The impact of climate change on the storm surges of the Mediterranean Sea: Coastal sea level responses to deep depression atmospheric systems

This study aims to systematically assess the impacts of projected climate change on episodic events of sea level elevation in coastal areas of the Mediterranean, induced by severe weather conditions identified as deep depressions. We try to add new insight in the long-term, climatic timescale, identification of affected parts of the Mediterranean coastal zone correlated to low atmospheric pressure systems, indicative of the Mediterranean basin during the 21st century. To achieve this goal, an integrated quantitative assessment is proposed by combining projections from available and established, green-house gasses emission/concentration scenarios (based on Representative Concentration Pathways; RCP 4.5 and 8.5) with advanced numerical modelling and statistical post-processing for the definition of cyclonic weather impacts on characteristic coastal zone hotspots. To this end, climate projections and outputs from three Regional Climate Models (RCMs) of the Med-CORDEX initiative at the Mediterranean basin scale are used and extensively evaluated against re-analysis data. These atmospheric datasets feed a robust storm surge model (MeCSS) for the simulation of barotropic hydrodynamics (sea level elevation and currents) thoroughly validated against in situ sea level observations by tide-gauges. Our results corroborate a projected storminess attenuation for the end of the 21st century, yet local differentiations in storm surge maxima around the Mediterranean coastal zone are pinpointed. Moreover, a slight reduction of average storm-induced Mean Sea Level (MSL; component attributed solely to the meteorological residual of sea level elevation) is also apparent towards the end of the 21st century. Extreme storm surge magnitudes range between 0.35 and 0.50 m in the Mediterranean with higher values along parts of its northern coasts (Venice lagoon, Gulf of Lions, northern Adriatic and Aegean Seas, etc.) and the Gulf of Gabes in its southern part. Overall, the spatial distributions of surge maxima are estimated to remain similar to those of the past throughout the entire Mediterranean coastal zone. Differentiations between the two scenarios (RCP4.5-8.5) used are obvious, not so much related to the spatiotemporal distribution of storm surge maxima, which shows a very stable pattern, but more in terms of their magnitudes. Indicatively, a decrease of surge maxima from -30% to -2% can be observed towards the end of the 21st century, especially for RCP8.5-driven MeCSS simulations. This is a spatially averaged estimation, yet for some specific coastal sites in Croatia, Spain Italy, and France, such as Rovinj, Bakar, Toulon, Trieste, Ajaccio, Genova, Marseilles, Naples, Venice, Cagliari, Ancona, Ibixa, and Barcelona, the storm surge maxima might increase from 1% to 22% under different RCM/RCP combinations towards the end of the 21st century. Our analysis leads to the quantification of deep depression systems’ effect on the coastal sea level elevation due to storm surges towards 2100. The strongest correlations of deep depression events to high sea levels are observed in several parts along the northern Mediterranean coasts (Gulfs of Valencia and Lions, Ligurian and northern Adriatic Seas). They are followed by mid-latitude areas around Corsica, Sardinia, the mid-zonal Italian Peninsula and the Adriatic, and the northern Aegean Sea. The influence of deep depressions on storm surges is lower for Sicily, South Italy, Peloponnese, Crete, the southern Aegean archipelago, and Alboran Sea. The only exceptions in the generally unaffected southern Mediterranean littorals are the Gulfs of Gabes and Alexandretta. These apply to the 20th century; however, they seem to repeat for the 21st century estimations, with even more pronounced differentiations between the southern and the northern parts. A projected northward shift of the main deep depression centres over the Mediterranean towards the end of the 21st century, is likely the reason for the latter. The climate change signal (difference of Future–Reference Period) of the deep cyclones’ effect on the episodic increases of coastal sea level seems to have a very clear pattern of slight attenuation in certain regions, i.e., Sardinia, Corsica, the Ligurian and Adriatic Seas, and the entire Italian peninsula for all RCM-fed implementations towards the end of the 21st century. Conditionally, this is the case for the Gulf of Valencia, the north-western African coasts, the Alboran, Ionian, Aegean, and Libyan Sea coasts, under specific combinations of RCM/RCP forcings. On the other hand, a possible increase of the Mediterranean deep depressions’ influence on the coastal storm surges might be the case for the Gulf of Lions, the Ionian, Aegean and Levantine Sea basins, covering the north-central and north-eastern coasts of Africa. In general, a positive influence of deep depressions to storm surge maxima would probably refer to areas of mid-to-high storm surge maxima (e.g., Aegean, Ionian, Gulf of Lions or Valencia or Gabes, etc.), but not the highest throughout the basin (e.g., Venice lagoon, Ligurian, Adriatic, etc.). In the latter coastal regions, however, intense local wind forcing mechanisms (i.e., Scirocco) are bound to play an essential role in the formation of high storm surges. The produced results can be used in focused studies for integrated hydrologic/hydrodynamic modelling under projected climate change conditions in the 21st century.