Modelling Present and Future Climate in the Mediterranean Sea: A Focus on Sea-Level Change

Modelling Present and Future Climate in the Mediterranean Sea: A Focus on Sea-Level Change | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Modelling Present and Future Climate in the Mediterranean Sea: A Focus on Sea-Level Change Gianmaria Sannino, Adriana Carillo, Roberto Iacono, Ernesto Napolitano, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-653703/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 21 Jan, 2022 Read the published version in Climate Dynamics → Version 1 posted 5 You are reading this latest preprint version Abstract We present results of three simulations of the Mediterranean Sea climate: a hindcast, a historical run, and a RCP8.5 scenario simulation reaching the year 2100. The simulations are performed with MED16, a new, tide-including implementation of the MITgcm model, which covers the Mediterranean - Black Sea system with a resolution of 1/16°, further increased at the Gibraltar and Turkish Straits. Validation of the hindcast simulation against observations and numerical reanalyses has given excellent results, proving that the model is also capable of reproducing near-shore sea level variations. Moreover, the spatial structure of the elevation field compares well with altimetric observations, especially in the Western basin, due to the use of improved sea level information at the Atlantic lateral boundary and to the adequate treatment of the complex, hydraulically driven dynamics across the Gibraltar Strait. Under the RCP8.5 future scenario, the temperature is projected to generally increase while the surface salinity decreases in the portion of the Mediterranean affected by the penetration of the Atlantic stream, and increases elsewhere. The warming of sea waters results in the partial inhibition of deep-water formation. The scenario simulation allows for a detailed characterization of the regional patterns of future sea level, arising from ocean dynamics, and indicates a relative sinking of the Mediterranean with respect to the Atlantic more pronounced than the current one. Explicit tidal forcing and an accurate resolution of the Gibraltar Strait are proved to be key features in the designing of numerical simulations for the Mediterranean Sea. Climatology Climate Analysis and Modeling Mediterranean Sea climate MED16 altimetric observations RCP8.5 Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 Figure 14 Figure 15 Figure 16 Figure 17 Figure 18 Figure 19 Figure 20 Figure 21 1. Introduction Future sea level rise (SLR) constitutes a threat for the coastal environment and economies, which is liable to be further exacerbated by the superposition of waves, atmospheric surge, and tides. Climate scientists are therefore becoming more and more committed to improve projections of SLR as to both resolution and accuracy, and to reduce current uncertainties, at the same time increasing our understanding of such a challenging scientific problem and enabling more accurate risk assessments. The latter, in particular, suffer from the difficulty in accounting for the cascading effects from sea-level rise to actual coastal impacts, across a range of spatial and temporal scales that demand complex and differentiated approaches (Thiéblemont et al., 2019 ). Likewise, the lack of comprehensive projections of extreme sea levels (ESL) that include mean sea level (MSL), tides, waves, and storm surges is lamented by Vousdoukas et al. ( 2017 ), who present a first attempt to provide an impact-oriented, regional-scale evaluation of ESL for the European shoreline. Their effort, albeit still inadequate for the design of specific measures, can nevertheless help highlight vulnerabilities and emerging research issues. As a matter of fact, the authors denounce the limitations inherent to their approach, mainly arising from the inadequacy of the regional atmospheric projections available at the time, as well as those of the projections derived from coupled global OGCMs. In addition, they highlight that explicitly solving the nonlinear interactions between waves, storm surge, and tides should be a key element of any future effort to reliably model ESL, rather than continuing to resort to linear combinations of the ESL components resulting from independent simulations. Besides being highly vulnerable to SLR due to the potentially disruptive impacts on its coastal economies (Jeftic et al., 1992 ), the Mediterranean basin is especially challenging from the scientific perspective, due to the inherent diversity of its geological history and the peculiar and complex features of its marine circulation and environment. In addition to the local dynamics and geological processes, interacting over a broad spectrum of scales, sea level in the basin is also constrained by the water mass exchange across the Strait of Gibraltar, which, in fact, regulates the hydraulic jump between the Mediterranean and the Atlantic Ocean and influences how sea level rise in the Atlantic is transmitted into the Mediterranean. In its turn, the connection with the Black Sea, through the Dardanelles, Sea of Marmara and the Bosphorus, couples the basin hydrology to the land-based hydrological cycle of a vast portion of continental Europe. Nevertheless, despite the recognized inability of coarse resolution GCMs to accurately represent the highly non-linear, small-scale processes in marginal seas such as the Mediterranean (Slangen et al., 2017a ; Marcos and Tsimplis, 2008 ), the global sterodynamic (following the terminology proposed by Gregory et al., 2019 ) sea-level projections for the Atlantic area near Gibraltar, are still often used to estimate the basin's internal sea level (Thiéblemont et al., 2019 ). At the global scale, the question is still open whether the strong trends observed in the last 25 years through satellite altimetry are part of longer-term tendencies, or reflect more recent changes in the atmosphere-ocean coupled system. Recent work by Dangendorf et al. ( 2019 ) provides evidence that the global trend accelerated at the end of the 1960s, and indicates that the resulting acceleration in the global sea level rise is linked to modifications of the southern hemispheric westerlies, leading to warming and circulation changes in the southern world ocean. Other factors have also been shown to play a key role, such as the change of terrestrial water storage, and the melting of ice sheets and glaciers, which has significantly increased in the last decades (see, e.g., Shugar et al., 2020 , which focuses on the growth of glacial lakes, and King et al., 2020 , where the mass loss from the Greenland Ice Sheet is analyzed). The work by Frederikse et al. ( 2020 ) provides a thorough review of the various contributions and their relative weight. As to future scenarios, since 1995 the coordinated worldwide CMIP effort (Coupled Model Intercomparison Project) has provided constantly updated projections of future sea level rise using earth-system models, to which the contributions from continental glaciers and geologic processes are added offline. The Phase 5 models (CMIP5) used for the fifth Intergovernmental Panel on Climate Change (IPCC) assessment report (AR5, IPCC 2013 ) project a global sea level rise between 26 and 97 cm (overall likely range) at the end of this century, depending on the climate scenario, with pretty large inter-model spread. Such estimates were revised in the 2019 Special Report on the Ocean and Cryosphere in a Changing Climate (SROCC), which assigns medium confidence to a likely range between 29 cm and 110 cm (IPCC, 2019 ). For the area of the North Atlantic in proximity of the Strait of Gibraltar, projections can range from about 40 cm to over 100 cm for the most pessimistic RCP8.5 scenario, with an ensemble mean of about 80 cm, and approximately from 30 cm to 80 cm for the intermediate RCP4.5 scenario, with an ensemble mean of 60 cm (Vousdoukas et al., 2017 ). On the other hand, Slangen et al., 2014 presented regional sea-level projections and their associated uncertainties up to the end of the 21st century, by combining model- and observation-based regional contributions of land ice, groundwater depletion and glacial isostatic adjustment, and CMIP5 projections of the changing ocean circulation, increased heat uptake and atmospheric pressure patterns, also accounting for gravitational effects due to mass redistribution, ending up with an estimate of about 70 cm for the Mediterranean Basin. However, while concluding that regional variations in sea-level can significantly differ from the global mean (up to 30% above and 50% below) and that the land ice contribution dominates the overall uncertainty, they also stressed the need for dedicated regional downscaling of each global climate scenario. In general, the inherent uncertainty in SLR projections has induced researchers to adopt a variety of alternative estimates in their impact assessments. Thiéblemont et al. ( 2019 ) analyze the effects of a median estimate of about 80 cm and two different high-end scenarios, resulting from two extreme estimates for the sea-level equivalent of melting glaciers (i.e the the upper limit of the likely range and the “worst model” projections). Antonioli et al. ( 2017 ), while reviewing possible alternative ranges for the projected high-end SLR at 2100, use the 530–970 mm interval reported in the IPCC AR5, and Rahmstorf’s semi-empirical estimate of about 1.400 mm (Rahmstorf, 2007 ). Improved projections from the next generation of global models are expected by 2022, when the next IPCC assessment is to be delivered. However, as remarked above, the spatial resolution of current global models is not sufficient to provide realistic estimates of local sea level rise in areas, such as the Mediterranean, where crucial processes are far from being explicitly resolved, and hardly allow a reliable parameterization (Sannino et al., 2009 ). Adloff et al. ( 2018 ) recently discussed the complexity inherent to projecting sea level in the Mediterranean, and analyzed the performances of four different regional hindcast simulations of the basin circulation (period 1980–2012), driven by realistic atmospheric forcing. The incorporation of improved sea level information at the Atlantic lateral boundary was found to significantly enhance the reliability of results, confirming that the correct representation of the interactions between the two basins is an important requirement for a successful numerical simulation. However, an adequate treatment of the complex, hydraulically driven dynamics across the Strait of Gibraltar was still missing, as well as a more refined treatment of the exchange between the Mediterranean Sea and the Black Sea through the Dardanelles Strait, at the model eastern boundary. In addition, model performances were mostly evaluated by computing basin averages, while altimeter data series clearly indicate that sea level rise has not been homogeneous in the basin over the last decades. By analyzing altimeter data over a 25 years period (1993–2017), Mohamed et al. ( 2019 ) showed that the observed increase in the average sea level anomaly (SLA) has been quite different in different sub-basins, ranging from a minimum of 1.95 mm/year, in the Ionian Sea, to a maximum of 3.73 mm/year, in the Aegean Sea. At a broader scale, the SLA positive trend appears to be significantly stronger in the Levantine basin than in the Western Mediterranean Sea, as well as characterized by distinctive features, with a fairly regular linear increase in the Western Mediterranean Sea, possibly attributable to the Atlantic contribution, and a more complex behavior in the Levantine basin. The latter, in particular, is liable to be influenced by the evolution of the Eastern Mediterranean Transient (EMT), the dramatic event occurred in the eastern Mediterranean at the beginning of the 1990s (Roether et al., 1996 ; Klein et al., 1999 , Theocharis et al., 2002 ). The difficulty in separating climate-change-induced variations from the local circulation variability is therefore evident, as well as the role played by small-scale features, either attributable to internal variability or determined by atmospheric forcing, in generating long-lasting differences across the Mediterranean sub-basins, amplifying or mitigating the effects of global sea level rise. Prompted by these considerations, we developed a regional ocean model for the long-term simulation of the Mediterranean Sea circulation (hereinafter MED16) which we used to obtain accurate projections of the Mediterranean sea level. The model represents the climate version of the high-resolution, tide-including ocean model described in Palma et al. ( 2020 ). The two models share the same computational domain, which includes the Black Sea, thus allowing to consistently compute water exchanges at the Dardanelles Strait and to avoid any ad hoc assumption at the Mediterranean eastern boundary. The climate version necessarily uses a coarser horizontal grid with respect to its operational counterpart, yet resolution is significantly increased in critical regions such as the Gibraltar, Dardanelles, and Bosphorus Straits. The explicit, high-resolution representation of inter-basin water exchanges at the boundaries constitutes a unique, distinguishing feature of the present implementation. In the following, we analyze the hindcast of the Mediterranean circulation forced by the SMHI-RCA4 regional downscaling of the ECMWF ERA-Interim reanalysis data (Dee et al. 2011 ), and the historical and RCP8.5 scenario forced by the SMHI-RCA4 regional downscaling of the HadGEM2-ES global projection (Collins et al., 2011 ). The overall basin-scale model performance is assessed through comparison with observations and reanalysis data, with special focus on the model ability to reliably represent the local sea level height. In particular, comparison with data from tide gauges allows us to evaluate the model performance in coastal regions, where the impacts of sea level rise really affect human communities and economies, and need to be specifically assessed. The paper is organized as follows. The main characteristics of the model and of its implementation are described in Sect. 2, whereas Sect. 3 – 5 are devoted to the detailed analysis of the simulations performed. In particular, Sect. 3 presents a validation of the hindcast and historical simulations in terms of transports at the main straits, hydrology, and circulation, obtained through comparison with observations and numerical reanalyses. The corresponding sea-level reconstructions are discussed in Sect. 4 , and validated using satellite altimeter data and coastal tidal gauge observations. Section 5 then describes the results of the scenario simulation, highlighting future changes in the basin hydrology and circulation, and discussing the projected sea-level. Finally, conclusions are drawn in Sect. 6 . 2. Data And Methods In order to produce the downscaled Mediterranean sea-level field under present climate (1980–2005) and the RCP8.5 future scenario (2006–2100), the MED16 model was forced using the correspondent regional downscaling experiments performed with the SMHI-RCA4 atmospheric model (Strandberg et al., 2014 ), alternatively constrained by a) the ERA-Interim reanalysis data, for the hindcast simulation, b) the HadGEM2-ES global model (Collins et al., 2011 ) for the historical simulation, and c) the HadGEM2-ES RCP8.5 for the future scenario. The current section describes in detail the atmospheric data used (Sect. 2.1 ) and the MED16 model characteristics and specific setup (Sect. 2.2 ). 2.1 Atmospheric forcing The atmospheric forcing is derived from the dynamically downscaled regional atmospheric fields produced by the Rossby Centre regional atmospheric model RCA4 (Strandberg et al., 2014 ), at 0.11° resolution (i.e. approx. 12.5 km grid spacing), over the EURO-CORDEX domain (Giorgi et al., 2009 ), whose boundary conditions are either provided by the ERA-Interim reanalysis, or by the atmospheric component of the CMIP5 global model HadGEM2-ES, which is commonly used for downscaled sea-level projections (Hermans et al., 2020 ). The hindcast experiment covers the period 1981–2010, the historical experiment covers the period 1981–2005, while the RCP-8.5 scenario spans years from 2006 to 2100. Up to 2005, observed greenhouse gas concentrations have been prescribed, which are then substituted by those indicated by Meinshausen et al. ( 2011 ) for the future business-as-usual scenario RCP8.5. To reproduce the surface heat fluxes, shortwave radiation from the atmospheric models is imposed, whereas the wind stress and the other heat flux components are computed via bulk formulas considering the sea surface temperature, the winds at 10 m height, the dry air temperature and the air pressure at 2 m, and the relative humidity as inputs. In particular the long-wave radiation is computed according to the formula proposed by Bignami et al. ( 1995 ), while the latent heat flux, the sensible heat flux and wind stress are computed according to the Large and Yeager ( 2004 ) bulk formula. Cloud cover is taken from the atmospheric model. The net freshwater flux is computed as precipitation (taken from the atmospheric model) minus evaporation (computed from the latent heat). To reduce salinity drift in the hindcast simulation the sea surface salinity (SSS) is restored toward the MEDHYMAP monthly climatology (Jordà et al., 2017a ) at a timescale of two days over the topmost model layer of 2 m thickness. At the surface, the model is forced by 6-hourly wind, and 3-hourly surface pressure, heat and freshwater fluxes computed via bulk formulae. The fresh water flux is prescribed in conjunction with the non-linear free surface numerical scheme. 2.2 The regional ocean model MED16 The numerical ocean model MED16 used to simulate the downscaled sea level is the updated version of the Mediterranean ocean model developed by Sannino et al. ( 2015 ). It is based on the MITgcm kernel developed by Marshall et al. ( 1997a , b ), and solves the Navier-Stokes equations under the Boussinesq approximation for an incompressible fluid. In the present study, the hydrostatic version of the model has been implemented, using a finite-volume spatial discretization on a staggered Arakawa-C grid, partial step topography, rescaled vertical height (z*) coordinate (Adcroft and Campin 2004 ), and an implicit non-linear free surface formulation (Campin et al. 2004 ). The source code and documentation are available at the following web site: https://github.com/MITgcm/MITgcm (last access: 20 April 2021). Model bathymetry for both the Mediterranean and the Black Sea was derived from the European Marine Observation and Data Network (EMODnet) 2016 dataset ( https://www . emodnet-bathymetry.eu), while the high-resolution digitalized chart of Sanz et al. ( 1991 ) provided data for the SoG, and the high-resolution bathymetry for the Bosphorus and Dardanelles Straits was made available by Erkan Gökaşan (Gökaşan et al. 2005 , 2007 ), with the permission of the Turkish Navy, Navigation, Hydrography and Oceanography Office. The datasets were combined through a bilinear interpolation on the computational grid, followed by manual check for isolated grid points, islands, and narrow passages (see Sannino et al., 2015 , 2017 ). Both equilibrium tide (i.e. the generating potential is incorporated in the momentum equations as a body force) and tide propagating from the Atlantic Ocean through the Atlantic open boundary, are explicitly applied as in Naranjo et al. 2014 , Sannino et al. 2015 , and Palma et al. 2020 . The equilibrium tide is composed of four tidal components: M2, O1, S2, K1, the semidiurnal and diurnal principal lunar tides, the principal semidiurnal solar tide, and diurnal lunisolar declination tide, respectively. The tidal values used to prescribe the Atlantic tide are derived from the OTIS global tide inverse model (Egbert and Erofeeva, 2002 ). Differently from the model configuration of Naranjo et al. 2014 , and Sannino et al. 2015 the updated version used in this study includes the Black Sea that is interactively connected to the Mediterranean Sea through the Turkish Straits (Dardanelles and Bosphorus). The model domain therefore covers the whole Mediterranean-Black Sea system and a small part of the Atlantic Ocean west of the Strait of Gibraltar (SoG hereafter), whose western limit corresponds to the only open-boundary condition prescribed. The horizontal computational grid has a uniform resolution of 1/16° (about 7 km), which is significantly increased in correspondence of the three straits where higher resolution is needed to reliably represent the local dynamics, i.e., the SoG (maximum resolution of about 200 m), and the Dardanelles and Bosphorus Straits, where a smoothly varying refinement in the latitudinal and longitudinal directions allows to reach a maximum resolution of 1/200° (about 555 m) (Fig. 1 ). The vertical domain is discretized using 100 z-levels, with grid spacing increasing from 2 m near the surface to 62 m at a depth of 1500 m; below 1500 m a uniform grid spacing of 62 m is used down to the sea floor. The time-step used is one minute. To calculate the vertical eddy viscosity and diffusivity coefficients, we used the 1.5 order Turbulent Kinetic Energy (TKE) closure scheme by Gaspar et al., 1990 , adapted from the atmospheric case developed by Bougeault and Lacarrere (1989). The background vertical eddy viscosity and diffusivity were set to 1.5 10 − 6 m 2 /s and 10 − 7 m 2 /s, respectively. The maximal value of diffusivity allowed to be generated by the turbulence scheme is 100 m 2 /s, in order to let it handle unstable vertical stratification and avoid the use of an enhanced vertical mixing parameterization. A spatial-dependent horizontal viscosity is obtained from the turbulence closure scheme by Leith ( 1968 ). Differently from the well-known scheme by Smagorinsky ( 1963 ), Leith’s scheme focuses on resolving the direct enstrophy cascade (cascade towards the smaller scales) that is characteristic of 2D turbulence (Fox-Kemper and Menemenlis, 2008 ). A constant horizontal diffusivity coefficient (2 m 2 sec − 1 ) is applied with a laplacian operator for the tracers (T,S). The no-slip conditions were used at the lateral and bottom solid boundaries, along with a quadratic bottom drag. The latter is calculated as a function of the velocity close to the bottom, with a dimensionless coefficient of 0.0025. The selected tracer advection scheme is a third-order direct space-time flux limited scheme. As the low spatial resolution of the global model did not allow an accurate representation of the dynamics within the Strait of Gibraltar, the Mediterranean basin was completely detached from the Atlantic Ocean in the global simulation, by artificially closing the Strait. The Mediterranean was therefore virtually represented as a lake in the global projection, thus impairing its use to initialize the regional ocean simulation. Initial conditions (namely temperature and salinity) were thus derived from MEDHYMAP climatological data. To reduce spurious drift, MED16 was spun-up 35 years, using the hindcast forcing for the 1980–1986 period in a five-time loop. The span-up model was then used as initial condition for both historical and hindcast simulations. At the Atlantic open boundary, consistently with the surface atmospheric forcing, the hindcast simulation was forced with ORAS4 global reanalysis (Balmaseda et al. 2013 ), while for historical and RCP8.5 scenario simulations monthly data from the HadGEM2-ES global model were used. In particular, the lateral boundary conditions consist of monthly mean temperature, salinity and sea surface height (SSH) which are interpolated onto the MED16 grid. Temperature and salinity are applied through a 3D relaxation with a relaxation time varying linearly from 2 hours at the western limit of the domain to 30 days over the first 30 grid points. The prescribed SSH on the Atlantic box includes different contributions depending on the forcing. For the hindcast simulation, the prescribed ORAS4 SSH includes sea level contributions from ice sheet mass loss, glaciers ice melt, changes in land water storage, as well as global thermal expansion. As ORAS4 underestimates the regional SSH seasonal cycle in the near Atlantic region compared with the multi-satellite products provided by the Climate Change Initiative (CCI-SLA doi: 10.5270/esa-sea_level_cci-MSLA-1993_2013-v_1.1-201412 ), we have applied a 12-month correction based on the difference between CCI-SLA and ORAS4 as described in Adloff et al. ( 2018 ). For both historical and scenario simulations, the prescribed SSH includes the dynamic sea level (DSL) and the global mean thermal expansion. DSL is the height of the ocean with respect to the time-invariant geoid that is determined by the dynamical balance associated with ocean density and circulation. DSL includes the regional variability of dynamic topography changes due to water mass advection, thermohaline circulation and to the wind-driven circulation (Gill and Niller 1973 ) and are part of the standard output in the CMIP5 simulations (variable zos ). zos was initially de-drifted by removing the linearly fitted HadGEM2-ES control run (which is forced by non-evolving pre-industrial conditions) from each grid point individually. This step removes any artificial signals associated with ongoing spin-up deep ocean and/or limitations in the representation of energy conservation in the model domain, as discussed by Sen Gupta et al. ( 2013 ). As the ocean component of HadGEM2-ES is based on the Boussinesq approximation and conserves volume rather than mass (Greatbatch, 1994 ), the value of zos was further corrected (at each time step and grid point) by subtracting its time dependent global mean. This correction guarantees that the resulting sea-level patterns only reflect fluctuations due to the joint effects of ocean density and circulation (Gregory et al., 2019 ) and thus, it is comparable to those resulting from non-Boussinesq models (Losch et al., 2004 ; Griffies and Greatbatch, 2012 ). The corresponding global thermal expansion time series (variable zostoga ) was also corrected for control drift by removing a quadratic fit to the control run’s thermal expansion time series before being added to the detrended and zero global mean zos field. The resulting sea level variable was then used as SSH lateral boundary condition for the MED16 historical and scenario simulations. As in Naranjo et al 2014 , Sannino et al 2015 , and the higher resolution Mediterranean version of Palma et al ( 2020 ), SSH and the additional 4 tidal constituents were directly prescribed at the western open boundary. The remaining sea level components for the historical and scenario simulations, namely the surface mass balance and dynamic ice sheet contributions from Greenland and Antarctica, the glacier and land water storage contributions, and the Glacial Isostatic Adjustment (GIA), were extracted from the Integrated Climate Data Center (ICDC) at Hamburg University ( https://icdc.cen.uni-hamburg.de/en/ar5-slr.html ), interpolated on the MED16 grid, and added offline. The ICDC dataset is based on the ensemble mean of CMIP5 climate models that were used in the regional sea-level projections of the AR5 (Church et al. 2013a ) of the IPCC. ICDC provides data on a global grid at 1°×1° spatial resolution for different RCP scenarios. In this study, we used central estimation data of the RCP8.5 ensemble mean. River discharge at the river outlets is prescribed by transporting the daily total runoff computed by the forcing atmospheric regional model, along the river network of the WBMplus model (Vörösmarty et al. 1998 ; Wisser et al. 2008 , 2010 ), by means of a Muskingum-Cunge type scheme that solves the Saint-Venant flow equations (Ponce 1994 ). Only for rivers discharging in the Black Sea, the so computed discharge was bias-corrected to reproduce monthly climatological values reported in the literature (Kara et al. 2008 ). 