Item

Abstract

With the increased frequency and intensity of extreme climates in Australia, developing reliable strategies for predicting droughts or sustained high flows has become increasingly important for water supply management. This study proposes a novel approach for forecasting rainfall anomalies over interannual to decadal time scales using the Wavelet System Prediction (WaSP) methodology. By transforming climate variables with the WaSP framework, we demonstrate improved accuracy in modelling the commonly adopted drought indices, such as the Standardized Precipitation Index (SPI). The forecasted drought indices based on the current climate are then combined with those derived from concurrent decadal climate forecasts from the CMIP6 model, demonstrating enhanced accuracy in predicting rainfall anomalies beyond what is currently possible. The hybrid modelling framework is applied to reservoir system in Sydney, NSW, with a focus on water storage level in the system. Our findings demonstrate the potential for WaSP to be used in water resource management applications for improved forecasting of medium to long-term rainfall anomalies. The improved forecasting of rainfall anomalies over interannual to decadal time scales can help water managers alter their plans, shift water from more resilient reservoirs to those that take longer to fill, and reduce the severity of water restrictions, ultimately benefiting downstream communities.