The Development of Model Estimation for Monthly Precipitation Using TRMM Satellite Data on Three Rainfall Regions within Indonesia

Abstract

Rainfall is one of the atmosphere parameters which is quite difficult to predict because of its very high variability on space and time. The recent development of rainfall observation technology nowadays is using rainfall satellite. This satellite has near real time rainfall monitoring, wide coverage and fast access. In this study, climatological analysis of rainfall pattern, statistical analysis to evaluate and validate TRMM data using selected ground stations and TRMM data, development of correction factor for TRMM data and monthly rainfall prediction using ARIMA model, have been done in three regions with different rainfall pattern. Those areas are monsoonal pattern in Lampung, East Java and Kalimantan Selatan; equatorial pattern in North Sumatera and West Kalimantan; and local pattern in Maluku. The validation analysis showed very high correlation with the gauge data for monsoonal pattern area (>0.80), high correlation for equatorial (>0.60) and local pattern area (>0.75). Each pattern area has a correction equation. In general, the error of corrected TRMM data decreased for monsoonal, equatorial and local pattern. The model of monthly rainfall prediction has been developed for corrected TRMM data with 95% level significance. The verification of model showed high coefficient correlation (r) in local pattern and monsoonal pattern, but lower on equatorial pattern.