A stepwise Principal Component Regression Model to predict Seasonal Rainfall over Idukki district of Kerala

Abstract

The study comprises of developing a mathematical model based on principal component regression and step-wise multiple linear regression over Idukki district of Kerala. Different parameters such as sea surface temperature, ocean heat content and wind are utilized for the development of model. 63 different models are constructed for this purpose. The observed rainfall data for the district has been collected from the Kerala Government. Results suggest that the rainfall in the region has become asymmetric with more rainfall in theAdarsh-Suvarna J and Archana Nair month of September. The efficiency of model is judged by root mean square error (RMSE) and it has been found that the model by taking only OHC in the region (0-50N,500E-750E) as parameter gives, the least RMSE.