5. Conclusion
Knowledge of the significant wave height is essential in the design of coastal structures. Physically collecting such data through long term deployment of wave buoys is an expensive process. To overcome this issue numerical hindcasting and wave transformation models have been developed for predicting the wave climate at a coastal location. However, these numerical approaches also have their limitations. These include: computational expense; sensitivity to input conditions that are often difficult to measure accurately in coastal locations; and neglect of key physical processes. As a result alternative methods for predicting significant wave height have been sought. In this paper, a method for estimating the significant wave height at a coastal location based upon spatial correlations has been put forward. In order to apply the model a limited record of the significant wave heights at the site of interest as well as simultaneously observed wave records at nearby locations are required. This requires less information and input data than that required to run and calibrate a numerical model. Furthermore, since wave buoys are often deployed for short periods at locations where the wave climate is of interest and networks of wave buoys are now commonplace along coasts under threat of flooding, the proposed method is likely to find many applications. It was shown to be able to predict the wave conditions at a particular location to a good approximation giving a similar level of performance to a medium or high resolution SWAN simulation.
