Improving streamflow forecasting lead-time for Australian rivers using oceanic-atmospheric oscillations and hydroclimatic variables

Abstract

It is essential to develop a long-lead streamflow forecast system for providing the prior signal for possible floods. Climatic variabilities such as oceanic-atmospheric global oscillations may possess tele-connectivity with Australian rainfall-runoff. This study identifies an ocean-atmospheric region connected with Australian rivers streamflow. By utilizing its persistence capacity, statistical and machine learning-based forecast models are developed, predicting inter-annual streamflow forecast of Australian river flows. This outcome will be beneficial for future water planning and mitigating flood risk.