Pengembangan model monsun indonesia berbasis hasil analisis data indeks monsun regional

Abstract

Monsoon is one of regional phenomenon that affects weather and climate in Indonesia. This is caused by the territory of Indonesia which is located in equatorial zone and between Asia in the north and Australia in the south. Due to the location, there are two monsoons named Asia which causes wet season (December-January-February) and Australia which causes dry season (June-July-August). Until this time, Indonesia doesn’t have monsoon index which is needed by BMKG (Meteorology, Climatology and Geophysics Agency). Regional monsoon index (RMI) data must be calibrated with data from Indonesia, EAR (Equatorial Atmosphere Radar) to examine the relation of EAR with IMR data. Selected time-series of IMR data is used to develop monsoon model. The method that used is monsoon analysis and spectral analysis with Matlab and cross correlation analysis and the Box-Jenkins method, ARIMA with SPSS 16.0. The results of the analysis show that AUSMI (Australian Monsoon Index) at 200 mb layer which oscillate around 12 months has high correlation with EAR data 0.781. The appropriate model to show the monsoon model is ARIMA (1,1,1). AUSMI period t is decided by one and two months latest data (Zt-1) and (Zt-2) and a month error latest data (at-1) with equation Zt = 1.047 Zt-1-.047Zt-2 + 0.093at-1. The validation results obtained using model with actual data IMR February 2004-December 2008 give average error 2.3906 and correlation 0.7139.