3. Model Validation Before analysing the model sea level reconstructions (see Sect. 4 ), here we perform a first validation of the hindcast and historical runs. In particular, we compare the simulated water transports at the straits and the Mediterranean hydrology and circulation with the observations, and with the results of a recent reanalysis of Mediterranean water properties by Escudier et al ( 2020 ). The latter is a product from the Copernicus CMEMS database ( http://marine.copernicus.eu ) that covers the period 1987–2019, and results from the assimilation of temperature and salinity vertical profiles and satellite Sea Level Anomaly along track data into a numerical system, consisting of the Nucleus for European Modelling of the Ocean (NEMO), and of a variational data assimilation scheme (OceanVAR). As our focus is on the Mediterranean Sea, we will mainly show results for this basin, and report the projected water exchanges at the Dardanelles as a boundary condition, determined by the Black Sea effectively acting as a reservoir. Nevertheless, it is worth noting that the model correctly reproduces the permanent basin-scale, cyclonic boundary current that characterizes the Black Sea circulation. 3.1 Transports at the main straits: Gibraltar, Dardanelles, Bosphorus The explicit treatment of water inflow and outflow at the open boundary of the Mediterranean Sea, coupled with explicit tidal forcing, allows to realistically constrain the mass, energy and momentum exchanges with adjacent basins, as well as to finely modulate the temperature and salinity properties of either flow, in both time and space (see Sannino et al, 2015 , for a detailed description). Table 1 reports the simulated averaged long-term (1980–2011) transports across the main straits (Gibraltar, Dardanelles, and Bosphorus), for both the hindcast and historical experiments. The sections used to compute transports are located at 5.4° W for the SoG, at 26.2° E for the Dardanelles Strait, and at 41° N for the Bosphorus Strait. Transport is positive for an eastward or northward flow. Table 1 Averaged water transport (in m 3 /s x 10 3 ) through the Strait of Gibraltar, Dardanelles and Bosphorus over the period 1982–2011. [Mean value and Monthly standard deviation]. A negative flow is directed towards the Atlantic for the Strait of Gibraltar, and into the Mediterranean Sea for the Dardanelles and the Bosphorus NET Gibraltar Dardanelles Bosphorus Hindcast 60 ± 90 -2 ± 7 -1 ± 8 Historical 60 ± 60 -4 ± 9 -4 ± 9 INFLOW Gibraltar Dardanelles Bosphorus Hindcast 930 ± 90 15 ± 2 5 ± 4 Historical 1080 ± 80 14 ± 3 5 ± 3 OUTFLOW Gibraltar Dardanelles Bosphorus Hindcast -870 ± 100 -17 ± 7 -7 ± 4 Historical -1020 ± 70 -18 ± 7 -9 ± 6 For the hindcast simulation the net average transport of 0.06 Sv at the Strait of Gibraltar results from a mean inflow from the Atlantic of about 0.93 Sv and a Mediterranean outflow of about − 0.87 Sv, and compares favorably with previous estimates based on the evaluation of the Mediterranean long-term hydrological deficit (E-P-R), which needs to be compensated by the excess Atlantic inflow. As a matter of fact, a recent work by Jordà et al. ( 2017b ) reports values ranging from 0.04 Sv (Soto-Navarro et al., 2010 ) to 0.06 Sv (Boutov et al., 2014 ), while proposing an intermediate estimate of 0.05 ± 0.03 Sv based on an independent evaluation of the hydrological budget, and a lower bound of 0.03 Sv deriving from the optimal interpolation (OI) of Acoustic Doppler Current Profiler (ADCP) and tide gauge data. The projected average outflow is also close to previous estimates reported in the literature, which range from − 0.67 Sv (Tsimplis and Bryden, 2000 ) to -0.97 Sv (Candela 2001 ), and include the − 0.78 Sv estimate of Sánchez Román et al. ( 2009 ), obtained by combining three years of ADCP observations and numerical modelling. A higher estimated value of − 0.865 ± 0.136 Sv has been recently obtained by García-Lafuente et al. ( 2021 ) from the analysis of experimental data recorded over the period 2004–2012. The averaged simulated Atlantic inflow is 0.93 Sv, which is also in the range of published values (e.g. 0.78 Sv in Tsimplis and Bryden 2000 ; 0.96 Sv in García-Lafuente et al. 2002 ). The net transport for the historical experiment is in excellent agreement with the hindcast value, although exhibiting reduced variability, yet it results from higher inflow and outflow. The net transport across both Turkish Straits (TSS) is on average negative, consistently with the observed net inflow of fresher waters into the Aegean Sea from the Black Sea, where precipitation and river discharge exceed evaporation. The magnitude of the inflow appears to be lower than previous observational and numerical estimates, which range from 0.006 Sv to 0.01 Sv (Ünlüata et al., 1990 ; Simonov and Altman, 1991 ; Peneva et al., 2001 ; Kara et al., 2008 ; Sannino et al., 2017 ). However, fluxes through the TSS are difficult to accurately measure, due to the large variability in the observed currents and the consequent low signal-to-noise ratio, which might need longer-term monitoring to be adequately increased (Sannino et al., 2017 ). On the other hand, the simulated net transport is expected to significantly benefit by further refining the grid in correspondence of both Straits, as demonstrated by Sannino et al. ( 2017 ), where a very-high-resolution simulation was performed to accurately reproduce transports with the prescription of climatological barotropic forcing at the two open boundaries. Increased resolution together with a specific treatment of local viscosity and diffusivity is therefore liable to enhance water exchanges across the straits, and to further improve the already good model performance. 3.2 Basin hydrology and circulation We first compare model climatologies of the winter and summer sea surface temperature (SST) and SSS with reanalysis data; mean fields are computed over the common period 1987–2005. In Fig. 2 , panels (a-b) are for the Mediterranean Optimally Interpolated Sea Surface Temperature (OISST) reanalysis of Marullo et al. ( 2007 ), panels (c-d) for the Copernicus reanalysis, panels (e-f) for the hindcast, and panels (g-h) for the historical simulation. The model simulations are in an overall good agreement with the reanalyses, even though MED16 systematically produces lower temperatures in winter (more so in the historical simulation). In summer, the SST from the hindcast is in excellent agreement with both reference fields, whereas the historical simulation displays slightly warmer temperatures in the western basin, in the central part of the eastern basin and in the Adriatic Sea, and lower ones in the Ionian, and in the easternmost part of the Levantine. In Fig. 3 , the SSS model climatologies (panels c-d, hindcast; panels e-f, historical), are compared with the Copernicus renalysis (panels a-b). In the western basin, the hindcast salinity appears to slightly overestimate the observations in both seasons, although less notably so in summer. In the eastern basin, the salinity field is generally well reproduced in winter, while in summer some local negative biases are present (e.g, in the Gulf of Gabes, in the Adriatic Sea, and in the north Aegean). The historical experiment, on the contrary, exhibits larger salinity values in both seasons and basins, yet reducing the positive bias around the Balearic Islands, coherently with the limited intrusion of the Atlantic Water (AW) into the Tyrrhenian Sea, the Sicily channel and the Gulf of Gabes, and with the more pronounced northward spreading of the Atlantic inflow after exiting the Alboran Sea. Specific local differences between the numerical experiments and the reanalysis can be attributed to the different treatment of rivers and/or to the different magnitudes of the runoff provided by the driving atmospheric simulations (e.g. along the eastern Adriatic coast and in the North Adriatic Sea), while the different behaviour of the two MED16 runs in the North Aegean is probably related to the representation of the local circulation, as no significant difference was detected in the representation of the net inflow from the Dardanelles between the hindcast and the historical experiment. This discrepancy will be the object of future research. The hydrological structure and the circulation produced by the hindcast and historical simulations were further validated via comparison with in situ and remote observations. Vertical means of the modelled and reanalysed temperature and salinity were computed over the 1987–2005 time window, for three different vertical layers, and averaged over the whole Mediterranean basin (MED), and the western and eastern sub-basins (WMED and EMED). Table 2 shows the reference values from the reanalysis, together with the differences between model and reference for the two experiments, for the surface layer (0 -150m), the intermediate layer (150-600m), and the deep layer (600-3500m). Table 2 For each geographical area in the leftmost column - the whole Mediterranean (MED) and the Western (WMED) and Eastern (EMED) sub-basins - the first row reports the average temperature ( °C ) and salinity ( Psu ) at various depths, from the Copernicus Reanalysis dataset; the second and third rows (labelled “ hindcast ” and “ historical ”) respectively report the associated differences between the hindcast and the historical simulations and the reanalyses. All spatial averages are taken over the period 1987–2005 Temperature Salinity Depth [m] Depth [m] 0-150 150–600 600–3500 0-150 150–600 600–3500 MED Reanalysis 16,54 14,00 13,32 38,39 38,81 38,65 Hindcast -0,64 -0,14 0,05 -0,12 -0,15 -0,04 Historical -0,78 -0,27 0,18 0,04 -0,08 0,04 WMED Reanalysis 15,37 13,45 12,93 37,83 38,49 38,47 Hindcast -0,39 -0,29 0,01 -0,04 -0,10 0,01 Historical -0,57 -0,38 0,09 0,01 -0,05 0,07 EMED Reanalysis 17,17 14,34 13,55 38,70 38,84 38,80 Hindcast -0,74 -0,06 0,08 -0,15 -0,03 0,00 Historical -0,86 -0,21 0,26 0,07 0,05 0,09 Despite the differences highlighted in the spatial patterns, the agreement between the hindcast and historical simulation and the reference data is good. However, the modelled temperatures are always slightly lower than the observations in the two upper layers and slightly higher in the deep layer. For both simulations and all geographical areas, maximum discrepancies are anyway found in the surface layer, with largest differences in the Eastern basin. When the whole basin is considered, modelled salinity also generally presents negative anomalies, with maximum negative differences in the surface and intermediate layers of the hindcast simulation, while slightly positive anomalies can be observed in the surface and bottom layers for the historical simulation. When separately considering the two sub-basins, the hindcast experiment exhibits more marked negative biases in the intermediate layer of the Western basin and in the surface layer of the Eastern basin, while the historical simulation projects higher salinity in the surface layer of the Eastern basin with respect to the reanalysis. We note that the biases for temperature and salinity are well inside the ensemble spread computed by Llasses et al. ( 2018 ) from the Med-Cordex simulations. In order to evaluate the ability of the hindcast experiment in reproducing the time evolution of water temperature, Fig. 4 shows the time series (1987–2010) of the mean annual temperature anomalies in the surface and intermediate layers, for the whole basin and the western and eastern sub-basins, as simulated by the hincast experiment. Comparison with the historical simulation is omitted here as time coherence with the reanalysis could not be expected. The observed interannual variability is generally well captured by the hindcast simulation, both in the surface and in the intermediate layer. The simulation correctly reproduces the observed interannual variability. The EMED time series, in particular, captures the positive trend in both the upper and intermediate layers. Moreover, the simulation well represents the strong cooling during the years 1992 and 1993, that is the signature of the Eastern Mediterranean Transient (EMT) phenomenon. Analogous considerations apply to the time series of mean annual salinity anomalies (Fig. 5 ). The general trend of simulated salinity and reanalysis are also similar, except for the intermediate layer of WMED where the simulated salinity does not follow the reanalysis trend. In the WMED the model captures both the sudden drop in salinity around 1990 and the maximum in 2005. The first event is due to an excess of freshwater input and is confined to the upper layers, whereas the latter is related to a well-known event of intense dense water formation in the Gulf of Lion. A good representation of the interannual variability of the salinity field is also achieved in the EMED, with a local maximum during the 90s and an increasing trend in the second half of the simulation. A notable discrepancy occurs in the years from 2001 to 2004, in which the reanalysis displays a relative minimum, particularly strong in the WMED, which is not present in the model. However, the corresponding data from the MEDHYMAP dataset show a very good agreement with MED16 also for the period 2000–2005. Figure 6 shows the surface (15 m of depth) and intermediate (300 m of depth) model circulations, averaged over the simulation time-span (1982–2010). At both depths, the model circulation is close to the 1987–2007 time-mean flow displayed in Pinardi et al. ( 2015 ), which results from a reanalysis of the Mediterranean Sea circulation (see their Figs. 3 and 4 , displaying the circulation at 15 m of depth, and the average circulation in the layer from 200 m and 300 m of depth). The main Mediterranean surface current systems, either driven by the local wind forcing, or by the inflow of AW at the Gibraltar Strait, are present both in our hindcast and in the reanalysis, with similar shapes. This also applies to most of the robust cyclonic and anticyclonic circulation structures, even though there are a few features, such as the quasi-permanent cyclonic circulation to the southeast of Sardinia, and some smaller vortices in the eastern basin, that are not well defined in the hindcast. Overall, considering that the latter is the result of a “free” simulation spanning almost three decades, whereas the reanalysis relies on the assimilation of observed data, the representation of the average Mediterranean circulation provided by our model can be considered quite satisfactory. Three hindcast sub-regional surface circulation maps are displayed in Fig. 7 . Panel (a) of the figure highlights the two permanent cyclonic circulations that characterize the northern portion of the western Mediterranean, that is, the wide gyre formed by the Liguro-Provençal current, occupying the western Ligurian Sea, and the Bonifacio cyclone, in the northern part of the Tyrrhenian Sea, to the east of Corsica. Interestingly, between the two cyclonic circulations, there is the weak signature of an anticyclonic circulation, in the area of the Corsica Channel, between Cap Corse (the long peninsula in the northern tip of Corsica) and the Italian coast. This structure, which was denoted as Ligurian anticyclone in Ciuffardi et al. ( 2016 ), appears to be a robust feature of the local dynamics during the summer months (see also recent works by Poulain et al., 2020 , and by Iacono and Napolitano, 2020 ). As shown in Fig. 7 b, the model correctly reproduces the wide cyclonic structure, with two poles, that occupies the southern, deeper portion of the Adriatic Sea (see, e.g., recent work by Palma et al., 2020 , and references therein). Finally, we note that the surface dynamics of the Marmara Sea (Fig. 7 c) are dominated by a central, elongated gyre, more attached to the northern coast, in agreement with what was found in a recent high-resolution numerical investigation of the Turkish Straits System by Sannino et al. ( 2017 ). The time evolution of the Mixed Layer Depth (MLD - Fig. 8 ) highlights how deep-water-formation processes in the Gulf of Lion and the Adriatic basin are well represented both in the hindcast and in the historical experiments. 4. Hindcast And Historical Sea-level Analysis The Mediterranean Sea level average and variability simulated in the hindcast experiment are compared to their observed counterparts, derived from altimetric satellite sea level anomaly (SLA) measurements. The model ability to reproduce nearshore sea level is assessed by means of tide gauge data. 4.1 Comparison with satellite observations The altimeter data used are daily maps of gridded L4 SLA and derived variables reprocessed (Copernicus portal, http://marine.copernicus.eu , last access april 2021), obtained from the merging of track data from different satellites. These maps, especially developed for the Mediterranean Sea, have a nominal resolution of 1/8° × 1/8°, which is twice the resolution of the products that cover the world ocean. They have been available since 1993, overlapping the last 18 years of the simulation. A different altimeter dataset was used in Adloff et al. ( 2018 ) (see Ablain et al., 2015 , for details), which consists of monthly SLA maps with a spatial resolution of 1/4°. The average seasonal cycle reproduced by the hindcast is compared to that of the observations in Fig. 9 , where the solid lines with black dots refer to model results and the dashed lines with diamonds to altimeter data (the averages are over the period 1993–2010). Panel (a) shows that the model cycle over the whole basin is fairly close to the observations, with some overestimation in winter and some underestimation at the end of summer and in part of the autumn. Overall, the model cycle is close to that obtained from the NEMO-MED12 simulation discussed in Adloff et al. ( 2018 ), which was found to provide the best SL reconstruction among the four considered in that work. The two other panels in the figure show the same comparison for the western and eastern sub-basins, respectively. The cycle for the western Mediterranean is even closer to the observations than the basin cycle, whereas in the eastern basin, where the amplitude of the observed cycle is slightly larger, the model underestimation during SON is also larger. This may be due to the fact that the steric contribution to the SL, which is in principle absent in most state-of-the art circulation models, is here introduced indirectly through the Atlantic boundary conditions. As we have seen, this works pretty well, particularly in the western basin, but it may be not sufficient to account for additional sea-level seasonal variations induced by local heat fluxes, especially in the eastern sub-basin. The interannual variability of the sea-level anomaly of the hindcast simulation is shown in Fig. 10 , where the same notations of the previous figure are adopted. The plotted values are yearly anomalies with respect to the mean over the period covered by the altimeter (1993–2010), averaged over four different regions: the Atlantic box, the whole Mediterranean basin, the WMED, and the Levantine sub-basin. The first panel is relative to the Atlantic box, i.e., the small domain to the west of the Gibraltar Strait where boundary conditions are prescribed, making use of the available altimetric data (see Sect. 2 for details). From 1993 on, the model time series appears well correlated with the observations, showing that the boundary conditions are correctly applied in our model. In fact, the cross-correlation coefficients of Table 3 show that the model time series are very well correlated with the corresponding altimeter-derived time series in all the basins in consideration. Table 3 Cross-correlation between the yearly sea level anomaly time series and that of the SLA, over the period 1993–2010, for four different basins Atlantic box Mediterranean Western Med. Levantine Correlation with SLA 0.81 0.81 0.78 0.87 The time series of the average Mediterranean SL anomaly (top right panel in Fig. 10 ) reproduces the growing basin-scale trend displayed by the altimetric observations (see Bonaduce et al., 2016 , and the more recent investigation by Mohamed et al., 2019 , where trends of the SL in the Mediterranean were discussed, based on 25 years of altimeter data). It also captures short-time variations, such as the sharp increase seen in 2010, which results from a large wintertime jump of the SL over the whole Mediterranean Sea. As shown in Landerer and Volkov ( 2013 ), this was due to a basin-wide barotropic fluctuation induced by the variation of the wind stress in the Gibraltar area and in the neighboring Atlantic sector, associated with inflow of Atlantic water through the Gibraltar Strait. A similar event occurred during the following winter of 2011. Comparison with Fig. 4 of Adloff et al. ( 2018 ) shows that the model time series of Mediterranean anomalies also captures the main interannual variability observed before 1993; the reference for this period is provided by two reconstructions of the Mediterranean SL by Calafat and Jordà ( 2011 ), and Meyssignac et al. ( 2011 ), based on data from coastal tide gauges. The series is also in good agreement with that obtained from the NEMO-MED12 simulation of Adloff et al. ( 2018 ). This is due to the fact that, despite the many differences, the two models apply the same Atlantic boundary conditions, and use as local atmospheric forcing two different downscaling of the same dataset (ERA-INTERIM). In some cases, the model series displays stronger SL variations; this may be ascribed to model differences, but could also partly be due to the contribution of mesoscale features that are better resolved by the model, thanks to the higher horizontal resolution (1/16°). This holds a fortiori, when compared with the altimeter data, whose nominal resolution is 1/8°x1/8°. The lower row of Fig. 10 shows the SL anomalies for the sub-basin that is closest to the Atlantic (western Mediterranean) and for that more far away from it (Levantine basin, to the east of 22° of longitude). The signals are quite different in the two sub-basins; the series in the western Mediterranean resembles that in the Atlantic box, with a fairly regular increase of the SL over the simulation period, whereas that in the Levantine shows a decrease of the SL at the end of the 80’s and beginning of the 90’s, followed by a sharp increase, leading to a different state over the last 10 years of the simulation. The latter behavior is of course related to the Eastern Mediterranean Transient (EMT), the major event that has affected the circulation and water mass properties of the Mediterranean in the last decades. The hindcast is capable of reproducing the rapid SL rise as a result of strong anomalous mass modification in the deep layers of the EMED during the end of the 80s and the beginning of the 90s; the effects of the EMT on sea level has been discussed for the first time by Tsimplis and Josey ( 2001 ). The model correlates very well with the altimeter observations after the transient, which display a definitely nonlinear trend, first evidenced in Amitai et al. ( 2010 ). We can conclude that the hindcast simulation not only captures the basin scale variability of the SL over the hindcast period, but also some expected, significant differences between the main Mediterranean sub-basins. In Fig. 11 , the modelled (both hindcast and historical) SSH fields, averaged over the period 1993–2011, are compared with the corresponding climatological average of the absolute dynamic topography (ADT) from altimeter data (AVISO, http://www.aviso.oceanobs.com , last access May 10th 2021). In all cases, the respective regional climatological means have been subtracted to ease the comparison of patterns. Being a long-term average, the field derived from the AVISO data is expected to be very close to the Mean Dynamic Topography (MDT), that is, to the mean signal, representative of the average circulation, which produces the ADT when added to the zero-mean remotely measured sea level anomaly (SLA). The reference observational field shown in Fig. 11 is indeed very close to that of Rio et al. ( 2014 ), which was constructed using observations and model results, and validated using independent observations. The spatial pattern of the modelled SSH field generally appears to be in good agreement with the observations, and is also similar to that obtained by the NEMO-MED12 model, the best performing among those considered in Adloff et al. ( 2018 ). As to the hindcast experiment, the MED16 model outperforms NEMO-MED12 in the western portion of the basin, indicating that the stretched grid at Gibraltar indeed allows for a better representation of the complex dynamics inside the Strait. Nevertheless, the signatures of the Rhodes gyre and of the western Cretan gyre are weaker than expected. On the other hand, neither the MED16 simulation nor the four hindcasts analyzed in Adloff et al. ( 2018 ) resolve the Ierapetra gyre, possibly due to the still insufficient spatial resolution or to the quality of the downscaled wind forcing in the area, the Ierapetra gyre being a wind driven feature. The SSH values obtained for the Black Sea (not shown here) exhibit larger spatial variability with respect to those of the Mediterranean Sea, approximately ranging from − 6 cm to 18 cm, and compare well with the MDT reconstructions by Knysh et al. ( 2002 ) and by Kubryakov and Stanichny ( 2011 ). In particular, Kubryakov and Stanichny ( 2011 ) find localized, deeper minima in the core of the dominant cyclonic cell, while MED16 produces significantly higher SSH values than either reconstruction in the westernmost portion of the same cell. 4.2 Comparison with tide gauges Hindcast sea surface heights along the coast have been compared with tide gauge data retrieved from the Permanent Service for Mean Sea Level (PSMSL) (Holgate et al., 2013 ; PSMSL, 2020). In particular, we used time series of