Modelling of Multi Input Transfer Function for Rainfall Forecasting in Batu City

Abstract

The aim of this research is to model and forecast the rainfall in Batu City using multi input transfer function model based on air temperature, humidity, wind speed and cloud. Transfer function model is a multivariate time series model which consists of an output series (Y_t) sequence expected to be effected by an input series (X_t) and other inputs in a group called a noise series (N_t). Multi input transfer function model obtained is (b₁,s₁,r₁) (b₂,s₂,r₂) (b₃,s₃,r₃) (b₄,s₄,r₄)(p_n,q_n) = (0,0,0) (23,0,0) (1,2,0) (0,0,0) ([5,8],2) and shows that air temperature on t-day affects rainfall on t-day, rainfall on t-day is influenced by air humidity in the previous 23 days, rainfall on t-day is affected by wind speed in the previous day , and rainfall on day t is affected by clouds on day t. The results of rainfall forecasting in Batu City with multi input transfer function model can be said to be accurate, because it produces relatively small RMSE value. The value of RMSE data forecasting training is 7.7921 while forecasting data testing is 4.2184. Multi-input transfer function model is suitable for rainfall in Batu City.