DOWNSCALING MONTHLY PRECIPITATION USING LINEAR AND NON-LINEAR REGRESSION MODELS

Numerous regression techniques have been proposed to downscaling rainfall, temperature and other climatic parameters. Performance of downscaling model depends on scale and characteristics of climatic variable of a region. This emphasizes the need to identify the suitable model for downscaling climate at a region. The objective of the present study is to assess the performance of linear and non-linear parametric regression models in downscaling monthly precipitation in the East coast of Peninsular Malaysia. For the this purpose, three downscaling models based on linear regression model (LM), generalized linear model (GLM), generalized additive model (GAM) have been developed and evaluated in this study. Different statistical measures have been used to assess the performance of downscaling models. The obtained results show that LM based downscaling model performs well in downscaling monthly precipitation in the study area. It is remarked that due to near normal distribution of monthly rainfall in a tropical region makes the LM based downscaling model much strong compared to other models developed considered different distribution of data.

Statistical downscaling, multiple linear regression, generalized linear model, generalized additive model.