sea level monthly means from the Revised Local Reference dataset (RLR), which are independently calibrated for each station in the PSMSL according to the tide gauge history provided by each local supplying authority. However, no correction is made by PSMSL for vertical land movements (VLM), although tide gauges without known benchmarks are omitted from this data set by PSMSL prior to release (Hamlington et al., 2016 ). These data therefore measure the relative movement of the sea surface with respect to the land, disregarding the multiple processes over a variety of time scales that can affect MSL estimates, some of natural origin, such as the glacial isostatic adjustment (GIA) and plate tectonics, and some related to human activities (e.g. ground-water pumping). In the following, RLR data have been used without any further correction for VLM, selecting all the available Mediterranean stations which covered at least 50% of the simulated period. In order to avoid relative biases due to different zero-level assumptions, monthly data from the model grid point nearest to the tide gauge have been compared with the monthly mean tide gauge data after subtraction of the respective mean values, over the common time interval, from both modelled and tide gauge time series. The 34 locations for which the comparison was carried out are shown in Fig. 12 . The cross-correlation coefficients obtained are detailed in Table 4 , where the geographic location of the stations and the number of observations are also indicated. The coefficients are pretty high in most of the stations; they are smaller than 0.65 only in 8 stations (17, 23, 25, 26, 28, 29, 31, 34) located in the Aegean Sea or nearby, and one in the Black Sea. This is quite encouraging, considering that the model values are not at the stations’ sites, which are very close to the coast, and that the model freely evolves for three decades. It shows that, in part of the basin, the model is capable of providing a fairly accurate description of the seal level variability on long time scales, not only over large spatial scales, but also nearshore, where the consequences of sea level change are stronger. The fact that the model performs less well in the Aegean Sea is likely partly due to difficulties in reproducing small scale details of the near coast dynamics, induced by the very complex bathymetry of the area. Table 4 List of tide gauges used with the corresponding code number and coordinates. Correlation between the observed time series and model data is shown together with the standard deviation of the two series of data and the number of times considered number station code longitude latitude correlation n. times std station std model 1 TARIFA 488 -5,616 36,020 0,651 321 6,821 4,177 2 ALGECIRAS 490 -5,422 36,132 0,717 249 5,974 3,936 3 CEUTA 498 -5,299 35,910 0,659 330 5,502 4,205 4 MALAGA 496 -4,407 36,668 0,777 349 7,606 4,469 5 PORT VENDRES 1469 3,094 42,594 0,766 210 7,952 6,113 6 L'ESTARTIT 1764 3,281 42,031 0,825 241 7,472 6,630 7 MARSEILLE 61 5,344 43,281 0,797 316 7,207 6,318 8 TOULON 980 5,906 42,969 0,711 217 7,041 6,772 9 NICE 1468 7,281 43,656 0,717 300 7,101 6,373 10 VENEZIA (PUNTA DELLA SALUTE) 168 12,406 45,281 0,813 239 8,804 8,607 11 ROVINJ 761 13,656 45,094 0,799 349 8,246 8,630 12 TRIESTE 154 13,781 45,656 0,810 349 8,450 8,718 13 SPLIT - GRADSKA LUKA 352 16,406 43,344 0,761 349 7,925 8,293 14 SUCURAJ 1706 16,906 43,031 0,732 221 8,431 8,321 15 DUBROVNIK 760 18,031 42,656 0,754 345 7,568 8,049 16 LEVKAS 1239 20,719 38,906 0,702 310 9,384 7,356 17 PREVEZA 410 20,719 38,906 0,585 301 7,579 7,380 18 KATAKOLON 1240 21,344 37,656 0,717 300 7,720 7,412 19 KALAMAI 411 22,156 36,969 0,770 276 7,947 7,677 20 THESSALONIKI 373 22,844 40,536 0,680 329 8,847 8,014 21 NORTH SALAMINOS 1604 23,531 37,906 0,735 180 8,305 7,856 22 SOUDHAS 1232 24,094 35,781 0,678 338 7,668 7,607 23 SIROS 1234 24,969 37,531 0,577 317 8,687 7,644 24 IRAKLION 634 25,156 35,531 0,669 235 10,500 7,208 25 ALEXANDROUPOLIS 1238 25,906 40,825 0,635 312 8,132 7,820 26 KHIOS 408 26,156 38,656 0,622 316 8,108 7,746 27 MENTES/IZMIR 1679 26,719 38,594 0,686 265 8,062 7,785 28 LEROS 1233 26,844 37,094 0,522 300 6,432 6,800 29 RHODOS 1243 28,219 36,469 0,649 220 7,741 6,415 30 ALEXANDRIA 503 29,906 31,281 0,708 301 9,142 7,470 31 ERDEK 1598 27,856 40,383 0,478 255 9,030 6,460 32 ANTALYA II 1681 30,719 36,781 0,736 253 9,541 7,811 33 TUAPSE 215 39,125 44,026 0,724 346 9,428 9,200 34 POTI 41 41,622 42,192 0,533 336 12,538 9,128 Sea surface height time series for six stations (red circles in Fig. 12 ) where VLM can be considered to be negligible are plotted in Fig. 13 . The main interannual variability is reproduced fairly well at all stations, and especially at Malaga, Trieste, Preveza and Alexandria. 5. Scenario Simulation 5.1 Future basin hydrology and circulation For the RCP8.5 scenario, Fig. 14 shows the differences in the future surface temperature climatologies with respect to the reference period. Averages are computed over years 2046–2065 (left) and 2081–2100 (right), for winter (top) and summer (bottom). The sustained increase in temperature is evident in both seasons. As for the correspondent salinity fields (Fig. 15 ), the salinification of the Eastern basin is apparent, especially in the north, as well as the northward displacement of the Atlantic stream in the Western basin, yet accompanied by a more efficient penetration of the AW into the Ionian basin under the future scenario with respect to the historical simulation. The temperature-induced additional buoyancy is evident in the mixed layer depth which exhibits significantly lower relative maxima at dense water formation sites. The deep convection in the Gulf of Lion is almost neutralised, whereas in the Adriatic Sea is characterised by enhanced variability (Fig. 16 ). Basin-wide averages of T and S are summarized in Table 5 , where substantial increases in temperature generally characterize the surface and intermediate layers, while variations in salinity appear to be most relevant and coherent in the intermediate layer. Results are consistent with the findings of Soto-Navarro et al. ( 2020 ). Table 5 Average temperature ( °C ) and salinity ( Psu ) at different depths. Differences between values averaged over two periods (2046–2065 and 2081–2100) of the scenario simulation and the whole period (1981–2005) of the historical simulation. Averages are over the whole Mediterranean Sea (MED), and over the western and eastern sub-basins (WMED and EMED) Temperature Salinity Depth [m] Depth [m] 0-150 150–600 600–3500 0-150 150–600 600–3500 MED RCP 2046–2065 1,72 1,48 0,25 -0,03 0,22 0,00 RCP 2081–2100 3,31 2,76 0,56 0,20 0,44 0,06 WMED RCP 2046–2065 1,46 1,44 0,45 -0,10 0,27 0,07 RCP 2081–2100 2,98 2,76 0,83 -0,01 0,47 0,15 EMED RCP 2046–2065 1,87 1,49 0,11 0,02 0,20 -0,03 RCP 2081–2100 3,50 2,76 0,36 0,32 0,42 0,02 5.2 Impact of dynamical downscaling on the projected SLC Figure 17 separately quantifies the magnitude and sign of the components, either local or global, that contribute to a projected mean sea level trend of about 7 mm/yr in the Mediterranean, and allows the evaluation of their relative weight. Under RCP8.5 and from 2005 to 2100, it shows the time evolution of each respective central estimate, the ocean dynamic component being derived from an ensemble of 16 GCM participating in the CMIP5 programme, after excluding simulations that do not explicitly represent the open connection between the Mediterranean Sea and the Atlantic Ocean ( https://icdc.cen.uni-hamburg.de/en/ar5-slr.html , Slangen et al., 2014 ). Prior to spatial averaging, all the variables have been interpolated onto the MED16 grid via a bilinear interpolation, with nearest-neighbor interpolation near the coasts, results from the MED16 regional projection are also shown. The zero level is set at the 1986–2005 mean that Slangen et al. ( 2014 ) consider as the common zero SSH anomaly value. The most relevant contribution is that from the sterodynamic term (dark red; solid line = GCM, dashed line = MED16), which accounts for changes in the circulation and density distribution of sea water and for the inverse barometer (IB) correction associated with atmospheric pressure patterns (Gregory et al., 2019 ). Positive changes (here listed in decreasing order) also derive from distributed glacier melting (aside from Greenland and Antarctica ice sheets - green), Antarctic dynamic ice loss (pink), Greenland surface mass balance (light blue), Greenland dynamic ice loss (blue) and glacial isostatic adjustment (red), ground water storage (yellow), while the Antarctic surface mass balance (dark green) determines a reduction in sea level. At the end of the XXI century, these components are projected to respectively account for 51%, 23%, 13%, 10%, 6%, 2.5%, 2%, -7% of the total, with an overall contribution of 16% from Greenland and 6% from Antarctica. Recent works (Spada, 2017 ; Santamaría-Gómez et al., 2017 ) have dwelled on the necessity to consistently include the GIA effect in updated estimates of SLC in the Mediterranean, based on the alternative datasets provided by Lambeck (Lambeck et al., 1998 ) and Peltier (Peltier et al., 2004), whose limits and reliability have been the subject of a lively scientific discussion (Peltier, 2002; Lambeck et al., 2002 ). Here, the GIA contribution is the one computed by Church et al. ( 2013b ) by averaging results from the ICE-5G model (Peltier et al., 2004) and the ANU ice sheet model (Lambeck et al. 1998 and subsequent improvements), using the updated SELEN code for the sea level equation (Spada and Stocchi 2006 , 2007 ). In the Mediterranean basin, however, despite its comparative importance for the assessment of relative sea level rise for present times, the GIA contribution appears to progressively lose relevance by the end of the XXI century, unless ongoing research revises the long-term ranking of individual contributions. Although only representing one of the numerous possible realizations that can be obtained by different GCM-RCM modelling chains, the MSL trend estimated via the MED16 scenario simulation is very similar to that of the GCM ensemble mean from Slangen at al. (2014), although it slightly accelerates from 2040 onwards. With respect to the coarse-resolution global average, the MED16 downscaling better characterizes the regional patterns of SLC arising from ocean dynamics, and it consequently retains enhanced variability. The time evolution of the difference in SSH between the Mediterranean basin interior and the neighbouring Atlantic Ocean (average over the region just outside the Gibraltar Strait, from 9W to 6W) is reported in Fig. 19 , for both the MED16 model (left panel) and the global ensemble mean (right panel). The red curves correspond to the sterodynamic term (i.e., to the dark red curves of Fig. 17 ), the yellow curve in the left panel represents the historical reference experiment, while the AVISO observations are plotted in black. Beside the good agreement between the AVISO data and the historical experiment, it is worth noting that the regional simulation is capable of capturing the hydraulic jump across the Gibraltar Strait (Fig. 18 ) that is completely missed by the global projection, due to the insufficient representation of the local dynamics (Sannino et al. 2009 , 2015 ). Nevertheless, the global experiment also exhibits the relative “sinking” of the Mediterranean with respect to the Atlantic, albeit following a trend that is half that of the regional experiment, and characterized by lower variability. Some coherence is yet detectable between the respective typical time-scales of variability, possibly when driven by large-scale patterns, as well as a comparable signal-to-noise ratio. However, the global projection seems to be reaching a new stable state towards the end of the XXI century that is not discernible in the regional experiment, a longer run being needed to confirm such a hypothesis. So far, the change in the relative level between the two basins can be estimated to be within centimeters, over an expected sea level increase on the order of tens of centimetres by the end of the century. Figure 20 illustrates the spatial distribution of the progressive increase in sea level under RCP8.5 with respect to the historical period 1985–2005, which qualitatively retains the overall pattern of the dynamic topography under current climate, where the largest increments apparently correspond to the Atlantic signal, mirroring the AW progress into the basin interior. The main notable difference appears to be the sustained amplified signal around the Balearic Islands, presumably itself a consequence of the northward displacement of the Atlantic inflow that numerical models associate with the newly detected Western Mid-Mediterranean Current (WMMC). This current corresponds to the northward meandering stream arising from the bifurcation of the Atlantic inflow after exiting the Alboran Sea and overcoming the Almera-Oran front (Arnone et al., 1990 , Pinardi et al., 2015 ). The WMMC branch of the Atlantic inflow has also been detected in the hindcast and historical experiments (Fig. 6 ), yet it has never been well documented in observational datasets. Pinardi et al. ( 2015 ) tentatively interpret it as a residual current resulting from multi-year averaging. Figure 21 reports the differences between MED16 and the GCM ensemble mean for the two targeted time horizons. The overall pattern is maintained in time, with localized yet roughly constant negative biases (i.e., CGM estimates are higher) in the Levantine basin, and positive biases in the Ionian Sea and the Western basin, generally not exceeding 10÷12 cm, and an increasingly more pessimistic regional projection for the Balearic region, going from a deviation of 15÷17cm at mid-century, to one of 17÷22 cm in 2100. 6. Conclusions We have here analysed the results of three climatic simulations of the Mediterranean Sea circulation and sea level, performed with MED16, an updated version of the tide-including, climatic ocean model presented in Sannino et al., ( 2015 ). These consist of a hindcast simulation, covering three recent decades (1981–2010), a historical run (1981–2005), and a future climate simulation, under the RCP8.5 scenario, which starts from the end of the historical run and proceeds until the end of the present century (2006–2100). The hindcast simulation has been used to assess the ability of the new model to reproduce the past evolution of the basin circulation, with excellent results. This simulation has a value in itself, because it is the first pluri-decadal, eddy-resolving simulation of the Mediterranean Sea circulation that includes the main tidal effects, together with an accurate treatment of the complex dynamics occurring in the Gibraltar Strait area. We have therefore preliminarily analysed the simulated transports at the Gibraltar, Dardanelles and Bosphorus straits, and the basin hydrology and circulation. We found that the model correctly represents the transport evolution at Gibraltar, the average circulation in the surface and intermediate layers, and the main variability of the Mediterranean hydrologic structure during the simulation period. The simulated SLA was found in good agreement with satellite observations, exhibiting a positive trend in the Western Mediterranean Sea, and a more complex behavior in the Levantine basin, where sea level variability has been strongly influenced by the EMT and its evolution. The spatial structure of the hindcast average SSH field is well reproduced over the entire basin, especially in the western Mediterranean where results are much closer to that of the MDT, due to the fact that the use of improved sea level information at the Atlantic lateral boundary significantly enhances the reliability of results. A further improvement has been obtained through the adequate treatment of the complex, hydraulically driven dynamics across the Gibraltar Strait.The evolution of the hindcast SSH near coast compares very well with tide gauge observation along most of the basin’s coasts. This shows that MED16, if adequately forced, can provide useful information on the sea level variability in coastal areas, which are the more exposed to SLR, even though the current horizontal resolution of the model does not allow to exactly resolve the locations of the tide gauge stations. This also provides an indirect check of the general quality of the atmospheric forcing from the SMHI downscaling. This suggests that further improvement of the forcing may be needed to obtain an even more realistic description of the circulation structure. The historical simulation, which sets the reference baseline for the future scenario projections, also correctly reproduces the main mean features of the Mediterranean circulation and hydrology, a result that is all the more satisfactory when considering that no specific constraint or relaxation is prescribed. Under the RCP8.5 future scenario, the temperature is projected to generally increase. The spatial pattern of variations in surface salinity is affected by the penetration of the Atlantic stream into the basin interior, with the westernmost regions exhibiting appreciable decreases. The warming of sea waters results in the inhibition or enhanced convection variability in the main deep-water formation sites. As regards future sea level, the overall DSL trend estimated by MED16 is comparable to the ensemble mean computed from the Global Circulation Models analysed in Slangen at al. (2014). Nevertheless, our regional dynamical downscaling provides a better characterization of regional patterns arising from ocean dynamics and of their variability, and gives a more pronounced relative sinking of the Mediterranean with respect to the Atlantic. The regional differences in SLR between the presented downscaling and the GCM mean projection are generally constant in time, yet an acceleration emerges in the regional projection for the Balearic area. Indeed, state-of-the-art numerical estimates show that future sea level rise in the Mediterranean basin will crucially depend on the evolution of the sterodynamic component, as determined by the propagation of the Atlantic signal through the Strait of Gibraltar and by the local circulation. MED16 effectively provides an accurate and reliable representation of the basin’s DSL, by accounting for the numerous and complex processes that concur to determine it. Nevertheless, in view of the significant numerical resources it demands, the assessment is recommended of the minimum requirements MED16, or any other similar model, must comply with, in terms of resolution and explicit process representation. In accordance with previous works, we demonstrate that both explicit tidal forcing and an accurate resolution of the Gibraltar Strait are key features that cannot be disregarded in designing numerical tools suitable for the DSL simulation in the Mediterranean Sea. Declarations Author Contributions Conceived the work [GS] Implemented the model and performed the run [GS, AC] Collected data [MP, EN, RI, AC] Designed the analysis [GS, RI, AC, EN, GP, MVS] Performed the analysis [RI, AC, MP, EN] Wrote the paper [ALL] Acknowledgements Regional sea level data from IPCC AR5 distributed in netCDF format by the Integrated Climate Data Center (ICDC, icdc.cen.uni-hamburg.de) University of Hamburg, Hamburg, Germany. SROCC sea-level rise data are available at https://www.ipcc.ch/srocc/download-report/. References Ablain M, Cazenave A, Larnicol G, Balmaseda M, Cipollini P, Faugère Y, Fernandes MJ, Henry O, Johannessen JA, Knudsen P, Andersen O, Legeais J, Meyssignac B, Picot N, Roca M, Rudenko S, Scharffenberg MG, Stammer D, Timms G, and Benveniste J (2015) Improved sea level record over the satellite altimetry era (1993–2010) from the Climate Change Initiative project. Ocean Sci. 11:67–82. https://doi.org/10.5194/os-11-67-2015, 2015. Adcroft A, Campin JM (2004) Rescaled height coordinates for accurate representation of free-surface flows in ocean circulation models. Ocean Model 7(34):269–284 Adloff F, Jordà G, Somot S, Sevault F, Arsouze T, Meyssignac B, Li L, and Planton S (2018) Improving sea level simulation in Mediterranean regional climate models. Clim Dyn, 51:1167–1178 Amitai Y, Lehahn Y, Lazar A, Heifetz E (2010) Surface circulation of the eastern Mediterranean Levantine basin: Insights from analyzing 14 years of satellite altimetry data. J. Geophys. Res. 115, C10058. Antonioli F et al (2017) Sea-level rise and potential drowning of the Italian coastal plains. Flooding risk scenarios for 2100. Quaternary Science Review 158:29–43, doi: 10.1016/j.quascirev.2016.12.021. Arnone RA, Wiesenburg DA, Saunders KD (1990) The origin and characteristics of the algerian current. Journal of Geophysical Research 95 (C2):1587–1598. http://dx.doi.org/10.1029/JC095iC02p01587. Balmaseda MA, Mogensen K, Weaver AT (2013) Evaluation of the ECMWF ocean reanalysis system oras4. QJR Meteorol Soc 139(674):1132–1161. Bignami F, Marullo S, Santoleri R, and Schiano ME (1995) Longwave radiation budget in the Mediterranean Sea. J. Geophys. Res 100:2501–2514. Bonaduce A, Pinardi N, Oddo P, Spada G, and Larnicol G (2016) Sea-level variability in the Mediterranean Sea from altimetry and tide gauges. Climate Dynamics 47(9–10): 2851–2866. Bougeault P, and Lacarrère P (1989) Parameterization of Orography-Induced Turbulence in a Mesobeta-Scale Model. Monthly Weather Review 117:1872-1890. http://dx.doi.org/10.1175/1520-0493(1989)117 Boutov D, Peliz A, Miranda PMA, Soares PMM, Cardoso RM (2014) Inter-annual variability and long term predictability of exchanges through the Strait of Gibraltar. Glob Planet Change 114:23–37. doi:10.1016/j.gloplacha.2013.12.009 Calafat FM, Jordà G (2011) A mediterranean sea level reconstruction (1950–2008) with error budget estimates. Glob. Planet. Change 79:118–133. Campin JM, Adcroft A, Hill C, Marshall J (2004) Conservation of prop- erties in a free-surface model. Ocean Model 6:221–244. Candela J (2001) Mediterranean water and global circulation. In: Siedler G, Church J, Gould J (eds) Ocean circulation and climate. Academic, San Diego, pp 419–429 Church JA, Clark PU, Cazenave A, Gregory J, et al (2013a) Sea level change, in Climate Change 2013: The Physical Science Basis. Stocker TF, Qin D, Plattner GK, et al (ed) Cambridge University Press, Cambridge, UK and New York, NY. USA. Church JA, Clark PU, Cazenave A, Gregory J, et al (2013b) Sea Level Change Supplementary Material. In: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Available from www.climatechange2013.org and www.ipcc.ch. Ciuffardi T, Napolitano E, Iacono R, Reseghetti F, Raiteri G, Bordone A (2016) Analysis of surface circulation structures along a frequently repeated XBT transect crossing the Ligurian and Tyrrhenian seas. Ocean Dyn. 66 (6–7): 767–783. Collins WJ, Bellouin N, Doutriaux-Boucher M, Gedney N, Halloran P, Hinton T, Hughes J, Jones CD et al (2011) Development and evaluation of an Earth-System model – HadGEM2. Geoscient Model Dev 4:1051–1075. https:// doi.org/10.5194/gmd-4-1051-2011 Dangendorf S, Hay C, Calafat FM, Marcos M, Piecuch CG, Berk K, and Jensen J (2019) Persistent acceleration in global sea-level rise since the 1960s. Nature Climate Change 9:705-710 Dee D, Uppala S, Simmons A, Berrisford P, Poli P, Kobayashi S et al (2011) The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Q. J. R. Meteorol. Soc. 137:553–597. doi: 10.1002/qj.828 Egbert GD, Erofeeva SY (2002) Efficient inverse modeling of barotropic ocean tides. Journal of Atmospheric and Oceanic Technology 19:183–204. doi:10.1175/1520-0426(2002)019¡0183:EIMOBO¿2.0.CO;2. Escudier R, Clementi E, Omar M et al (2020) Mediterranean Sea Physical Reanalysis (CMEMS MED-Currents) (Version 1).Copernicus Monitoring Environment Marine Service (CMEMS). https://doi.org/10.25423/CMCC/MEDSEA_MULTIYEAR_PHY_006_004_E3R1 Frederikse T, Landerer F, Caron L, Adhikari S et al (2020) The causes of sea-level rise since 1900. Nature 584:393—397 Fox-Kemper B, Menemenlis D (2008) Can large eddy simulation techniques improve mesoscale rich ocean models? Geophys Monogr:319–337 García-Lafuente J, Alvarez-Fanjul E, Vargas J, Ratsimandresy A (2002) Subinertial variability through the Strait of Gibraltar. J Geophys Res 107(C10):3168. doi:10.1029/2001JC001104 García-Lafuente J, Sammartino S, Huertas IE, Flecha S, Sánchez-Leal RF, Naranjo C, Nadal I and Bellanco MJ (2021) Hotter and Weaker Mediterranean Outflow as a Response to Basin-Wide Alterations. Front. Mar. Sci. 8:613444. doi: 10.3389/fmars.2021.613444 Gaspar P, Gregoris Y, Lefevre JM (1990) A simple eddy kinetic energy model for simulations of the oceanic vertical mixing: Tests at station Papa and long- term upper ocean study site. Journal of Geophysical Research 95 16179. doi:10.1029/jc095ic09p16179. Gill AE, and Niller PP (1973) The theory of the seasonal variability in the ocean. Deep-Sea Res. Oceanogr. Abstr. 20: 141– 177, doi:10.1016/0011-7471(73)90049-1. Giorgi F, Jones C, Asrar GR (2009) Addressing climate information needs at the regional level: the CORDEX framework. WMO Bull 58(3) Gökaşan E, Tur H, Ecevitoğlu B, Görüm T, Türker A, Tok B, Çağlak F, Birkan H, Şimşek M (2005) Evidence and implications of massive erosion along the Strait of İstanbul (Bosphorus). Geo-Mar Lett 25: 324–342 Gökaşan E, Ergin M et al (2007) Factors controlling the morphological evolution of the Çanakkale Strait (Dardanelles, Turkey). Geo-Mar Lett 28:107–129 Greatbatch RJ (1994) A note on the representation of steric sea level in models that conserve volume rather than mass. J. Geophys. Res. 99(C6):12767–12771. Gregory JM, Griffies SM, Hughes CW et al (2019) Concepts and terminology for sea level: mean, variability and change, both local and global. Surv Geophys 40:1251–1289 https://doi.org/10.1007/s10712-019-09525-z Griffies SM, Greatbatch RJ (2012) Physical processes that impact the evolution of global mean sea level in ocean climate models. Ocean Modell. 51:37–72, doi:10.1016/ j.ocemod.2012.04.003 Hamlington BD, Thompson P, Hammond WC, Blewitt G, and Ray RD (2016) Assessing the impact of vertical land motion on twentieth century global mean sea level estimates, J. Geophys. Res. Oceans 121:4980– 4993, doi:10.1002/2016JC011747. Hermans THJ, Tinker J, Palmer MD et al (2020) Improving sea-level projections on the Northwestern European shelf using dynamical downscaling. Clim Dyn 54:1987–2011. https://doi.org/10.1007/s00382-019-05104-5 Holgate SJ, Matthews A, Woodworth PL, Rickards LJ, Tamisiea ME, Bradshaw E, Foden PR, Gordon KM, Jevrejeva S, and Pugh J (2013) New data systems and products at the Permanent Service for Mean Sea Level. Journal of Coastal Research 29(3):493–504 (DOI: 10.2112/JCOASTRES-D-12-00175.1). Iacono R, Napolitano E (2020) Aspects of the summer circulation in the eastern Ligurian Sea, Deep Sea Research Part I: Oceanographic Research Papers, Volume 166, 103407, ISSN 0967-0637, https://doi.org/10.1016/j.dsr.2020.103407 IPCC (2013) Summary for Policymakers. In: Climate Change 2013: The Physical Science Basis. