscaling is one method to feed the large-scale output of GCM simulation into a statistical model to estimate the corresponding local and regional climate characteristics. In this paper, Multi Linear Regression (MLR) and Support Vector Machine (SVM) approaches were applied for statistical downscaling for precipitation and temperature variables in Roorkee area. The results are encouraging.