ABSTRACT
General Circulation Models (GCMs) are numerical models representing the general circulation of atmosphere and ocean, as they have gained importance in hydro environment fields, and render one to understand how climate changes on a regional scale, and are of great importance in planning for our habitat. In many hydrologic studies, the hydrologic variables of interest are usually generated from the predicted atmospheric variables, wherein the GCM predictor variables at coarser grid resolution are downscaled to finer resolution at the basin level, either by statistical or dynamical downscaling. Statistical downscaling is the commonly used downscaling method mainly due to its computational efficiency. Even though studies utilizing GCM outputs for watershed level studies for certain basins in India are available, only few studies are reported so far in the peninsular region of India. The downscaled variable values also vary between GCMs and between regions. The study involves the statistical downscaling of precipitation for the Bharathapuzha basin of Kerala state using NCEP/NCAR reanalysis data and projecting the precipitation for a future period using CGCM3.1 of T63 resolution for B2 scenario. Multivariate regression using Principal Component Analysis (PCA) was used for downscaling. The zone of influence of predictors was determined prior to downscaling to get better performance. The results indicate that this method can be reliably used for downscaling the precipitation for the study region.