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press: Cambridge and New York, USA and UK, 2014, available online: https://www.ipcc.ch/site/assets/uploads/2018/02/WG1AR5_all_final.pdf. IPCC (2019) Technical Summary [H.-O. Pörtner, D.C. Roberts, V. Masson-Delmotte, P. Zhai, E. Poloczanska, K. Mintenbeck, M. Tignor, A. Alegría, M. Nicolai, A. Okem, J. Petzold, B. Rama, N.M. Weyer (eds.)]. In: IPCC Special Report on the Ocean and Cryosphere in a Changing Climate [H.- O. Pörtner, D.C. Roberts, V. Masson-Delmotte, P. Zhai, M. Tignor, E. Poloczanska, K. Mintenbeck, A. Alegría, M. Nicolai, A. Okem, J. Petzold, B. Rama, N.M. Weyer (eds.)], https://www.ipcc.ch/srocc/ Jeftic L, Milliman JD and Sestini, G (1992) Climatic change and the Mediterranean: environmental and societal impacts of climatic change and sea-level rise in the Mediterranean region. United Nations Environment Programme, http://hdl.handle.net/20.500.11822/29285 Jordà G, Gomis D (2013) On the interpretation of the steric and mass components of sea level variability: The case of the Mediterranean basin. J. Geophys. Res. Oceans 118:953– 963, doi:10.1002/jgrc.20060. Jordà G, Von Schuckmann K, Josey SA, Caniaux G et al (2017a) The Mediterranean Sea heat and mass budgets: Estimates, uncertainties and perspectives. Progress in Oceanography 156:174–208. doi:10.1016/j.pocean.2017.07.001 Jordà G, Sánchez-Román A, Gomis D (2017b) Reconstruction of transports through the Strait of Gibraltar from limited observations. Clim Dyn 48:851–865, https://doi.org/10.1007/s00382-016-3113-8 Kara AB, Wallcraft AJ, Hulbert HE, Stanev EV (2008) Air‐sea fluxes and river discharges in the Black Sea with a focus on the Danube and Bosphorus. J. Mar. Syst. 74:74–95. King MD, Howat IM, Candela SG, Noh MJ et al (2020) Dynamic ice loss from the Greenland Ice Sheet driven by sustained glacier retreat. Commun Earth Environ 1, 1. https://doi.org/10.1038/s43247-020-0001-2 Klein B, Roether W, Manca BB, Bregant D, Beitzel V, Kovacevic V, Luchetta A (1999) The large deep water transient in the Eastern Mediterranean. Deep Sea Research Part I 46: 371-414. Knysh, VV, Demyshev SG, Korotaev GK (2002) Method of reconstructing the climatic seasonal circulation of the Black Sea on the basis of assimilation of hydrologic data in models. Marine Hydrophys. J. 2:36–52, (in Russian). Kubryakov AA, Stanichny SV (2011) Mean Dynamic Topography of the Black Sea, computed from altimetry, drifter measurements and hydrology data. Ocean Sci. 7:745–753. https://doi.org/10.5194/os-7-745-2011 Lambeck K, Smither C, Johnston P (1998) Sea-level change, glacial rebound and mantle viscosity for northern Europe. Geophys. J. Int. 134:102–144. Lambeck K, Yokoyama Y, Purcell A, Johnston P (2002) Reply to the comment by WR Peltier. Quaternary Science Reviews, 21:415-418. Landerer FW, Volkov DL (2013) The anatomy of recent large sea level fluctuations in the Mediterranean Sea. Geophys. Res. Lett. 40:553–557 Large WG, Yeager SG (2004) Diurnal to decadal global forcing for ocean and sea-ice models: the data sets and flux climatologies. National Center for Atmospheric Research Boulder. http://dx.doi.org/10.5065/D6KK98Q6 Leith C (1968) Parameterization of vertical mixing in numerical models of tropical oceans. Phys Fluids 10:1409–1416 Llasses J, Jordà G, Gomis D, Adloff F et al (2018) Heat and salt redistribution within the Mediterranean Sea in the Med‐CORDEX model ensemble. Clim. Dyn. 51:1119-1143. Losch M, Adcroft A, Campin JM (2004) How sensitive are coarse general circulation models to fundamental approximations in the equations of motion? J. Phys. Oceanogr., 34:306–319. Marullo S, Buongiorno Nardelli B, Guarracino M, Santoleri R (2007) Observing the Mediterranean Sea from space: 21 years of Pathfinder-AVHRR sea surface temperatures (1985 to 2005): re-analysis and validation. Ocean Sci. 3: 299-310 Meinshausen M, Smith SJ, Calvin K, Daniel JS, Kainuma MLT, Lamarque JF, Matsumoto K, Montzka SA, Raper SCB, Riahi K, Thomson A, Velders GJM, van Vuuren DPP (2011) The RCP greenhouse gas concentrations and their extensions from 1765 to 2300. Clim Change 109:210–241. https://doi.org/10.1007/s1058 4-011-0156-z Marcos M, Tsimplis MN (2008) Coastal sea level trends in Southern Europe. Geophys. J. Int. 175:70–82. Marshall J, Hill C, Perelman L, Adcroft A (1997a) Hydrostatic, quasi- hydrostatic, and nonhydrostatic ocean modeling. J Geophys Res 102:5733–5752 Marshall J, Adcroft A, Hill C, Perelman L, Heisey C (1997b) A finite- volume, incompressible Navier Stokes model for studies of the ocean on parallel computers. J Geophys Res 102:5753–5766 Meyssignac B, Calafat FM, Somot S, Rupolo V, Stocchi P, Llovel W, Cazenave A (2011) Two-dimensional reconstruction of the mediterranean sea level over 1970–2006 from tide gage data and regional ocean circulation model outputs. Glob. Planet. Change 77(1–2): 49–61. Mohamed B, Mohamed A et al (2019) Inter-Annual Variability and Trends of Sea Level and Sea Surface Temperature in the Mediterranean Sea over the Last 25 Years. Pure Appl. Geophys. 176:3787–3810. Naranjo C, Garcia-Lafuente J, Sannino G, Sanchez-Garrido JC (2014) How much do tides affect the circulation of the Mediterranean Sea? From local processes in the Strait of Gibraltar to basin-scale effects, Progress in Oceanography 127:108-116 /doi.org/10.1016/j.pocean.2014.06.005. Palma M, Iacono R, Sannino G, Bargagli A, Carillo A, Fekete BM, Lombardi E, Napolitano E, Pisacane G, Struglia MV (2020) Short-term, linear and non-linear local effects of the tides on the surface dynamics in a new, high-resolution model of the Mediterranean Sea circulation. Ocean Dyn. 70:935-963 https://doi.org/10.1007/s10236-020-01364-6. Peltier WR (2002) Comments on the paper of Yokoyama et al. (2000) entitled ‘‘Timing of the Last Glacial Maximum from observed sea level minima’’. Quaternary Science Reviews 21:409-414 Peltier WR (2004) Global glacial isostasy and the surface of the ice-age earth: The ICE-5G (VM2) model and GRACE. Annu. Rev. Earth Planet. Sci. 32:111–149 doi:10.1146/annurev.earth.32.082503.144359. Peneva E, Stanev E, Belokopytov V, Le Traon PY (2001) Water transport in the Bosphorus Strait estimated from hydro‐meteorological and altimeter data: Seasonal and decadal variability. J. Mar. Syst. 31:21–33. Pinardi N, Zavatarelli M, Adani M, Coppini G, Fratianni C, Oddo P, Simoncelli S, Tonani M, Lyubartsev V, Dobricic S, Bonaduce A (2015) Mediterranean Sea large-scale low-frequency ocean variability and water mass formation rates from 1987 to 2007: a retrospective analysis. Prog Oceanogr., 132: 318-332. https://doi.org/10.1016/j.pocean.2013.11.003. Ponce VM (1994) Engineering hydrology: principles and practices. Prentice Hall Poulain PM, Mauri E, Gerin R, Chiggiato J et al (2020) On the dynamics in the southeastern Ligurian Sea in summer 2010. Cont. Shelf Res. 196:104083 Rahmstorf S (2007) A semi-empirical approach to projecting future sea-level rise, Science 315(8010): 368-370, doi: 10.1126/science.1135456. Rio MH, Pascual A, Poulain PM, Menna M, Barcelo’ B, Tintoré J (2014) Computation of a new mean dynamic topography for the Mediterranean Sea from model outputs, altimeter measurements and oceanographic in situ data. Ocean Sci. 10:731-744. Roether W, Manca BB, Klein B et al (1996) Recent changes in eastern Mediterranean deep waters. Science 271: 333–335. doi:10.1126/science.271.5247.333. Sánchez Román A, Sannino G, García Lafuente J et al (2009) Transport estimates at the western section of the Strait of Gibraltar: A combined experimental and numerical modeling study. Journal of Geophysical Research: Oceans 114, C06002. https://doi.org/10.1029/2008JC005023. Sannino G, Sözer A, Özsoy E (2017) A high-resolution modelling study of the Turkish Straits System. Ocean Dynamics 67(3):397–432 Sannino G, Carillo A, Pisacane G, Naranjo C (2015) On the relevance of tidal forcing in modelling the Mediterranean thermohaline circulation. Prog Oceanogr 134:304–329 Sannino G, Herrmann M, Carillo A, Rupolo V, Ruggiero V, Artale V, Heimbach P (2009) An eddy-permitting model of the Mediterranean sea with a two-way grid refinement at the Strait of Gibraltar. Ocean Modelling 30 (1): 56–72 Sannino G, Carillo A, Pratt L (2009) Hydraulic criticality of the exchange flow through the strait of Gibraltar. Journal of Physical Oceanography 39:2779–2799. doi:10.1175/2009JPO4075.1 Santamaría-Gómez, A, Gravelle M, Dangendorf S, Marcos M, Spada G, Wöppelmann G (2017) Uncertainty of the 20th century sea-level rise due to vertical land motion errors. Earth and Planetary Science Letters 473:24-32. https://doi.org/10.1016/j.epsl.2017.05.038. Sanz JL, Acosta J, Esteras M, Herranz P, Palomo C, Sandoval N (1991) Prospección geofísica del Estrecho de Gibraltar: resultados del Programa Hércules (1980–1983). Publ. espec. Inst. Esp. Oceanogr. Span. Inst. of Oceanogr., Madrid Sen Gupta A, Jourdain NC, Brown JN, Monselesan D (2013) Climate drift in the CMIP5 models. J. Climate 26:8597– 8615. doi:10.1175/JCLI-D-12-00521.1. Shugar DH, Burr A, Haritashya UK et al (2020) Rapid worldwide growth of glacial lakes since 1990. Nature Climate Change 10:939-945. https://doi.org/10.1038/s41558-020-0855-4 Simonov AI, Altman EN (1991) Hydrometeorology and hydrochemistry of the USSR seas, in Sensitivity to Change: Black Sea, Baltic Sea and North Sea. NATO/ASI Ser., vol. IV, edited by E. Özsoy and A. Mikaelyan, pp. 11–25, Kluwer Acad., Dordrecht, Netherlands. Slangen ABA, Adloff F, Jevrejeva S et al (2017a) Review of Recent Updates of Sea-Level Projections at Global and Regional Scales. Surv. Geophys., 38:385–406. Slangen ABA, Carson M, Katsman CA et al (2014) Projecting twenty-first century regional sea-level changes. Climatic Change 124:317–332. https://doi.org/10.1007/s10584-014-1080-9 Smagorinsky J (1963) General circulation experiments with the primitive equations. Mon Weather Rev 91:99–164 Soto-Navarro J, Criado‐Aldeanueva F, Lafuente JG, Sanchez-Roman A (2010) Estimation of the Atlantic inflow through the Strait of Gibraltar from climatological and in situ data. J. Geophys. Res. 115. https://doi.org/10.1029/2010JC006302 Soto-Navarro J, Jordá, G, Amores A et al (2020) Evolution of Mediterranean Sea water properties under climate change scenarios in the Med-CORDEX ensemble. Clim Dyn 54:2135–2165. https://doi.org/10.1007/s00382-019-05105-4 Spada G, Stocchi P (2006) The Sea Level Equation, Theory and Numerical Examples, Aracne, Roma, p. 96, ISBN: 88–548–0384–7. Spada G, Stocchi P (2007) SELEN: a Fortran 90 program for solving the ‘Sea Level Equation’. Comput. Geosci. 33(4): 538–562. doi:10.1016/j.cageo.2006.08.006. Spada G (2017) Glacial Isostatic Adjustment and Contemporary Sea Level Rise: An Overview. Surv Geophys 38:153–185. https://doi.org/10.1007/s10712-016-9379-x Strandberg G, Bärring L, Hansson U, Jansson C, Jones C (2014) CORDEX scenarios for Europe from the Rossby Centre regional climate model RCA4. Theocharis A, Klein B, Nittis K, Roether W (2002) Evolution and status of the Eastern Mediterranean Transient (1997–1999). J. Mar. Syst. 33-34: 91-116. Thiéblemont R, Le Cozannet G, Toimil A, Meyssignac B, Losada IJ (2019) Likely and High-End Impacts of Regional Sea-Level Rise on the Shoreline Change of European Sandy Coasts Under a High Greenhouse Gas Emissions Scenario. Water 2019, 11, 2607. Tsimplis MN, Josey SA (2001) Forcing of the Mediterranean Sea by atmospheric oscillations over the North Atlantic. Geophysical Research Letters 28:803-806. https://doi.org/10.1029/2000GL012098 Tsimplis MN, Bryden HL (2000) Estimation of the transports through the strait of Gibraltar. Deep-Sea Res I 47:2219–2242 Ünlüata Ü, Oğuz T, Latif MA, Özsoy E (1990) On the physical oceanography of the Turkish Straits, in The Physical Oceanography of Sea Straits. NATO/ASI Ser., edited by L. J. Pratt, pp. 25–60, Kluwer Acad., Dordrecht, Netherlands. Vörösmarty CJ, Federer CA, Schloss AL (1998) Potential evaporation functions compared on US watersheds: implications for global-scale water balance and terrestrial ecosystem modeling. J Hydrol 207: 147–169 Vousdoukas MI, Mentaschi L, Voukouvalas E, Verlaan M, Feyen L (2017) Extreme sea levels on the rise along Europe’s coasts. Earth’s Future 5:304-323. https://doi.org/10.1002/2016EF000505 Wisser D, Frolking S, Douglas EM, Fekete BM, Vörösmarty CJ, Schumann AH (2008) Global irrigation water demand: variability and uncertainties arising from agricultural and climate data sets. Geophys Res Lett 35:L24408 Wisser D, Fekete BM, Vörösmarty CJ, Schumann AH (2010) Reconstructing 20th century global hydrography: a contribution to the Global Terrestrial Network- Hydrology (GTN-H). Hydrol Earth Syst Sci 14:1–24 Cite Share Download PDF Status: Published Journal Publication published 21 Jan, 2022 Read the published version in Climate Dynamics → Version 1 posted Editorial decision: Major Revision 27 Sep, 2021 Reviews received at journal 16 Aug, 2021 Reviewers invited by journal 27 Jun, 2021 Editor assigned by journal 23 Jun, 2021 First submitted to journal 23 Jun, 2021 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-653703","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":36045010,"identity":"e2d92332-b1df-4b52-9b3d-2a1f911c7073","order_by":0,"name":"Gianmaria Sannino","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABB0lEQVRIiWNgGAWjYDACZgjFA+VKMDCwN6CIEKOF5wBEBLceDCCRgGoKOpB3Z3726EbNPRn+BvaHnwt3WOTxSz4+9uBDDYOMPQ4thofZzI1zjhXzSBzgMZaeeUaiWHJ2WrrhjGO4HWbYzGAmncOWwMNwgIdBmrdNInHD7RwzaR42fFrYv0nn/EvgkT/A/vg3WMvN89+k//zDrUWemcdMOrctgcfgANA6sJYbPGzSjG24tRgw85RJ5/Yl8Bge5jGzBmmZ2ZNmJtnbJ8EDDWxMW/qPb5PO+ZZgL3e8/fFt3ra6xH72w88kfnyzsYdGKaYtcKOYUSUkcDgLaAsOo0bBKBgFo2AUIAAAJEpK+vldhZIAAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0002-3985-9432","institution":"ENEA Casaccia Research Centre: ENEA Centro Ricerche Casaccia","correspondingAuthor":true,"submittingAuthor":false,"prefix":"","firstName":"Gianmaria","middleName":"","lastName":"Sannino","suffix":""},{"id":36045011,"identity":"1174a54e-85e7-4df9-9845-5e8de373ca93","order_by":1,"name":"Adriana Carillo","email":"","orcid":"","institution":"ENEA Casaccia Research Centre: ENEA Centro Ricerche Casaccia","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Adriana","middleName":"","lastName":"Carillo","suffix":""},{"id":36045012,"identity":"b6460f56-6488-4650-a808-ce2de6574eea","order_by":2,"name":"Roberto Iacono","email":"","orcid":"","institution":"ENEA Casaccia Research Centre: ENEA Centro Ricerche Casaccia","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Roberto","middleName":"","lastName":"Iacono","suffix":""},{"id":36045013,"identity":"0c97eebf-a88f-4593-8813-d22d8087dac2","order_by":3,"name":"Ernesto Napolitano","email":"","orcid":"","institution":"ENEA Casaccia Research Centre: ENEA Centro Ricerche Casaccia","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Ernesto","middleName":"","lastName":"Napolitano","suffix":""},{"id":36045014,"identity":"0daa73b0-7f90-411f-87ae-394660c9f5e5","order_by":4,"name":"Massimiliano Palma","email":"","orcid":"","institution":"ENEA Casaccia Research Centre: ENEA Centro Ricerche Casaccia","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Massimiliano","middleName":"","lastName":"Palma","suffix":""},{"id":36045015,"identity":"573b9b09-7170-4899-9c40-b404cf399a1d","order_by":5,"name":"Giovanna Pisacane","email":"","orcid":"","institution":"ENEA Casaccia Research Centre: ENEA Centro Ricerche Casaccia","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"Giovanna","middleName":"","lastName":"Pisacane","suffix":""},{"id":36045016,"identity":"ad64ed88-a694-4f09-8276-7aaed8ee95ee","order_by":6,"name":"MariaVittoria Struglia","email":"","orcid":"","institution":"ENEA Casaccia Research Centre: ENEA Centro Ricerche Casaccia","correspondingAuthor":false,"submittingAuthor":false,"prefix":"","firstName":"MariaVittoria","middleName":"","lastName":"Struglia","suffix":""}],"badges":[],"createdAt":"2021-06-23 18:28:21","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-653703/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-653703/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s00382-021-06132-w","type":"published","date":"2022-01-21T06:17:26+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":11024035,"identity":"73e36c91-3281-4f97-add6-025d8d25353e","added_by":"auto","created_at":"2021-07-01 21:57:39","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":605415,"visible":true,"origin":"","legend":"Model bathymetry: entire domain and high-resolution representation of the three straits (Gibraltar, Dardanelles, and Bosphorus) in which the spatial detail of the computational grid has been increased","description":"","filename":"Figure01.png","url":"https://assets-eu.researchsquare.com/files/rs-653703/v1/7b54c867e273059979c60c31.png"},{"id":11025093,"identity":"a2dcb00d-80ee-4fa7-a9c1-da5d080f7d1b","added_by":"auto","created_at":"2021-07-01 22:06:39","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":423301,"visible":true,"origin":"","legend":"Map of the SST climatology for the OISST (a,b), COPERNICUS reanalysis (c,d), hindcast simulation (e,f), and historical simulation (g,h) over the common period 1987-2005. Winter (summer) climatologies are on the left (right) panels ","description":"","filename":"Figure02.png","url":"https://assets-eu.researchsquare.com/files/rs-653703/v1/dc14791ae8a32256bdfca21a.png"},{"id":11024098,"identity":"26f5d027-4e7a-4aac-a39a-cb0c57ac8cc9","added_by":"auto","created_at":"2021-07-01 22:00:39","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":298504,"visible":true,"origin":"","legend":"Map of the SSS climatology for COPERNICUS reanalysis (a,b), hindcast simulation (c,d), and historical simulation (e,f) over the common period 1987-2005. Winter (summer) climatologies are on the left (right) panels","description":"","filename":"Figure03.png","url":"https://assets-eu.researchsquare.com/files/rs-653703/v1/43b5bcb4967b37f4bd42136c.png"},{"id":11024097,"identity":"65855ff0-ceea-4834-b788-5a939bd6140b","added_by":"auto","created_at":"2021-07-01 22:00:39","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":61766,"visible":true,"origin":"","legend":"Reanalysis (blue) and hindcast (red) time series of temperature anomalies (°C ; annual values) for the upper (0-150 m) and intermediate (150-600 m) layers, for the Mediterranean Sea, and the western and eastern sub-basins ","description":"","filename":"Figure04.png","url":"https://assets-eu.researchsquare.com/files/rs-653703/v1/35c3fc22299cccdc02aa7b5a.png"},{"id":11024100,"identity":"cad0a1b7-d22f-4464-bb76-821f719f19d6","added_by":"auto","created_at":"2021-07-01 22:00:39","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":58943,"visible":true,"origin":"","legend":"Reanalysis (blue) and hindcast simulation (red) time series of salinity anomalies (psu); annual values) for the upper (0-150 m) and intermediate (150-600 m) layers, for the Mediterranean Sea, and the western and eastern sub-basins ","description":"","filename":"Figure05.png","url":"https://assets-eu.researchsquare.com/files/rs-653703/v1/50b9e93d4c73653ff9c3b590.png"},{"id":11024175,"identity":"8f3be4f0-4eee-4e39-996f-bc7c428d0924","added_by":"auto","created_at":"2021-07-01 22:03:39","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":338294,"visible":true,"origin":"","legend":"Surface (15 m of depth) and intermediate (300 m of depth) circulation, averaged over the simulation periods of the hindcast (left panel) and of the historical (right panel) experiments","description":"","filename":"Figure06.png","url":"https://assets-eu.researchsquare.com/files/rs-653703/v1/56f89f2eabee39e98948015b.png"},{"id":11024037,"identity":"e286aeb9-13dd-4185-b9d9-5774877d3a56","added_by":"auto","created_at":"2021-07-01 21:57:39","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":332795,"visible":true,"origin":"","legend":"Hindcast surface (15 m of depth) circulation averaged over the simulation time-span (1982-2010) for three sub-basins: Liguro-Provençal (a), Adriatic (b), and Marmara Sea ©","description":"","filename":"Figure07.png","url":"https://assets-eu.researchsquare.com/files/rs-653703/v1/407d8d0832680c30fb304bce.png"},{"id":11024179,"identity":"9b1fb336-b052-454d-bab7-e15adbe6fbf1","added_by":"auto","created_at":"2021-07-01 22:03:40","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":84722,"visible":true,"origin":"","legend":"Time evolution of the monthly maximum MLD computed over the Gulf of Lion area (box from 3.34°E/39.97°N to 7.28°E/43.15°N) and the Southern Adriatic Sea (box from 16.47°E/40.28°N to 19.28°E/42.72°N). Left (right) panels are for the hindcast (historical) simulation","description":"","filename":"Figure08.png","url":"https://assets-eu.researchsquare.com/files/rs-653703/v1/b9f24c28e786f6e40a2f6b34.png"},{"id":11024099,"identity":"efff1207-1e7a-4810-a377-e521f8dfc77d","added_by":"auto","created_at":"2021-07-01 22:00:39","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":33683,"visible":true,"origin":"","legend":"Sea-level seasonal cycle; whole Mediterranean (panel a), western and eastern sub-basins (panels b-c). Black dots denote values computed from the hindcast simulation, and diamonds those from the observations ","description":"","filename":"Figure09.png","url":"https://assets-eu.researchsquare.com/files/rs-653703/v1/429cc21669fc2e66a8a2290b.png"},{"id":11024047,"identity":"a63b7ed5-267f-4200-84d0-9f617e9cb7cf","added_by":"auto","created_at":"2021-07-01 21:57:40","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":47117,"visible":true,"origin":"","legend":"Interannual variability of the sea-level anomaly in different basins. Symbols are as in Figure 9","description":"","filename":"Figure10.png","url":"https://assets-eu.researchsquare.com/files/rs-653703/v1/a667579daa90f0ec0b844194.png"},{"id":11024051,"identity":"18bb560f-59cc-4b54-b51a-8da6ee4bf2ae","added_by":"auto","created_at":"2021-07-01 21:57:40","extension":"png","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":237318,"visible":true,"origin":"","legend":"Averages of the SSH (1993-2010), for the hindcast (middle panel) and historical (lower panel) simulations, are compared with the average of the ADT for the same period (upper panel), which is very close to the mean dynamic topography. The historical simulation has been extended with the RCP8.5 scenario for 2006–2010. In all cases, the respective regional climatological means have been subtracted to ease the comparison of patterns","description":"","filename":"Figure11.png","url":"https://assets-eu.researchsquare.com/files/rs-653703/v1/0fdcc8ab2dc58305b941456f.png"},{"id":11024046,"identity":"795cc3d4-d780-4c4e-8c10-ecad587b95d7","added_by":"auto","created_at":"2021-07-01 21:57:40","extension":"png","order_by":12,"title":"Figure 12","display":"","copyAsset":false,"role":"figure","size":49241,"visible":true,"origin":"","legend":"Location of the tide gauge stations whose data are used to construct Table 4. Red circles indicate stations used to extract the time series displayed in Figure 13","description":"","filename":"Figure12.png","url":"https://assets-eu.researchsquare.com/files/rs-653703/v1/94a8ecf4d229a98ea7c79e9b.png"},{"id":11024050,"identity":"242445a9-a3ec-468c-b508-3efe6b8d8b26","added_by":"auto","created_at":"2021-07-01 21:57:40","extension":"png","order_by":13,"title":"Figure 13","display":"","copyAsset":false,"role":"figure","size":243064,"visible":true,"origin":"","legend":"Comparison between sea-surface height time series extracted from the model (red) with tide gauge time series, at 6 locations along the coasts of the computational domain, marked by red circles in the map of Figure 12","description":"","filename":"Figure13.png","url":"https://assets-eu.researchsquare.com/files/rs-653703/v1/ae6888328c81936cf46ad109.png"},{"id":11024106,"identity":"424e9e70-77f6-422e-880c-653e4f7bb799","added_by":"auto","created_at":"2021-07-01 22:00:39","extension":"png","order_by":14,"title":"Figure 14","display":"","copyAsset":false,"role":"figure","size":154590,"visible":true,"origin":"","legend":"SST difference between averages over the scenario and the historical simulations for winter (left panels) and summer (right panels). Upper panels show averages over the period 2046-2065, lower panels over the period 2081-2100 (right panel). Historical simulation has been averaged over the entire period 1981-2005","description":"","filename":"Figure14.png","url":"https://assets-eu.researchsquare.com/files/rs-653703/v1/e2a7b563f5dbe5af333c5a66.png"},{"id":11024103,"identity":"cfaf1c7f-2eec-4243-b4bd-9f6045707c8a","added_by":"auto","created_at":"2021-07-01 22:00:39","extension":"png","order_by":15,"title":"Figure 15","display":"","copyAsset":false,"role":"figure","size":141654,"visible":true,"origin":"","legend":"SSS difference between averages over the scenario and the historical simulations for winter (left panels) and summer (right panels). Upper panels show averages over the period 2046-2065, lower panels over the period 2081-2100 (right panel). Historical simulation has been averaged over the entire period 1981-2005","description":"","filename":"Figure15.png","url":"https://assets-eu.researchsquare.com/files/rs-653703/v1/b952120202877368fb1ffe03.png"},{"id":11025091,"identity":"c4bbba5d-c106-422b-bd2d-0b5d96c2e3cf","added_by":"auto","created_at":"2021-07-01 22:06:39","extension":"png","order_by":16,"title":"Figure 16","display":"","copyAsset":false,"role":"figure","size":50799,"visible":true,"origin":"","legend":"Time evolution of the monthly maximum MLD computed over the Gulf of Lion area and the Southern Adriatic Sea. Scenario simulation","description":"","filename":"Figure16.png","url":"https://assets-eu.researchsquare.com/files/rs-653703/v1/49893444a8d79aa2d57680e6.png"},{"id":11024052,"identity":"f06c58f6-0fa3-4b0b-bf31-fee27e0d3444","added_by":"auto","created_at":"2021-07-01 21:57:40","extension":"png","order_by":17,"title":"Figure 17","display":"","copyAsset":false,"role":"figure","size":59968,"visible":true,"origin":"","legend":"Time evolution of the components contributing to the projected mean sea level in the Mediterranean under the RCP8.5. Solid lines represent the central estimate over available models","description":"","filename":"Figure17.png","url":"https://assets-eu.researchsquare.com/files/rs-653703/v1/f713652b4432d347912a332b.png"},{"id":11024053,"identity":"da5a3a69-2eb8-4de7-a0b9-471b3ee4cac3","added_by":"auto","created_at":"2021-07-01 21:57:40","extension":"png","order_by":18,"title":"Figure 18","display":"","copyAsset":false,"role":"figure","size":138307,"visible":true,"origin":"","legend":"Evolution of salinity in a section of the Strait of Gibraltar at intervals of 1 h","description":"","filename":"Figure18.png","url":"https://assets-eu.researchsquare.com/files/rs-653703/v1/096b59f0cf01702818d1d3ee.png"},{"id":11024108,"identity":"82d9a323-514a-44cd-bd7b-82eedb4106e1","added_by":"auto","created_at":"2021-07-01 22:00:40","extension":"png","order_by":19,"title":"Figure 19","display":"","copyAsset":false,"role":"figure","size":86164,"visible":true,"origin":"","legend":"Differences between values averaged over the Mediterranean basin and over the Atlantic box for the historical (orange) and scenario (red) simulations, (in black superimposed AVISO ADT) (left panel). In the right panel, the same difference for the GLOBAL dynamic sea surface height. Annual values ","description":"","filename":"Figure19.png","url":"https://assets-eu.researchsquare.com/files/rs-653703/v1/dc91ad84ab1ca59fb9bc752b.png"},{"id":11025160,"identity":"cb4477ab-7135-4fca-820d-06486d1f1de9","added_by":"auto","created_at":"2021-07-01 22:09:39","extension":"png","order_by":20,"title":"Figure 20","display":"","copyAsset":false,"role":"figure","size":282199,"visible":true,"origin":"","legend":"Difference in the MED16 sterodynamic SLC computed over the period 2046-2065 (upper panel) and over the period 2080-2099 (lower panel) relative to the historical period","description":"","filename":"Figure20.png","url":"https://assets-eu.researchsquare.com/files/rs-653703/v1/39b4ed42a5da3e8b4bc84d29.png"},{"id":11024048,"identity":"c4bc2c7f-bb1f-4a3d-ba9d-505165288469","added_by":"auto","created_at":"2021-07-01 21:57:40","extension":"png","order_by":21,"title":"Figure 21","display":"","copyAsset":false,"role":"figure","size":227325,"visible":true,"origin":"","legend":"Difference in the sterodynamic SLC computed by the MED16 simulation and the ensemble mean from Slangen at al. (2014). Mean over the period 2046-2065 (upper panel), and over the period 2080-2099 (lower panel)","description":"","filename":"Figure21.png","url":"https://assets-eu.researchsquare.com/files/rs-653703/v1/9d09208c17588a2708923a94.png"},{"id":17522094,"identity":"1c9330eb-ecf6-40bc-930f-8318213b4779","added_by":"auto","created_at":"2022-01-21 06:17:36","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3377585,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-653703/v1/50f70dea-8737-4fba-8203-4834eaee0c51.pdf"}],"financialInterests":"","formattedTitle":"\u003cp\u003eModelling Present and Future Climate in the Mediterranean Sea: A Focus on Sea-Level Change\u003c/p\u003e","fulltext":[{"header":"1. Introduction","content":" \u003cp\u003eFuture sea level rise (SLR) constitutes a threat for the coastal environment and economies, which is liable to be further exacerbated by the superposition of waves, atmospheric surge, and tides. Climate scientists are therefore becoming more and more committed to improve projections of SLR as to both resolution and accuracy, and to reduce current uncertainties, at the same time increasing our understanding of such a challenging scientific problem and enabling more accurate risk assessments. The latter, in particular, suffer from the difficulty in accounting for the cascading effects from sea-level rise to actual coastal impacts, across a range of spatial and temporal scales that demand complex and differentiated approaches (Thi\u0026eacute;blemont et al., \u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Likewise, the lack of comprehensive projections of extreme sea levels (ESL) that include mean sea level (MSL), tides, waves, and storm surges is lamented by Vousdoukas et al. (\u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), who present a first attempt to provide an impact-oriented, regional-scale evaluation of ESL for the European shoreline. Their effort, albeit still inadequate for the design of specific measures, can nevertheless help highlight vulnerabilities and emerging research issues. As a matter of fact, the authors denounce the limitations inherent to their approach, mainly arising from the inadequacy of the regional atmospheric projections available at the time, as well as those of the projections derived from coupled global OGCMs. In addition, they highlight that explicitly solving the nonlinear interactions between waves, storm surge, and tides should be a key element of any future effort to reliably model ESL, rather than continuing to resort to linear combinations of the ESL components resulting from independent simulations.