Forecasting of Rainfall Using General Circulation Model - Statistical Modelling in Johor

Abstract

Changes in the spatial and temporal rainfall pattern which affected by the climate change need to be investigated as its significant characteristics are often used for managing water resources. This paper implemented the statistical modelling to study the spatial and temporal distribution changes of rainfall pattern in the long term analysis. A daily rainfall series from eight (8) stations in Johor capturing 30 years period (1988-2017) with less than 10% missing data were chosen. To build the predictor-predictand relationship, it depends on the monthly correlation coefficient (R) with the large scale of atmospheric characteristic and performances of calibrated and validated of statistical equation using Statistical Downscaling Model (SDSM). The results showed that the temperature (nceptemp), surface specific humidity (ncepshum) and near surface relative humidity (nceprhum) had strong influence in the local weather formations. As proved, all these stations were successfully produced low standard error (SE) ranged from 3.82% to 11.64%. The annual temporal rainfall was expected to increase to 3058.13 mm (15.4%) at the end of century. The eastern of Johor was expecting to receive higher rainfall intensity and then disperse to the western of Johor. The predictions of rainfall are a bit alarming, therefore, mitigation planning on climate change effects is needed.