Performance evaluation of a Wavewatch III-based wave forecasting system for Mexico.

Episodes of intense wind-generated waves are a phenomenon that affects the development of maritime activities and may even threaten the safety of human settlements near the coast. Under this premise, a wave forecasting system for alerting purposes was developed for the National Center for Disaster Prevention of Mexico (CENAPRED). The wave model used in this system was the WAVEWATCH III, which was implemented for the world ocean with nests in the Gulf of Mexico and the Mexican Pacific. The performance of the model was evaluated for the Gulf of Mexico by comparing the outputs of the model against data from buoys of the US National Data Buoy Center. The evaluation was carried out for the variables of significant wave height, wave period and wave direction. For the first two variables, bias, root mean square error, non-systematic error and Pearson's correlation coefficient were obtained; while for the direction the angular bias and the complex correlation coefficient were obtained. The results showed that the values predicted by the model fit very well to the observations in the first 48 hours of simulation. The monthly averages of the statistical metrics show that the forecast, in general, has a good performance. For the significant wave height variable, the absolute values of bias range from 0 to 0.12 m, the correlation coefficient ranges from 0.63 to 0.89 and the values of the index of agreement range from 0.45 to 0.78. The values for the period and the wave direction are also presented.