\u003c/p\u003e \u003cp\u003eBesides being highly vulnerable to SLR due to the potentially disruptive impacts on its coastal economies (Jeftic et al., \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e1992\u003c/span\u003e), the Mediterranean basin is especially challenging from the scientific perspective, due to the inherent diversity of its geological history and the peculiar and complex features of its marine circulation and environment. In addition to the local dynamics and geological processes, interacting over a broad spectrum of scales, sea level in the basin is also constrained by the water mass exchange across the Strait of Gibraltar, which, in fact, regulates the hydraulic jump between the Mediterranean and the Atlantic Ocean and influences how sea level rise in the Atlantic is transmitted into the Mediterranean. In its turn, the connection with the Black Sea, through the Dardanelles, Sea of Marmara and the Bosphorus, couples the basin hydrology to the land-based hydrological cycle of a vast portion of continental Europe. Nevertheless, despite the recognized inability of coarse resolution GCMs to accurately represent the highly non-linear, small-scale processes in marginal seas such as the Mediterranean (Slangen et al., \u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e2017a\u003c/span\u003e; Marcos and Tsimplis, \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2008\u003c/span\u003e), the global sterodynamic (following the terminology proposed by Gregory et al., \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) sea-level projections for the Atlantic area near Gibraltar, are still often used to estimate the basin's internal sea level (Thi\u0026eacute;blemont et al., \u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAt the global scale, the question is still open whether the strong trends observed in the last 25 years through satellite altimetry are part of longer-term tendencies, or reflect more recent changes in the atmosphere-ocean coupled system. Recent work by Dangendorf et al. (\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) provides evidence that the global trend accelerated at the end of the 1960s, and indicates that the resulting acceleration in the global sea level rise is linked to modifications of the southern hemispheric westerlies, leading to warming and circulation changes in the southern world ocean. Other factors have also been shown to play a key role, such as the change of terrestrial water storage, and the melting of ice sheets and glaciers, which has significantly increased in the last decades (see, e.g., Shugar et al., \u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e2020\u003c/span\u003e, which focuses on the growth of glacial lakes, and King et al., \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2020\u003c/span\u003e, where the mass loss from the Greenland Ice Sheet is analyzed). The work by Frederikse et al. (\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) provides a thorough review of the various contributions and their relative weight.\u003c/p\u003e \u003cp\u003eAs to future scenarios, since 1995 the coordinated worldwide CMIP effort (Coupled Model Intercomparison Project) has provided constantly updated projections of future sea level rise using earth-system models, to which the contributions from continental glaciers and geologic processes are added offline. The Phase 5 models (CMIP5) used for the fifth Intergovernmental Panel on Climate Change (IPCC) assessment report (AR5, IPCC \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2013\u003c/span\u003e) project a global sea level rise between 26 and 97 cm (overall likely range) at the end of this century, depending on the climate scenario, with pretty large inter-model spread. Such estimates were revised in the 2019 Special Report on the Ocean and Cryosphere in a Changing Climate (SROCC), which assigns medium confidence to a likely range between 29 cm and 110 cm (IPCC, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). For the area of the North Atlantic in proximity of the Strait of Gibraltar, projections can range from about 40 cm to over 100 cm for the most pessimistic RCP8.5 scenario, with an ensemble mean of about 80 cm, and approximately from 30 cm to 80 cm for the intermediate RCP4.5 scenario, with an ensemble mean of 60 cm (Vousdoukas et al., \u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). On the other hand, Slangen et al., \u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e2014\u003c/span\u003e presented regional sea-level projections and their associated uncertainties up to the end of the 21st century, by combining model- and observation-based regional contributions of land ice, groundwater depletion and glacial isostatic adjustment, and CMIP5 projections of the changing ocean circulation, increased heat uptake and atmospheric pressure patterns, also accounting for gravitational effects due to mass redistribution, ending up with an estimate of about 70 cm for the Mediterranean Basin. However, while concluding that regional variations in sea-level can significantly differ from the global mean (up to 30% above and 50% below) and that the land ice contribution dominates the overall uncertainty, they also stressed the need for dedicated regional downscaling of each global climate scenario.\u003c/p\u003e \u003cp\u003eIn general, the inherent uncertainty in SLR projections has induced researchers to adopt a variety of alternative estimates in their impact assessments. Thi\u0026eacute;blemont et al. (\u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) analyze the effects of a median estimate of about 80 cm and two different high-end scenarios, resulting from two extreme estimates for the sea-level equivalent of melting glaciers (i.e the the upper limit of the likely range and the \u0026ldquo;worst model\u0026rdquo; projections). Antonioli et al. (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), while reviewing possible alternative ranges for the projected high-end SLR at 2100, use the 530\u0026ndash;970 mm interval reported in the IPCC AR5, and Rahmstorf\u0026rsquo;s semi-empirical estimate of about 1.400 mm (Rahmstorf, \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). Improved projections from the next generation of global models are expected by 2022, when the next IPCC assessment is to be delivered. However, as remarked above, the spatial resolution of current global models is not sufficient to provide realistic estimates of local sea level rise in areas, such as the Mediterranean, where crucial processes are far from being explicitly resolved, and hardly allow a reliable parameterization (Sannino et al., \u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e2009\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAdloff et al. (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) recently discussed the complexity inherent to projecting sea level in the Mediterranean, and analyzed the performances of four different regional hindcast simulations of the basin circulation (period 1980\u0026ndash;2012), driven by realistic atmospheric forcing. The incorporation of improved sea level information at the Atlantic lateral boundary was found to significantly enhance the reliability of results, confirming that the correct representation of the interactions between the two basins is an important requirement for a successful numerical simulation. However, an adequate treatment of the complex, hydraulically driven dynamics across the Strait of Gibraltar was still missing, as well as a more refined treatment of the exchange between the Mediterranean Sea and the Black Sea through the Dardanelles Strait, at the model eastern boundary. In addition, model performances were mostly evaluated by computing basin averages, while altimeter data series clearly indicate that sea level rise has not been homogeneous in the basin over the last decades. By analyzing altimeter data over a 25 years period (1993\u0026ndash;2017), Mohamed et al. (\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) showed that the observed increase in the average sea level anomaly (SLA) has been quite different in different sub-basins, ranging from a minimum of 1.95 mm/year, in the Ionian Sea, to a maximum of 3.73 mm/year, in the Aegean Sea. At a broader scale, the SLA positive trend appears to be significantly stronger in the Levantine basin than in the Western Mediterranean Sea, as well as characterized by distinctive features, with a fairly regular linear increase in the Western Mediterranean Sea, possibly attributable to the Atlantic contribution, and a more complex behavior in the Levantine basin. The latter, in particular, is liable to be influenced by the evolution of the Eastern Mediterranean Transient (EMT), the dramatic event occurred in the eastern Mediterranean at the beginning of the 1990s (Roether et al., \u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e1996\u003c/span\u003e; Klein et al., \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e1999\u003c/span\u003e, Theocharis et al., \u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). The difficulty in separating climate-change-induced variations from the local circulation variability is therefore evident, as well as the role played by small-scale features, either attributable to internal variability or determined by atmospheric forcing, in generating long-lasting differences across the Mediterranean sub-basins, amplifying or mitigating the effects of global sea level rise.\u003c/p\u003e \u003cp\u003ePrompted by these considerations, we developed a regional ocean model for the long-term simulation of the Mediterranean Sea circulation (hereinafter MED16) which we used to obtain accurate projections of the Mediterranean sea level. The model represents the climate version of the high-resolution, tide-including ocean model described in Palma et al. (\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). The two models share the same computational domain, which includes the Black Sea, thus allowing to consistently compute water exchanges at the Dardanelles Strait and to avoid any ad hoc assumption at the Mediterranean eastern boundary. The climate version necessarily uses a coarser horizontal grid with respect to its operational counterpart, yet resolution is significantly increased in critical regions such as the Gibraltar, Dardanelles, and Bosphorus Straits. The explicit, high-resolution representation of inter-basin water exchanges at the boundaries constitutes a unique, distinguishing feature of the present implementation.\u003c/p\u003e \u003cp\u003eIn the following, we analyze the hindcast of the Mediterranean circulation forced by the SMHI-RCA4 regional downscaling of the ECMWF ERA-Interim reanalysis data (Dee et al. \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2011\u003c/span\u003e), and the historical and RCP8.5 scenario forced by the SMHI-RCA4 regional downscaling of the HadGEM2-ES global projection (Collins et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). The overall basin-scale model performance is assessed through comparison with observations and reanalysis data, with special focus on the model ability to reliably represent the local sea level height. In particular, comparison with data from tide gauges allows us to evaluate the model performance in coastal regions, where the impacts of sea level rise really affect human communities and economies, and need to be specifically assessed.\u003c/p\u003e \u003cp\u003eThe paper is organized as follows. The main characteristics of the model and of its implementation are described in Sect.\u0026nbsp;2, whereas Sect.\u0026nbsp;\u003cspan refid=\"Sec5\" class=\"InternalRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan refid=\"Sec11\" class=\"InternalRef\"\u003e5\u003c/span\u003e are devoted to the detailed analysis of the simulations performed. In particular, Sect.\u0026nbsp;\u003cspan refid=\"Sec5\" class=\"InternalRef\"\u003e3\u003c/span\u003e presents a validation of the hindcast and historical simulations in terms of transports at the main straits, hydrology, and circulation, obtained through comparison with observations and numerical reanalyses. The corresponding sea-level reconstructions are discussed in Sect.\u0026nbsp;\u003cspan refid=\"Sec8\" class=\"InternalRef\"\u003e4\u003c/span\u003e, and validated using satellite altimeter data and coastal tidal gauge observations. Section \u003cspan refid=\"Sec11\" class=\"InternalRef\"\u003e5\u003c/span\u003e then describes the results of the scenario simulation, highlighting future changes in the basin hydrology and circulation, and discussing the projected sea-level. Finally, conclusions are drawn in Sect.\u0026nbsp;\u003cspan refid=\"Sec14\" class=\"InternalRef\"\u003e6\u003c/span\u003e.\u003c/p\u003e "},{"header":"2. Data And Methods","content":" \u003cp\u003eIn order to produce the downscaled Mediterranean sea-level field under present climate (1980\u0026ndash;2005) and the RCP8.5 future scenario (2006\u0026ndash;2100), the MED16 model was forced using the correspondent regional downscaling experiments performed with the SMHI-RCA4 atmospheric model (Strandberg et al., \u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), alternatively constrained by a) the ERA-Interim reanalysis data, for the hindcast simulation, b) the HadGEM2-ES global model (Collins et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2011\u003c/span\u003e) for the historical simulation, and c) the HadGEM2-ES RCP8.5 for the future scenario. The current section describes in detail the atmospheric data used (Sect.\u0026nbsp;\u003cspan refid=\"Sec3\" class=\"InternalRef\"\u003e2.1\u003c/span\u003e) and the MED16 model characteristics and specific setup (Sect.\u0026nbsp;\u003cspan refid=\"Sec4\" class=\"InternalRef\"\u003e2.2\u003c/span\u003e).\u003c/p\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003e2.1 Atmospheric forcing\u003c/h2\u003e \u003cp\u003eThe atmospheric forcing is derived from the dynamically downscaled regional atmospheric fields produced by the Rossby Centre regional atmospheric model RCA4 (Strandberg et al., \u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), at 0.11\u0026deg; resolution (i.e. approx. 12.5 km grid spacing), over the EURO-CORDEX domain (Giorgi et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2009\u003c/span\u003e), whose boundary conditions are either provided by the ERA-Interim reanalysis, or by the atmospheric component of the CMIP5 global model HadGEM2-ES, which is commonly used for downscaled sea-level projections (Hermans et al., \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). The hindcast experiment covers the period 1981\u0026ndash;2010, the historical experiment covers the period 1981\u0026ndash;2005, while the RCP-8.5 scenario spans years from 2006 to 2100. Up to 2005, observed greenhouse gas concentrations have been prescribed, which are then substituted by those indicated by Meinshausen et al. (\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2011\u003c/span\u003e) for the future business-as-usual scenario RCP8.5.\u003c/p\u003e \u003cp\u003eTo reproduce the surface heat fluxes, shortwave radiation from the atmospheric models is imposed, whereas the wind stress and the other heat flux components are computed via bulk formulas considering the sea surface temperature, the winds at 10 m height, the dry air temperature and the air pressure at 2 m, and the relative humidity as inputs. In particular the long-wave radiation is computed according to the formula proposed by Bignami et al. (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e1995\u003c/span\u003e), while the latent heat flux, the sensible heat flux and wind stress are computed according to the Large and Yeager (\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2004\u003c/span\u003e) bulk formula. Cloud cover is taken from the atmospheric model. The net freshwater flux is computed as precipitation (taken from the atmospheric model) minus evaporation (computed from the latent heat). To reduce salinity drift in the hindcast simulation the sea surface salinity (SSS) is restored toward the MEDHYMAP monthly climatology (Jord\u0026agrave; et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2017a\u003c/span\u003e) at a timescale of two days over the topmost model layer of 2 m thickness. At the surface, the model is forced by 6-hourly wind, and 3-hourly surface pressure, heat and freshwater fluxes computed via bulk formulae. The fresh water flux is prescribed in conjunction with the non-linear free surface numerical scheme.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e2.2 The regional ocean model MED16\u003c/h2\u003e \u003cp\u003eThe numerical ocean model MED16 used to simulate the downscaled sea level is the updated version of the Mediterranean ocean model developed by Sannino et al. (\u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). It is based on the MITgcm kernel developed by Marshall et al. (\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e1997a\u003c/span\u003e, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003eb\u003c/span\u003e), and solves the Navier-Stokes equations under the Boussinesq approximation for an incompressible fluid. In the present study, the hydrostatic version of the model has been implemented, using a finite-volume spatial discretization on a staggered Arakawa-C grid, partial step topography, rescaled vertical height (z*) coordinate (Adcroft and Campin \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2004\u003c/span\u003e), and an implicit non-linear free surface formulation (Campin et al. \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2004\u003c/span\u003e). The source code and documentation are available at the following web site: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/MITgcm/MITgcm\u003c/span\u003e\u003c/span\u003e (last access: 20 April 2021).\u003c/p\u003e \u003cp\u003eModel bathymetry for both the Mediterranean and the Black Sea was derived from the European Marine Observation and Data Network (EMODnet) 2016 dataset (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www\u003c/span\u003e\u003c/span\u003e. emodnet-bathymetry.eu), while the high-resolution digitalized chart of Sanz et al. (\u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e1991\u003c/span\u003e) provided data for the SoG, and the high-resolution bathymetry for the Bosphorus and Dardanelles Straits was made available by Erkan G\u0026ouml;kaşan (G\u0026ouml;kaşan et al. \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2005\u003c/span\u003e, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2007\u003c/span\u003e), with the permission of the Turkish Navy, Navigation, Hydrography and Oceanography Office. The datasets were combined through a bilinear interpolation on the computational grid, followed by manual check for isolated grid points, islands, and narrow passages (see Sannino et al., \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2015\u003c/span\u003e, \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Both equilibrium tide (i.e. the generating potential is incorporated in the momentum equations as a body force) and tide propagating from the Atlantic Ocean through the Atlantic open boundary, are explicitly applied as in Naranjo et al. \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2014\u003c/span\u003e, Sannino et al. \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2015\u003c/span\u003e, and Palma et al. \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2020\u003c/span\u003e. The equilibrium tide is composed of four tidal components: M2, O1, S2, K1, the semidiurnal and diurnal principal lunar tides, the principal semidiurnal solar tide, and diurnal lunisolar declination tide, respectively. The tidal values used to prescribe the Atlantic tide are derived from the OTIS global tide inverse model (Egbert and Erofeeva, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). Differently from the model configuration of Naranjo et al. \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2014\u003c/span\u003e, and Sannino et al. \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2015\u003c/span\u003e the updated version used in this study includes the Black Sea that is interactively connected to the Mediterranean Sea through the Turkish Straits (Dardanelles and Bosphorus). The model domain therefore covers the whole Mediterranean-Black Sea system and a small part of the Atlantic Ocean west of the Strait of Gibraltar (SoG hereafter), whose western limit corresponds to the only open-boundary condition prescribed. The horizontal computational grid has a uniform resolution of 1/16\u0026deg; (about 7 km), which is significantly increased in correspondence of the three straits where higher resolution is needed to reliably represent the local dynamics, i.e., the SoG (maximum resolution of about 200 m), and the Dardanelles and Bosphorus Straits, where a smoothly varying refinement in the latitudinal and longitudinal directions allows to reach a maximum resolution of 1/200\u0026deg; (about 555 m) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The vertical domain is discretized using 100 z-levels, with grid spacing increasing from 2 m near the surface to 62 m at a depth of 1500 m; below 1500 m a uniform grid spacing of 62 m is used down to the sea floor. The time-step used is one minute. To calculate the vertical eddy viscosity and diffusivity coefficients, we used the 1.5 order Turbulent Kinetic Energy (TKE) closure scheme by Gaspar et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e1990\u003c/span\u003e, adapted from the atmospheric case developed by Bougeault and Lacarrere (1989). The background vertical eddy viscosity and diffusivity were set to 1.5 10\u003csup\u003e\u0026minus;\u0026thinsp;6\u003c/sup\u003e m\u003csup\u003e2\u003c/sup\u003e/s and 10\u003csup\u003e\u0026minus;\u0026thinsp;7\u003c/sup\u003e m\u003csup\u003e2\u003c/sup\u003e/s, respectively. The maximal value of diffusivity allowed to be generated by the turbulence scheme is 100 m\u003csup\u003e2\u003c/sup\u003e/s, in order to let it handle unstable vertical stratification and avoid the use of an enhanced vertical mixing parameterization. A spatial-dependent horizontal viscosity is obtained from the turbulence closure scheme by Leith (\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e1968\u003c/span\u003e). Differently from the well-known scheme by Smagorinsky (\u003cspan citationid=\"CR91\" class=\"CitationRef\"\u003e1963\u003c/span\u003e), Leith\u0026rsquo;s scheme focuses on resolving the direct enstrophy cascade (cascade towards the smaller scales) that is characteristic of 2D turbulence (Fox-Kemper and Menemenlis, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). A constant horizontal diffusivity coefficient (2 m\u003csup\u003e2\u003c/sup\u003e sec\u003csup\u003e\u0026minus;\u0026thinsp;1\u003c/sup\u003e) is applied with a laplacian operator for the tracers (T,S). The no-slip conditions were used at the lateral and bottom solid boundaries, along with a quadratic bottom drag. The latter is calculated as a function of the velocity close to the bottom, with a dimensionless coefficient of 0.0025. The selected tracer advection scheme is a third-order direct space-time flux limited scheme.