Accuracy of wave forecasts as a function of forecast time horizon in south-western Australia

In a quest to find alternatives to traditional fossil fuel powered energy systems, there is a global thrust to develop renewable energy sources such as those from wind, wave, tides and solar energy. Waves offer the most consistent and reliably predictable energy source and have recently become of great interest as many pilot wave power plants have been installed in Australia. A reliable forecast system is essential for these power plants to be successful, which is why the global forecasting system WAVE WATCH III (WW3), has been analyzed in its accuracy of forecasting significant wave heights over a forecast horizon of 120 hours in this paper. To do so, forecast data for Rottnest Island, WA Australia, was used, leading to the conclusion, that WAVE WATCH III provides a forecast of significant wave heights with a standard deviation of <10% for the first 72 hours of forecast, after which it nearly linearly increases to 16% deviation at 120 hours of forecast for Rottnest Island. This indicates a shortcoming of WW3 in forecasting significant wave heights over an extended period in the study region. However, it must be considered that the area examined is on the continental shelf, for which WW3 takes no account as it respects merely wind force as the wave driving force. WW3 was found to predict peak waves during calmer seas with the least accuracy, demonstrating a tendency to underestimate the significant wave height. Periods of constant higher significant wave heights were well predicted, contradicting the positive relationship between significant wave height and standard error and again indicating the shortcoming of WW3 when applied for regions along the continental shelf. It is suggested to conduct further studies at various locations along the coast to enforce this statement as well as gain more data in order to correct for any inaccurate forecasts, possibly by integrating near shore forecasting models such as SWAN.