Abstract

The generation and propagation of waves towards the coastal regions during storm events can substantially increase coastal hazards associated with extreme sea levels. While the Mediterranean Sea is characterized by a fetch-limited environment, the progression of extra-tropical cyclones over its surface often engenders powerful waves. As climate numerical models consistently converge towards a global warming climate over the past few decades, the present wave climate is expected to undergo alterations. However, the reliability of the model projections differ among climate variables, exhibiting for instance higher confidence in the temperature than in precipitation variables.

This study investigates future changes in the wave climate across the Mediterranean region using an extensive ensemble of wave numerical simulations. These simulations were forced with wind fields from thirty-one GCM-RCMs (general circulation - regional climate models) of the European Coordinated Regional Climate Downscaling Experiment (EURO-CORDEX), integrating WaveWatch III and SCHISM numerical models. Future changes in the mean and intense (quantile 0.95) wave climate of significant wave height (Hs), peak wave period (Tp), peak wave direction (Dp) are assessed. Furthermore, we evaluate changes in 100-year return levels of Hs toward the end of the century. Extreme events from each GCM-RCM are aggregated into a single coherent distribution, following a bias correction procedure assuming the Cumulative Distribution Function (CDF) of extreme events to adhere to either a parametric Gumbel or Generalized Extreme Value (GEV) CDF, individually for each model.

Return levels are then computed by fitting a GEV distribution to the unified distribution for both historical and future periods. Our findings reveal an intensification of extreme waves towards the end of the century in several areas of the Mediterranean basin. Despite limitations inherent to bias-correction methods and return level computation, our study underscores the contrasting outcomes between analyzing the entire statistical distribution versus focusing solely on the tail, emphasizing the importance of considering both aspects in wave climate projections.