Storm Characterization Using a BME Approach

Abstract

The storm that occurred at the Spanish coast of the Mediterranean Sea at the end of January of 2017 produced severe coastal floods. The space–time evolution of seawave heights during that event is analyzed in space and time using a combination of the spatiotemporal random field (S/TRF) theory and the Bayesian maximum entropy (BME) method. Observed were combined with hindcasted datasets from Puertos del Estado (Spain) to assess modeling accuracy and improve the analysis. The mean absolute error and root mean square error of the tenfold cross-validation technique were found to be equal to 7.90 $\cdot$ 10$^{-2}$ m and 9.59 $\cdot$ 10$^{-2}\mathrm{m}$, respectively. The results are presented in the form of spatial maps of seawave height statistics (mean and variance) at the study domain. The mean wave height during the storm propagation is fairly well reproduced. The variance shows two regions of permanent maximum variance at the Mazarron Bay and between the Azahar coast and the north face of the Balearic islands. Some indicators were computed based on S/TRF of the mean wave height maps. The storm shape and a suitable storm determination threshold for the definition of the storm can be inferred from the results. The classification of several storms based on this methodology improves the assessment of the potential damage caused by the storm event, thus enabling the development of management strategies in coastal areas.