Statistical Downscaling of Sea Level by Support Vector Machine and Regression Tree Approaches

Abstract

Projections of future climate from various climate models indicate that global temperatures are continuously rising. This, in turn, may result in a significant rise in the water levels in the oceans, adversely impacting coastal aquifers. Apart from temperature, some other climatic variables also influence sea level. For better management of coastal aquifers, it is necessary to predict the sea level with a reasonable degree of accuracy. The repercussions of projected climate change on sea level rise can be investigated by projecting future sea level values for different representative concentration pathways (RCPs) using global climate models (GCMs). GCMs are run at a coarser scale; hence for regional-scale studies these projections have to be downscaled before being input into hydrologic models. This paper presents the details and results of a study in which support vector regression (SVR) and regression tree (RT) techniques were applied for statistical downscaling of sea level using climatic variables. The results of both these techniques were compared. It was observed that the performance of the SVR model was better than that of the RT technique.