The Modeling of Downscaling GCM output and SST Anomaly Nino 3.4 Using Support Vector Regression (A Case Study of The monthly Rainfall In Indramayu).
Abstract
The objective of this research is to develop a downscaling model GCM output and SST anomaly Nino 3.4 as input in the training to predict a rainfall monthly in Indramayu. The techniques of a downscaling is used for a phenomenon indicators of El Nino and Southern Oscillation (ENSO) climate anomaly such as a Global Circulation Model (GCM) and Sea Surface Temperature (SST) nino 3.4 are commonly used as a primary study learn and understand the climate system. This research propose a method for developing a downscaling model GCM output and SST anomaly Nino 3.4 by using Support Vector Regression (SVR). The research showed that GCM output and SST anomaly Nino 3.4 can be approach the average value of monthly rainfall. The best result of prediction is Bondan station which has average correlation that is 0.700.