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAs the low spatial resolution of the global model did not allow an accurate representation of the dynamics within the Strait of Gibraltar, the Mediterranean basin was completely detached from the Atlantic Ocean in the global simulation, by artificially closing the Strait. The Mediterranean was therefore virtually represented as a lake in the global projection, thus impairing its use to initialize the regional ocean simulation. Initial conditions (namely temperature and salinity) were thus derived from MEDHYMAP climatological data. To reduce spurious drift, MED16 was spun-up 35 years, using the hindcast forcing for the 1980\u0026ndash;1986 period in a five-time loop. The span-up model was then used as initial condition for both historical and hindcast simulations.\u003c/p\u003e \u003cp\u003eAt the Atlantic open boundary, consistently with the surface atmospheric forcing, the hindcast simulation was forced with ORAS4 global reanalysis (Balmaseda et al. \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2013\u003c/span\u003e), while for historical and RCP8.5 scenario simulations monthly data from the HadGEM2-ES global model were used. In particular, the lateral boundary conditions consist of monthly mean temperature, salinity and sea surface height (SSH) which are interpolated onto the MED16 grid. Temperature and salinity are applied through a 3D relaxation with a relaxation time varying linearly from 2 hours at the western limit of the domain to 30 days over the first 30 grid points. The prescribed SSH on the Atlantic box includes different contributions depending on the forcing. For the hindcast simulation, the prescribed ORAS4 SSH includes sea level contributions from ice sheet mass loss, glaciers ice melt, changes in land water storage, as well as global thermal expansion. As ORAS4 underestimates the regional SSH seasonal cycle in the near Atlantic region compared with the multi-satellite products provided by the Climate Change Initiative (CCI-SLA doi:\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.5270/esa-sea_level_cci-MSLA-1993_2013-v_1.1-201412\u003c/span\u003e\u003c/span\u003e), we have applied a 12-month correction based on the difference between CCI-SLA and ORAS4 as described in Adloff et al. (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). For both historical and scenario simulations, the prescribed SSH includes the dynamic sea level (DSL) and the global mean thermal expansion. DSL is the height of the ocean with respect to the time-invariant geoid that is determined by the dynamical balance associated with ocean density and circulation. DSL includes the regional variability of dynamic topography changes due to water mass advection, thermohaline circulation and to the wind-driven circulation (Gill and Niller \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e1973\u003c/span\u003e) and are part of the standard output in the CMIP5 simulations (variable \u003cem\u003ezos\u003c/em\u003e). \u003cem\u003ezos\u003c/em\u003e was initially de-drifted by removing the linearly fitted HadGEM2-ES control run (which is forced by non-evolving pre-industrial conditions) from each grid point individually. This step removes any artificial signals associated with ongoing spin-up deep ocean and/or limitations in the representation of energy conservation in the model domain, as discussed by Sen Gupta et al. (\u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). As the ocean component of HadGEM2-ES is based on the Boussinesq approximation and conserves volume rather than mass (Greatbatch, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e1994\u003c/span\u003e), the value of zos was further corrected (at each time step and grid point) by subtracting its time dependent global mean. This correction guarantees that the resulting sea-level patterns only reflect fluctuations due to the joint effects of ocean density and circulation (Gregory et al., \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) and thus, it is comparable to those resulting from non-Boussinesq models (Losch et al., \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; Griffies and Greatbatch, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2012\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe corresponding global thermal expansion time series (variable \u003cem\u003ezostoga\u003c/em\u003e) was also corrected for control drift by removing a quadratic fit to the control run\u0026rsquo;s thermal expansion time series before being added to the detrended and zero global mean \u003cem\u003ezos\u003c/em\u003e field. The resulting sea level variable was then used as SSH lateral boundary condition for the MED16 historical and scenario simulations. As in Naranjo et al \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e2014\u003c/span\u003e, Sannino et al \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2015\u003c/span\u003e, and the higher resolution Mediterranean version of Palma et al (\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), SSH and the additional 4 tidal constituents were directly prescribed at the western open boundary.\u003c/p\u003e \u003cp\u003eThe remaining sea level components for the historical and scenario simulations, namely the surface mass balance and dynamic ice sheet contributions from Greenland and Antarctica, the glacier and land water storage contributions, and the Glacial Isostatic Adjustment (GIA), were extracted from the Integrated Climate Data Center (ICDC) at Hamburg University (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://icdc.cen.uni-hamburg.de/en/ar5-slr.html\u003c/span\u003e\u003c/span\u003e), interpolated on the MED16 grid, and added offline. The ICDC dataset is based on the ensemble mean of CMIP5 climate models that were used in the regional sea-level projections of the AR5 (Church et al. \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2013a\u003c/span\u003e) of the IPCC. ICDC provides data on a global grid at 1\u0026deg;\u0026times;1\u0026deg; spatial resolution for different RCP scenarios. In this study, we used central estimation data of the RCP8.5 ensemble mean.\u003c/p\u003e \u003cp\u003eRiver discharge at the river outlets is prescribed by transporting the daily total runoff computed by the forcing atmospheric regional model, along the river network of the WBMplus model (V\u0026ouml;r\u0026ouml;smarty et al. \u003cspan citationid=\"CR103\" class=\"CitationRef\"\u003e1998\u003c/span\u003e; Wisser et al. \u003cspan citationid=\"CR105\" class=\"CitationRef\"\u003e2008\u003c/span\u003e, \u003cspan citationid=\"CR106\" class=\"CitationRef\"\u003e2010\u003c/span\u003e), by means of a Muskingum-Cunge type scheme that solves the Saint-Venant flow equations (Ponce \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e1994\u003c/span\u003e). Only for rivers discharging in the Black Sea, the so computed discharge was bias-corrected to reproduce monthly climatological values reported in the literature (Kara et al. \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2008\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e "},{"header":"3. Model Validation","content":" \u003cp\u003eBefore analysing the model sea level reconstructions (see Sect.\u0026nbsp;\u003cspan refid=\"Sec8\" class=\"InternalRef\"\u003e4\u003c/span\u003e), here we perform a first validation of the hindcast and historical runs. In particular, we compare the simulated water transports at the straits and the Mediterranean hydrology and circulation with the observations, and with the results of a recent reanalysis of Mediterranean water properties by Escudier et al (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). The latter is a product from the Copernicus CMEMS database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://marine.copernicus.eu\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e)\u003c/span\u003e that covers the period 1987\u0026ndash;2019, and results from the assimilation of temperature and salinity vertical profiles and satellite Sea Level Anomaly along track data into a numerical system, consisting of the Nucleus for European Modelling of the Ocean (NEMO), and of a variational data assimilation scheme (OceanVAR).\u003c/p\u003e \u003cp\u003eAs our focus is on the Mediterranean Sea, we will mainly show results for this basin, and report the projected water exchanges at the Dardanelles as a boundary condition, determined by the Black Sea effectively acting as a reservoir. Nevertheless, it is worth noting that the model correctly reproduces the permanent basin-scale, cyclonic boundary current that characterizes the Black Sea circulation.\u003c/p\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003e3.1 Transports at the main straits: Gibraltar, Dardanelles, Bosphorus\u003c/h2\u003e \u003cp\u003eThe explicit treatment of water inflow and outflow at the open boundary of the Mediterranean Sea, coupled with explicit tidal forcing, allows to realistically constrain the mass, energy and momentum exchanges with adjacent basins, as well as to finely modulate the temperature and salinity properties of either flow, in both time and space (see Sannino et al, \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2015\u003c/span\u003e, for a detailed description). Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e reports the simulated averaged long-term (1980\u0026ndash;2011) transports across the main straits (Gibraltar, Dardanelles, and Bosphorus), for both the hindcast and historical experiments. The sections used to compute transports are located at 5.4\u0026deg; W for the SoG, at 26.2\u0026deg; E for the Dardanelles Strait, and at 41\u0026deg; N for the Bosphorus Strait. Transport is positive for an eastward or northward flow.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eAveraged water transport (in m\u003csup\u003e3\u003c/sup\u003e/s x 10\u003csup\u003e3\u003c/sup\u003e) through the Strait of Gibraltar, Dardanelles and Bosphorus over the period 1982\u0026ndash;2011. [Mean value and Monthly standard deviation]. A negative flow is directed towards the Atlantic for the Strait of Gibraltar, and into the Mediterranean Sea for the Dardanelles and the Bosphorus\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNET\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGibraltar\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDardanelles\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eBosphorus\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eHindcast\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e60\u0026thinsp;\u0026plusmn;\u0026thinsp;90\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-2\u0026thinsp;\u0026plusmn;\u0026thinsp;7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-1\u0026thinsp;\u0026plusmn;\u0026thinsp;8\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eHistorical\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e60\u0026thinsp;\u0026plusmn;\u0026thinsp;60\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-4\u0026thinsp;\u0026plusmn;\u0026thinsp;9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-4\u0026thinsp;\u0026plusmn;\u0026thinsp;9\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eINFLOW\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eGibraltar\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003eDardanelles\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eBosphorus\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eHindcast\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e930\u0026thinsp;\u0026plusmn;\u0026thinsp;90\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e15\u0026thinsp;\u0026plusmn;\u0026thinsp;2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5\u0026thinsp;\u0026plusmn;\u0026thinsp;4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eHistorical\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1080\u0026thinsp;\u0026plusmn;\u0026thinsp;80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e14\u0026thinsp;\u0026plusmn;\u0026thinsp;3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5\u0026thinsp;\u0026plusmn;\u0026thinsp;3\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOUTFLOW\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eGibraltar\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003eDardanelles\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003eBosphorus\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eHindcast\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-870\u0026thinsp;\u0026plusmn;\u0026thinsp;100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-17\u0026thinsp;\u0026plusmn;\u0026thinsp;7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-7\u0026thinsp;\u0026plusmn;\u0026thinsp;4\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eHistorical\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-1020\u0026thinsp;\u0026plusmn;\u0026thinsp;70\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-18\u0026thinsp;\u0026plusmn;\u0026thinsp;7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-9\u0026thinsp;\u0026plusmn;\u0026thinsp;6\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eFor the hindcast simulation the net average transport of 0.06 Sv at the Strait of Gibraltar results from a mean inflow from the Atlantic of about 0.93 Sv and a Mediterranean outflow of about \u0026minus;\u0026thinsp;0.87 Sv, and compares favorably with previous estimates based on the evaluation of the Mediterranean long-term hydrological deficit (E-P-R), which needs to be compensated by the excess Atlantic inflow. As a matter of fact, a recent work by Jord\u0026agrave; et al. (\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2017b\u003c/span\u003e) reports values ranging from 0.04 Sv (Soto-Navarro et al., \u003cspan citationid=\"CR92\" class=\"CitationRef\"\u003e2010\u003c/span\u003e) to 0.06 Sv (Boutov et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), while proposing an intermediate estimate of 0.05\u0026thinsp;\u0026plusmn;\u0026thinsp;0.03 Sv based on an independent evaluation of the hydrological budget, and a lower bound of 0.03 Sv deriving from the optimal interpolation (OI) of Acoustic Doppler Current Profiler (ADCP) and tide gauge data. The projected average outflow is also close to previous estimates reported in the literature, which range from \u0026minus;\u0026thinsp;0.67 Sv (Tsimplis and Bryden, \u003cspan citationid=\"CR101\" class=\"CitationRef\"\u003e2000\u003c/span\u003e) to -0.97 Sv (Candela \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2001\u003c/span\u003e), and include the \u0026minus;\u0026thinsp;0.78 Sv estimate of S\u0026aacute;nchez Rom\u0026aacute;n et al. (\u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e2009\u003c/span\u003e), obtained by combining three years of ADCP observations and numerical modelling. A higher estimated value of \u0026minus;\u0026thinsp;0.865\u0026thinsp;\u0026plusmn;\u0026thinsp;0.136 Sv has been recently obtained by Garc\u0026iacute;a-Lafuente et al. (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) from the analysis of experimental data recorded over the period 2004\u0026ndash;2012. The averaged simulated Atlantic inflow is 0.93 Sv, which is also in the range of published values (e.g. 0.78 Sv in Tsimplis and Bryden \u003cspan citationid=\"CR101\" class=\"CitationRef\"\u003e2000\u003c/span\u003e; 0.96 Sv in Garc\u0026iacute;a-Lafuente et al. \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). The net transport for the historical experiment is in excellent agreement with the hindcast value, although exhibiting reduced variability, yet it results from higher inflow and outflow.\u003c/p\u003e \u003cp\u003eThe net transport across both Turkish Straits (TSS) is on average negative, consistently with the observed net inflow of fresher waters into the Aegean Sea from the Black Sea, where precipitation and river discharge exceed evaporation. The magnitude of the inflow appears to be lower than previous observational and numerical estimates, which range from 0.006 Sv to 0.01 Sv (\u0026Uuml;nl\u0026uuml;ata et al., \u003cspan citationid=\"CR102\" class=\"CitationRef\"\u003e1990\u003c/span\u003e; Simonov and Altman, \u003cspan citationid=\"CR88\" class=\"CitationRef\"\u003e1991\u003c/span\u003e; Peneva et al., \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e2001\u003c/span\u003e; Kara et al., \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2008\u003c/span\u003e; Sannino et al., \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eHowever, fluxes through the TSS are difficult to accurately measure, due to the large variability in the observed currents and the consequent low signal-to-noise ratio, which might need longer-term monitoring to be adequately increased (Sannino et al., \u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). On the other hand, the simulated net transport is expected to significantly benefit by further refining the grid in correspondence of both Straits, as demonstrated by Sannino et al. (\u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), where a very-high-resolution simulation was performed to accurately reproduce transports with the prescription of climatological barotropic forcing at the two open boundaries. Increased resolution together with a specific treatment of local viscosity and diffusivity is therefore liable to enhance water exchanges across the straits, and to further improve the already good model performance.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e3.2 Basin hydrology and circulation\u003c/h2\u003e \u003cp\u003eWe first compare model climatologies of the winter and summer sea surface temperature (SST) and SSS with reanalysis data; mean fields are computed over the common period 1987\u0026ndash;2005. In Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, panels (a-b) are for the Mediterranean Optimally Interpolated Sea Surface Temperature (OISST) reanalysis of Marullo et al. (\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2007\u003c/span\u003e), panels (c-d) for the Copernicus reanalysis, panels (e-f) for the hindcast, and panels (g-h) for the historical simulation. The model simulations are in an overall good agreement with the reanalyses, even though MED16 systematically produces lower temperatures in winter (more so in the historical simulation). In summer, the SST from the hindcast is in excellent agreement with both reference fields, whereas the historical simulation displays slightly warmer temperatures in the western basin, in the central part of the eastern basin and in the Adriatic Sea, and lower ones in the Ionian, and in the easternmost part of the Levantine.\u003c/p\u003e \u003cp\u003eIn Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e, the SSS model climatologies (panels c-d, hindcast; panels e-f, historical), are compared with the Copernicus renalysis (panels a-b). In the western basin, the hindcast salinity appears to slightly overestimate the observations in both seasons, although less notably so in summer. In the eastern basin, the salinity field is generally well reproduced in winter, while in summer some local negative biases are present (e.g, in the Gulf of Gabes, in the Adriatic Sea, and in the north Aegean). The historical experiment, on the contrary, exhibits larger salinity values in both seasons and basins, yet reducing the positive bias around the Balearic Islands, coherently with the limited intrusion of the Atlantic Water (AW) into the Tyrrhenian Sea, the Sicily channel and the Gulf of Gabes, and with the more pronounced northward spreading of the Atlantic inflow after exiting the Alboran Sea.\u003c/p\u003e \u003cp\u003eSpecific local differences between the numerical experiments and the reanalysis can be attributed to the different treatment of rivers and/or to the different magnitudes of the runoff provided by the driving atmospheric simulations (e.g. along the eastern Adriatic coast and in the North Adriatic Sea), while the different behaviour of the two MED16 runs in the North Aegean is probably related to the representation of the local circulation, as no significant difference was detected in the representation of the net inflow from the Dardanelles between the hindcast and the historical experiment. This discrepancy will be the object of future research.\u003c/p\u003e \u003cp\u003eThe hydrological structure and the circulation produced by the hindcast and historical simulations were further validated via comparison with in situ and remote observations. Vertical means of the modelled and reanalysed temperature and salinity were computed over the 1987\u0026ndash;2005 time window, for three different vertical layers, and averaged over the whole Mediterranean basin (MED), and the western and eastern sub-basins (WMED and EMED). Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e shows the reference values from the reanalysis, together with the differences between model and reference for the two experiments, for the surface layer (0 -150m), the intermediate layer (150-600m), and the deep layer (600-3500m).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eFor each geographical area in the leftmost column - the whole Mediterranean (MED) and the Western (WMED) and Eastern (EMED) sub-basins - the first row reports the average temperature (\u003cem\u003e\u0026deg;C\u003c/em\u003e) and salinity (\u003cem\u003ePsu\u003c/em\u003e) at various depths, from the Copernicus Reanalysis dataset; the second and third rows (labelled \u0026ldquo;\u003cem\u003ehindcast\u003c/em\u003e\u0026rdquo; and \u0026ldquo;\u003cem\u003ehistorical\u003c/em\u003e\u0026rdquo;) respectively report the associated differences between the hindcast and the historical simulations and the reanalyses. All spatial averages are taken over the period 1987\u0026ndash;2005\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"8\"\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e \u003cp\u003eTemperature\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"3\" nameend=\"c8\" namest=\"c6\"\u003e \u003cp\u003eSalinity\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e \u003cp\u003eDepth [m]\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c8\" namest=\"c6\"\u003e \u003cp\u003eDepth [m]\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003e0-150\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003e150\u0026ndash;600\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003e600\u0026ndash;3500\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cb\u003e0-150\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e\u003cb\u003e150\u0026ndash;600\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e\u003cb\u003e600\u0026ndash;3500\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003e\u003cb\u003eMED\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eReanalysis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e16,54\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e14,00\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e13,32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e38,39\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e38,81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e38,65\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHindcast\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0,64\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0,14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0,05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0,12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-0,15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e-0,04\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHistorical\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0,78\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0,27\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0,18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0,04\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-0,08\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0,04\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003e\u003cb\u003eWMED\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eReanalysis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e15,37\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e13,45\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e12,93\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e37,83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e38,49\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e38,47\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHindcast\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0,39\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0,29\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0,01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0,04\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-0,10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0,01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHistorical\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0,57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0,38\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0,09\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0,01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-0,05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0,07\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003e\u003cb\u003eEMED\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eReanalysis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e17,17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e14,34\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e13,55\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e38,70\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e38,84\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e38,80\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHindcast\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0,74\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0,06\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0,08\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0,15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-0,03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0,00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHistorical\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0,86\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-0,21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0,26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0,07\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0,05\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0,09\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eDespite the differences highlighted in the spatial patterns, the agreement between the hindcast and historical simulation and the reference data is good. However, the modelled temperatures are always slightly lower than the observations in the two upper layers and slightly higher in the deep layer. For both simulations and all geographical areas, maximum discrepancies are anyway found in the surface layer, with largest differences in the Eastern basin. When the whole basin is considered, modelled salinity also generally presents negative anomalies, with maximum negative differences in the surface and intermediate layers of the hindcast simulation, while slightly positive anomalies can be observed in the surface and bottom layers for the historical simulation. When separately considering the two sub-basins, the hindcast experiment exhibits more marked negative biases in the intermediate layer of the Western basin and in the surface layer of the Eastern basin, while the historical simulation projects higher salinity in the surface layer of the Eastern basin with respect to the reanalysis. We note that the biases for temperature and salinity are well inside the ensemble spread computed by Llasses et al. (\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) from the Med-Cordex simulations.