On Prediction of Subang, Selangor Daily Rainfall Data: An Application of Local Approximation Method

Abstract

Prediction has been done on the daily rainfall data taken from Subang Meteorology Station, Selangor, located in Malaysia. Firstly, the original daily rainfall data need to be systemized through the reconstruction of phase space approach (RPS). Then, the system was utilized to predict rainfall via local approximation method (LAM). Out of seven days predictions, the results are found to be in good agreement with the observed ones for five days. Even though the results seem well, the values of overall prediction errors are quite high. So, some analysis have been done, a few remarks are highlighted and a number of suggestions are proposed in order to produce better prediction.