Development of Seasonal Rainfall Forecasting Models for Western Australia using Large Scale Climate Drivers

This study explored the seasonal variability of rainfall in Western Australia, considering the effects of dominant climate drivers. The climate data were analyzed to develop prediction models using potential climate drivers and their associated indices. The analyzing techniques included Multiple Linear Regression (MLR), Auto-Regressive Integrated Moving Average with Exogenous input (ARIMAX), Gene Expression Programming (GEP), and a hybrid system. Among the developed models, GEP-ARIMAX hybrid models demonstrated outstanding prediction performance with high statistical significance. Using the developed models, rainfall can be predicted up to four months in advance which can be of great strategic and economic benefit for Australia.