Water is a vital resource whose management in this era of climate change is greatly dependant on better predictive tools that can assist in timely decision making. Rivers and streams play a critical role in the hydrologic cycle and accurate short term forecasting of streamflow, especially for periods of adverse rainfall, will allow better management of river flows and reservoir system operations. Specifically, this would assist in better preparedness and planning of water allocations, hydropower operations, flow harvesting and distribution, flood forecasting and timely environmental watering regimes.

There are several hydrological parameters that effect stream flows and a major challenge with any prediction methodology is to understand these parameter interdependencies, correlations and their individual effects. A robust methodology is, thus, required for accurate prediction of streamflow under usually unique, waterway-specific conditions using available data.

This research presents the application of dynamic Artificial Neural Network (ANN) to provide short-term streamflow forecasting using nonlinear autoregressive with exogenous input (NARX). The past values of streamflow, together with other input parameters, are used to train, validate and optimise neural networks to improve prediction of streamflow forecasting. The optimisation steps of the methodology are discussed and the predicted outputs are compared and analysed with respect to the actual streamflow values.