Mediterranean Sea level forced by atmospheric pressure and wind: Variability of the present climate and future projections for several period bands

This paper documents the present climate and future climate projections of sea-level variability (SLV) induced by atmospheric pressure and wind for period bands ranging from 1 to 100days in the Mediterranean Sea. The results are based on output from a regional barotropic numerical ocean model. Present climate SLV was simulated using air pressure and wind fields from a dynamical downscaling of the ERA40 reanalysis for 1958–2008 (hindcast) and validated with long-term tide gauge records. Climate projections for the 21st century were forced with output from the atmospheric ARPEGE model using scenarios B1, A1B and A2. Additionally, we performed a control simulation forced by observed SSTs, GHGs and aerosol concentrations for the period 1950–2000. SLV has been assessed over three different period bands (1–3, 3–10 and 10–100days) that were obtained by applying digital band-pass filtering procedures to both the present day and climate projections. Of the three time periods, the 1–3-day SLV had the worst comparability between the hindcast and the observations. We found that most of the SLV trends were insignificant in the present climate but were negative in the projections over all of the periods, especially for the medium- (A1B) and high-emission (A2) scenarios. Projections of 3–10- and 10–100-day extreme SLV, estimated from high-order percentiles, indicate increases in extreme SLV events versus the average. This result does not apply to the 1–3-day extreme SLV. The largest negative SLV trends were projected for autumn and for the 1–3-day period band, except for the low-emission (B1) scenario. However, the 1–3-day period band contributed less than 12% to the overall SLV, while the largest variance (55–75%) was present in the 10–100-day period band, indicating the importance of correct simulation of planetary-scale atmospheric processes with long temporal scales for the reliability of the climate projections.