\u003c/p\u003e \u003cp\u003eIn order to evaluate the ability of the hindcast experiment in reproducing the time evolution of water temperature, Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e shows the time series (1987\u0026ndash;2010) of the mean annual temperature anomalies in the surface and intermediate layers, for the whole basin and the western and eastern sub-basins, as simulated by the hincast experiment. Comparison with the historical simulation is omitted here as time coherence with the reanalysis could not be expected. The observed interannual variability is generally well captured by the hindcast simulation, both in the surface and in the intermediate layer.\u003c/p\u003e \u003cp\u003eThe simulation correctly reproduces the observed interannual variability. The EMED time series, in particular, captures the positive trend in both the upper and intermediate layers. Moreover, the simulation well represents the strong cooling during the years 1992 and 1993, that is the signature of the Eastern Mediterranean Transient (EMT) phenomenon.\u003c/p\u003e \u003cp\u003eAnalogous considerations apply to the time series of mean annual salinity anomalies (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). The general trend of simulated salinity and reanalysis are also similar, except for the intermediate layer of WMED where the simulated salinity does not follow the reanalysis trend. In the WMED the model captures both the sudden drop in salinity around 1990 and the maximum in 2005. The first event is due to an excess of freshwater input and is confined to the upper layers, whereas the latter is related to a well-known event of intense dense water formation in the Gulf of Lion. A good representation of the interannual variability of the salinity field is also achieved in the EMED, with a local maximum during the 90s and an increasing trend in the second half of the simulation.\u003c/p\u003e \u003cp\u003eA notable discrepancy occurs in the years from 2001 to 2004, in which the reanalysis displays a relative minimum, particularly strong in the WMED, which is not present in the model. However, the corresponding data from the MEDHYMAP dataset show a very good agreement with MED16 also for the period 2000\u0026ndash;2005.\u003c/p\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e shows the surface (15 m of depth) and intermediate (300 m of depth) model circulations, averaged over the simulation time-span (1982\u0026ndash;2010).\u003c/p\u003e \u003cp\u003eAt both depths, the model circulation is close to the 1987\u0026ndash;2007 time-mean flow displayed in Pinardi et al. (\u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2015\u003c/span\u003e), which results from a reanalysis of the Mediterranean Sea circulation (see their Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, displaying the circulation at 15 m of depth, and the average circulation in the layer from 200 m and 300 m of depth). The main Mediterranean surface current systems, either driven by the local wind forcing, or by the inflow of AW at the Gibraltar Strait, are present both in our hindcast and in the reanalysis, with similar shapes. This also applies to most of the robust cyclonic and anticyclonic circulation structures, even though there are a few features, such as the quasi-permanent cyclonic circulation to the southeast of Sardinia, and some smaller vortices in the eastern basin, that are not well defined in the hindcast. Overall, considering that the latter is the result of a \u0026ldquo;free\u0026rdquo; simulation spanning almost three decades, whereas the reanalysis relies on the assimilation of observed data, the representation of the average Mediterranean circulation provided by our model can be considered quite satisfactory.\u003c/p\u003e \u003cp\u003eThree hindcast sub-regional surface circulation maps are displayed in Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e. Panel (a) of the figure highlights the two permanent cyclonic circulations that characterize the northern portion of the western Mediterranean, that is, the wide gyre formed by the Liguro-Proven\u0026ccedil;al current, occupying the western Ligurian Sea, and the Bonifacio cyclone, in the northern part of the Tyrrhenian Sea, to the east of Corsica. Interestingly, between the two cyclonic circulations, there is the weak signature of an anticyclonic circulation, in the area of the Corsica Channel, between Cap Corse (the long peninsula in the northern tip of Corsica) and the Italian coast. This structure, which was denoted as Ligurian anticyclone in Ciuffardi et al. (\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), appears to be a robust feature of the local dynamics during the summer months (see also recent works by Poulain et al., \u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e2020\u003c/span\u003e, and by Iacono and Napolitano, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAs shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eb, the model correctly reproduces the wide cyclonic structure, with two poles, that occupies the southern, deeper portion of the Adriatic Sea (see, e.g., recent work by Palma et al., \u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e2020\u003c/span\u003e, and references therein).\u003c/p\u003e \u003cp\u003eFinally, we note that the surface dynamics of the Marmara Sea (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003ec) are dominated by a central, elongated gyre, more attached to the northern coast, in agreement with what was found in a recent high-resolution numerical investigation of the Turkish Straits System by Sannino et al. (\u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe time evolution of the Mixed Layer Depth (MLD - Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003e) highlights how deep-water-formation processes in the Gulf of Lion and the Adriatic basin are well represented both in the hindcast and in the historical experiments.\u003c/p\u003e \u003c/div\u003e "},{"header":"4. Hindcast And Historical Sea-level Analysis","content":" \u003cp\u003eThe Mediterranean Sea level average and variability simulated in the hindcast experiment are compared to their observed counterparts, derived from altimetric satellite sea level anomaly (SLA) measurements. The model ability to reproduce nearshore sea level is assessed by means of tide gauge data.\u003c/p\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e4.1 Comparison with satellite observations\u003c/h2\u003e \u003cp\u003eThe altimeter data used are daily maps of gridded L4 SLA and derived variables reprocessed (Copernicus portal, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://marine.copernicus.eu\u003c/span\u003e\u003c/span\u003e, last access april 2021), obtained from the merging of track data from different satellites. These maps, especially developed for the Mediterranean Sea, have a nominal resolution of 1/8\u0026deg; \u0026times; 1/8\u0026deg;, which is twice the resolution of the products that cover the world ocean. They have been available since 1993, overlapping the last 18 years of the simulation. A different altimeter dataset was used in Adloff et al. (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) (see Ablain et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2015\u003c/span\u003e, for details), which consists of monthly SLA maps with a spatial resolution of 1/4\u0026deg;.\u003c/p\u003e \u003cp\u003eThe average seasonal cycle reproduced by the hindcast is compared to that of the observations in Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e, where the solid lines with black dots refer to model results and the dashed lines with diamonds to altimeter data (the averages are over the period 1993\u0026ndash;2010). Panel (a) shows that the model cycle over the whole basin is fairly close to the observations, with some overestimation in winter and some underestimation at the end of summer and in part of the autumn.\u003c/p\u003e \u003cp\u003eOverall, the model cycle is close to that obtained from the NEMO-MED12 simulation discussed in Adloff et al. (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), which was found to provide the best SL reconstruction among the four considered in that work.\u003c/p\u003e \u003cp\u003eThe two other panels in the figure show the same comparison for the western and eastern sub-basins, respectively. The cycle for the western Mediterranean is even closer to the observations than the basin cycle, whereas in the eastern basin, where the amplitude of the observed cycle is slightly larger, the model underestimation during SON is also larger. This may be due to the fact that the steric contribution to the SL, which is in principle absent in most state-of-the art circulation models, is here introduced indirectly through the Atlantic boundary conditions. As we have seen, this works pretty well, particularly in the western basin, but it may be not sufficient to account for additional sea-level seasonal variations induced by local heat fluxes, especially in the eastern sub-basin.\u003c/p\u003e \u003cp\u003eThe interannual variability of the sea-level anomaly of the hindcast simulation is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003e, where the same notations of the previous figure are adopted. The plotted values are yearly anomalies with respect to the mean over the period covered by the altimeter (1993\u0026ndash;2010), averaged over four different regions: the Atlantic box, the whole Mediterranean basin, the WMED, and the Levantine sub-basin. The first panel is relative to the Atlantic box, i.e., the small domain to the west of the Gibraltar Strait where boundary conditions are prescribed, making use of the available altimetric data (see Sect.\u0026nbsp;2 for details). From 1993 on, the model time series appears well correlated with the observations, showing that the boundary conditions are correctly applied in our model. In fact, the cross-correlation coefficients of Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e show that the model time series are very well correlated with the corresponding altimeter-derived time series in all the basins in consideration.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eCross-correlation between the yearly sea level anomaly time series and that of the SLA, over the period 1993\u0026ndash;2010, for four different basins\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAtlantic box\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eMediterranean\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eWestern Med.\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eLevantine\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCorrelation with SLA\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.81\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.78\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0.87\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eThe time series of the average Mediterranean SL anomaly (top right panel in Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003e) reproduces the growing basin-scale trend displayed by the altimetric observations (see Bonaduce et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2016\u003c/span\u003e, and the more recent investigation by Mohamed et al., \u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e2019\u003c/span\u003e, where trends of the SL in the Mediterranean were discussed, based on 25 years of altimeter data). It also captures short-time variations, such as the sharp increase seen in 2010, which results from a large wintertime jump of the SL over the whole Mediterranean Sea. As shown in Landerer and Volkov (\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2013\u003c/span\u003e), this was due to a basin-wide barotropic fluctuation induced by the variation of the wind stress in the Gibraltar area and in the neighboring Atlantic sector, associated with inflow of Atlantic water through the Gibraltar Strait. A similar event occurred during the following winter of 2011.\u003c/p\u003e \u003cp\u003eComparison with Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e of Adloff et al. (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) shows that the model time series of Mediterranean anomalies also captures the main interannual variability observed before 1993; the reference for this period is provided by two reconstructions of the Mediterranean SL by Calafat and Jord\u0026agrave; (\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2011\u003c/span\u003e), and Meyssignac et al. (\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2011\u003c/span\u003e), based on data from coastal tide gauges. The series is also in good agreement with that obtained from the NEMO-MED12 simulation of Adloff et al. (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). This is due to the fact that, despite the many differences, the two models apply the same Atlantic boundary conditions, and use as local atmospheric forcing two different downscaling of the same dataset (ERA-INTERIM). In some cases, the model series displays stronger SL variations; this may be ascribed to model differences, but could also partly be due to the contribution of mesoscale features that are better resolved by the model, thanks to the higher horizontal resolution (1/16\u0026deg;). This holds a fortiori, when compared with the altimeter data, whose nominal resolution is 1/8\u0026deg;x1/8\u0026deg;.\u003c/p\u003e \u003cp\u003eThe lower row of Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003e shows the SL anomalies for the sub-basin that is closest to the Atlantic (western Mediterranean) and for that more far away from it (Levantine basin, to the east of 22\u0026deg; of longitude). The signals are quite different in the two sub-basins; the series in the western Mediterranean resembles that in the Atlantic box, with a fairly regular increase of the SL over the simulation period, whereas that in the Levantine shows a decrease of the SL at the end of the 80\u0026rsquo;s and beginning of the 90\u0026rsquo;s, followed by a sharp increase, leading to a different state over the last 10 years of the simulation. The latter behavior is of course related to the Eastern Mediterranean Transient (EMT), the major event that has affected the circulation and water mass properties of the Mediterranean in the last decades. The hindcast is capable of reproducing the rapid SL rise as a result of strong anomalous mass modification in the deep layers of the EMED during the end of the 80s and the beginning of the 90s; the effects of the EMT on sea level has been discussed for the first time by Tsimplis and Josey (\u003cspan citationid=\"CR100\" class=\"CitationRef\"\u003e2001\u003c/span\u003e). The model correlates very well with the altimeter observations after the transient, which display a definitely nonlinear trend, first evidenced in Amitai et al. (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2010\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eWe can conclude that the hindcast simulation not only captures the basin scale variability of the SL over the hindcast period, but also some expected, significant differences between the main Mediterranean sub-basins.\u003c/p\u003e \u003cp\u003eIn Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e11\u003c/span\u003e, the modelled (both hindcast and historical) SSH fields, averaged over the period 1993\u0026ndash;2011, are compared with the corresponding climatological average of the absolute dynamic topography (ADT) from altimeter data (AVISO, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.aviso.oceanobs.com\u003c/span\u003e\u003c/span\u003e, last access May 10th 2021). In all cases, the respective regional climatological means have been subtracted to ease the comparison of patterns. Being a long-term average, the field derived from the AVISO data is expected to be very close to the Mean Dynamic Topography (MDT), that is, to the mean signal, representative of the average circulation, which produces the ADT when added to the zero-mean remotely measured sea level anomaly (SLA). The reference observational field shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e11\u003c/span\u003e is indeed very close to that of Rio et al. (\u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e2014\u003c/span\u003e), which was constructed using observations and model results, and validated using independent observations.\u003c/p\u003e \u003cp\u003eThe spatial pattern of the modelled SSH field generally appears to be in good agreement with the observations, and is also similar to that obtained by the NEMO-MED12 model, the best performing among those considered in Adloff et al. (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). As to the hindcast experiment, the MED16 model outperforms NEMO-MED12 in the western portion of the basin, indicating that the stretched grid at Gibraltar indeed allows for a better representation of the complex dynamics inside the Strait. Nevertheless, the signatures of the Rhodes gyre and of the western Cretan gyre are weaker than expected. On the other hand, neither the MED16 simulation nor the four hindcasts analyzed in Adloff et al. (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2018\u003c/span\u003e) resolve the Ierapetra gyre, possibly due to the still insufficient spatial resolution or to the quality of the downscaled wind forcing in the area, the Ierapetra gyre being a wind driven feature.\u003c/p\u003e \u003cp\u003eThe SSH values obtained for the Black Sea (not shown here) exhibit larger spatial variability with respect to those of the Mediterranean Sea, approximately ranging from \u0026minus;\u0026thinsp;6 cm to 18 cm, and compare well with the MDT reconstructions by Knysh et al. (\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2002\u003c/span\u003e) and by Kubryakov and Stanichny (\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). In particular, Kubryakov and Stanichny (\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2011\u003c/span\u003e) find localized, deeper minima in the core of the dominant cyclonic cell, while MED16 produces significantly higher SSH values than either reconstruction in the westernmost portion of the same cell.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e4.2 Comparison with tide gauges\u003c/h2\u003e \u003cp\u003eHindcast sea surface heights along the coast have been compared with tide gauge data retrieved from the Permanent Service for Mean Sea Level (PSMSL) (Holgate et al., \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; PSMSL, 2020).\u003c/p\u003e \u003cp\u003eIn particular, we used time series of sea level monthly means from the Revised Local Reference dataset (RLR), which are independently calibrated for each station in the PSMSL according to the tide gauge history provided by each local supplying authority. However, no correction is made by PSMSL for vertical land movements (VLM), although tide gauges without known benchmarks are omitted from this data set by PSMSL prior to release (Hamlington et al., \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). These data therefore measure the relative movement of the sea surface with respect to the land, disregarding the multiple processes over a variety of time scales that can affect MSL estimates, some of natural origin, such as the glacial isostatic adjustment (GIA) and plate tectonics, and some related to human activities (e.g. ground-water pumping). In the following, RLR data have been used without any further correction for VLM, selecting all the available Mediterranean stations which covered at least 50% of the simulated period. In order to avoid relative biases due to different zero-level assumptions, monthly data from the model grid point nearest to the tide gauge have been compared with the monthly mean tide gauge data after subtraction of the respective mean values, over the common time interval, from both modelled and tide gauge time series.\u003c/p\u003e \u003cp\u003eThe 34 locations for which the comparison was carried out are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003e.\u003c/p\u003e \u003cp\u003eThe cross-correlation coefficients obtained are detailed in Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e, where the geographic location of the stations and the number of observations are also indicated.\u003c/p\u003e \u003cp\u003eThe coefficients are pretty high in most of the stations; they are smaller than 0.65 only in 8 stations (17, 23, 25, 26, 28, 29, 31, 34) located in the Aegean Sea or nearby, and one in the Black Sea. This is quite encouraging, considering that the model values are not at the stations\u0026rsquo; sites, which are very close to the coast, and that the model freely evolves for three decades. It shows that, in part of the basin, the model is capable of providing a fairly accurate description of the seal level variability on long time scales, not only over large spatial scales, but also nearshore, where the consequences of sea level change are stronger. The fact that the model performs less well in the Aegean Sea is likely partly due to difficulties in reproducing small scale details of the near coast dynamics, induced by the very complex bathymetry of the area.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eList of tide gauges used with the corresponding code number and coordinates. Correlation between the observed time series and model data is shown together with the standard deviation of the two series of data and the number of times considered\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"9\"\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003enumber\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003estation\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003ecode\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003elongitude\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003elatitude\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003ecorrelation\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003en. times\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003estd station\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c9\"\u003e \u003cp\u003estd model\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003e1\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eTARIFA\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e488\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-5,616\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e36,020\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,651\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e321\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e6,821\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e4,177\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eALGECIRAS\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e490\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-5,422\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e36,132\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,717\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e249\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e5,974\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e3,936\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eCEUTA\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e498\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-5,299\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e35,910\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,659\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e330\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e5,502\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e4,205\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eMALAGA\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e496\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-4,407\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e36,668\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,777\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e349\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e7,606\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e4,469\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003ePORT VENDRES\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1469\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3,094\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e42,594\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,766\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e210\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e7,952\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e6,113\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e6\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eL'ESTARTIT\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1764\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3,281\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e42,031\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,825\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e241\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e7,472\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e6,630\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e7\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eMARSEILLE\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e61\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5,344\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e43,281\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,797\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e316\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e7,207\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e6,318\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e8\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eTOULON\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e980\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5,906\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e42,969\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,711\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e217\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e7,041\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e6,772\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e9\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eNICE\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1468\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e7,281\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e43,656\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,717\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e300\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e7,101\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e6,373\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eVENEZIA (PUNTA DELLA SALUTE)\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e168\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e12,406\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e45,281\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,813\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e239\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e8,804\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e8,607\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eROVINJ\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e761\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e13,656\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e45,094\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,799\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e349\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e8,246\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e8,630\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e12\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eTRIESTE\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e154\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e13,781\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e45,656\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,810\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e349\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e8,450\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e8,718\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e13\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eSPLIT - GRADSKA LUKA\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e352\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e16,406\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e43,344\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,761\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e349\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e7,925\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e8,293\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e14\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eSUCURAJ\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1706\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e16,906\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e43,031\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,732\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e221\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e8,431\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e8,321\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e15\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eDUBROVNIK\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e760\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e18,031\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e42,656\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,754\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e345\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e7,568\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e8,049\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e16\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eLEVKAS\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1239\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e20,719\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e38,906\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,702\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e310\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e9,384\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e7,356\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e17\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003ePREVEZA\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e410\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e20,719\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e38,906\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,585\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e301\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e7,579\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e7,380\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e18\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eKATAKOLON\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1240\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e21,344\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e37,656\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,717\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e300\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e7,720\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e7,412\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e19\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eKALAMAI\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e411\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e22,156\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e36,969\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,770\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e276\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e7,947\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e7,677\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eTHESSALONIKI\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e373\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e22,844\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e40,536\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,680\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e329\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e8,847\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e8,014\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e21\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eNORTH SALAMINOS\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1604\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e23,531\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e37,906\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,735\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e180\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e8,305\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e7,856\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e22\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eSOUDHAS\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1232\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e24,094\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e35,781\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,678\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e338\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e7,668\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e7,607\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e23\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eSIROS\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1234\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e24,969\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e37,531\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,577\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e317\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e8,687\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e7,644\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e24\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eIRAKLION\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e634\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e25,156\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e35,531\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,669\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e235\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e10,500\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e7,208\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eALEXANDROUPOLIS\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1238\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e25,906\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e40,825\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,635\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e312\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e8,132\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e7,820\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e26\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eKHIOS\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e408\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e26,156\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e38,656\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,622\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e316\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e8,108\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e7,746\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e27\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eMENTES/IZMIR\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1679\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e26,719\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e38,594\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,686\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e265\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e8,062\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e7,785\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e28\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eLEROS\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1233\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e26,844\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e37,094\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,522\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e300\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e6,432\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e6,800\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e29\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eRHODOS\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1243\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e28,219\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e36,469\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,649\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e220\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e7,741\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e6,415\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eALEXANDRIA\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e503\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e29,906\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e31,281\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,708\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e301\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e9,142\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e7,470\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eERDEK\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1598\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e27,856\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e40,383\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,478\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e255\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e9,030\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e6,460\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eANTALYA II\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1681\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e30,719\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e36,781\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,736\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e253\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e9,541\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e7,811\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e33\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003eTUAPSE\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e215\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e39,125\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e44,026\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,724\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e346\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e9,428\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e9,200\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e34\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e\u003cb\u003ePOTI\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e41\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e41,622\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e42,192\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0,533\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e \u003cp\u003e336\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e \u003cp\u003e12,538\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c9\"\u003e \u003cp\u003e9,128\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eSea surface height time series for six stations (red circles in Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003e) where VLM can be considered to be negligible are plotted in Fig.\u0026nbsp;\u003cspan refid=\"Fig13\" class=\"InternalRef\"\u003e13\u003c/span\u003e. The main interannual variability is reproduced fairly well at all stations, and especially at Malaga, Trieste, Preveza and Alexandria.\u003c/p\u003e \u003c/div\u003e "},{"header":"5. Scenario Simulation","content":" \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e5.1 Future basin hydrology and circulation\u003c/h2\u003e \u003cp\u003eFor the RCP8.5 scenario, Fig.\u0026nbsp;\u003cspan refid=\"Fig14\" class=\"InternalRef\"\u003e14\u003c/span\u003e shows the differences in the future surface temperature climatologies with respect to the reference period. Averages are computed over years 2046\u0026ndash;2065 (left) and 2081\u0026ndash;2100 (right), for winter (top) and summer (bottom). The sustained increase in temperature is evident in both seasons. As for the correspondent salinity fields (Fig.\u0026nbsp;\u003cspan refid=\"Fig15\" class=\"InternalRef\"\u003e15\u003c/span\u003e), the salinification of the Eastern basin is apparent, especially in the north, as well as the northward displacement of the Atlantic stream in the Western basin, yet accompanied by a more efficient penetration of the AW into the Ionian basin under the future scenario with respect to the historical simulation.\u003c/p\u003e \u003cp\u003eThe temperature-induced additional buoyancy is evident in the mixed layer depth which exhibits significantly lower relative maxima at dense water formation sites. The deep convection in the Gulf of Lion is almost neutralised, whereas in the Adriatic Sea is characterised by enhanced variability (Fig.\u0026nbsp;\u003cspan refid=\"Fig16\" class=\"InternalRef\"\u003e16\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eBasin-wide averages of T and S are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab5\" class=\"InternalRef\"\u003e5\u003c/span\u003e, where substantial increases in temperature generally characterize the surface and intermediate layers, while variations in salinity appear to be most relevant and coherent in the intermediate layer. Results are consistent with the findings of Soto-Navarro et al. (\u003cspan citationid=\"CR93\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab5\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eAverage temperature (\u003cem\u003e\u0026deg;C\u003c/em\u003e) and salinity (\u003cem\u003ePsu\u003c/em\u003e) at different depths. Differences between values averaged over two periods (2046\u0026ndash;2065 and 2081\u0026ndash;2100) of the scenario simulation and the whole period (1981\u0026ndash;2005) of the historical simulation. Averages are over the whole Mediterranean Sea (MED), and over the western and eastern sub-basins (WMED and EMED)\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"8\"\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e \u003cp\u003eTemperature\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"3\" nameend=\"c8\" namest=\"c6\"\u003e \u003cp\u003eSalinity\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c5\" namest=\"c3\"\u003e \u003cp\u003e\u003cb\u003eDepth [m]\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"3\" nameend=\"c8\" namest=\"c6\"\u003e \u003cp\u003e\u003cb\u003eDepth [m]\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e\u003cb\u003e0-150\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cb\u003e150\u0026ndash;600\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cb\u003e600\u0026ndash;3500\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cb\u003e0-150\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e\u003cb\u003e150\u0026ndash;600\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e\u003cb\u003e600\u0026ndash;3500\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u003cb\u003eMED\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRCP 2046\u0026ndash;2065\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1,72\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1,48\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0,25\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0,03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0,22\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0,00\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRCP 2081\u0026ndash;2100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3,31\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2,76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0,56\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0,20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0,44\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0,06\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u003cb\u003eWMED\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRCP 2046\u0026ndash;2065\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1,46\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1,44\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0,45\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0,10\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0,27\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0,07\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRCP 2081\u0026ndash;2100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2,98\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2,76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0,83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-0,01\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0,47\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0,15\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u003cb\u003eEMED\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRCP 2046\u0026ndash;2065\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1,87\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1,49\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0,11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0,02\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0,20\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e-0,03\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRCP 2081\u0026ndash;2100\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e3,50\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2,76\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e0,36\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0,32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e0,42\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003e0,02\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003e5.2 Impact of dynamical downscaling on the projected SLC\u003c/h2\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig17\" class=\"InternalRef\"\u003e17\u003c/span\u003e separately quantifies the magnitude and sign of the components, either local or global, that contribute to a projected mean sea level trend of about 7 mm/yr in the Mediterranean, and allows the evaluation of their relative weight. Under RCP8.5 and from 2005 to 2100, it shows the time evolution of each respective central estimate, the ocean dynamic component being derived from an ensemble of 16 GCM participating in the CMIP5 programme, after excluding simulations that do not explicitly represent the open connection between the Mediterranean Sea and the Atlantic Ocean (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://icdc.cen.uni-hamburg.de/en/ar5-slr.html\u003c/span\u003e\u003c/span\u003e, Slangen et al., \u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e2014\u003c/span\u003e).\u003c/p\u003e \u003cp\u003ePrior to spatial averaging, all the variables have been interpolated onto the MED16 grid via a bilinear interpolation, with nearest-neighbor interpolation near the coasts, results from the MED16 regional projection are also shown. The zero level is set at the 1986\u0026ndash;2005 mean that Slangen et al. (\u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) consider as the common zero SSH anomaly value.\u003c/p\u003e \u003cp\u003eThe most relevant contribution is that from the sterodynamic term (dark red; solid line\u0026thinsp;=\u0026thinsp;GCM, dashed line\u0026thinsp;=\u0026thinsp;MED16), which accounts for changes in the circulation and density distribution of sea water and for the inverse barometer (IB) correction associated with atmospheric pressure patterns (Gregory et al., \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Positive changes (here listed in decreasing order) also derive from distributed glacier melting (aside from Greenland and Antarctica ice sheets - green), Antarctic dynamic ice loss (pink), Greenland surface mass balance (light blue), Greenland dynamic ice loss (blue) and glacial isostatic adjustment (red), ground water storage (yellow), while the Antarctic surface mass balance (dark green) determines a reduction in sea level. At the end of the XXI century, these components are projected to respectively account for 51%, 23%, 13%, 10%, 6%, 2.5%, 2%, -7% of the total, with an overall contribution of 16% from Greenland and 6% from Antarctica. Recent works (Spada, \u003cspan citationid=\"CR96\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Santamar\u0026iacute;a-G\u0026oacute;mez et al., \u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) have dwelled on the necessity to consistently include the GIA effect in updated estimates of SLC in the Mediterranean, based on the alternative datasets provided by Lambeck (Lambeck et al., \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e1998\u003c/span\u003e) and Peltier (Peltier et al., 2004), whose limits and reliability have been the subject of a lively scientific discussion (Peltier, 2002; Lambeck et al., \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2002\u003c/span\u003e). Here, the GIA contribution is the one computed by Church et al. (\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2013b\u003c/span\u003e) by averaging results from the ICE-5G model (Peltier et al., 2004) and the ANU ice sheet model (Lambeck et al. \u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e1998\u003c/span\u003e and subsequent improvements), using the updated SELEN code for the sea level equation (Spada and Stocchi \u003cspan citationid=\"CR94\" class=\"CitationRef\"\u003e2006\u003c/span\u003e, \u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). In the Mediterranean basin, however, despite its comparative importance for the assessment of relative sea level rise for present times, the GIA contribution appears to progressively lose relevance by the end of the XXI century, unless ongoing research revises the long-term ranking of individual contributions.\u003c/p\u003e \u003cp\u003eAlthough only representing one of the numerous possible realizations that can be obtained by different GCM-RCM modelling chains, the MSL trend estimated via the MED16 scenario simulation is very similar to that of the GCM ensemble mean from Slangen at al. (2014), although it slightly accelerates from 2040 onwards. With respect to the coarse-resolution global average, the MED16 downscaling better characterizes the regional patterns of SLC arising from ocean dynamics, and it consequently retains enhanced variability.\u003c/p\u003e \u003cp\u003eThe time evolution of the difference in SSH between the Mediterranean basin interior and the neighbouring Atlantic Ocean (average over the region just outside the Gibraltar Strait, from 9W to 6W) is reported in Fig.\u0026nbsp;\u003cspan refid=\"Fig19\" class=\"InternalRef\"\u003e19\u003c/span\u003e, for both the MED16 model (left panel) and the global ensemble mean (right panel). The red curves correspond to the sterodynamic term (i.e., to the dark red curves of Fig.\u0026nbsp;\u003cspan refid=\"Fig17\" class=\"InternalRef\"\u003e17\u003c/span\u003e), the yellow curve in the left panel represents the historical reference experiment, while the AVISO observations are plotted in black. Beside the good agreement between the AVISO data and the historical experiment, it is worth noting that the regional simulation is capable of capturing the hydraulic jump across the Gibraltar Strait (Fig.\u0026nbsp;\u003cspan refid=\"Fig18\" class=\"InternalRef\"\u003e18\u003c/span\u003e) that is completely missed by the global projection, due to the insufficient representation of the local dynamics (Sannino et al. \u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e2009\u003c/span\u003e, \u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Nevertheless, the global experiment also exhibits the relative \u0026ldquo;sinking\u0026rdquo; of the Mediterranean with respect to the Atlantic, albeit following a trend that is half that of the regional experiment, and characterized by lower variability. Some coherence is yet detectable between the respective typical time-scales of variability, possibly when driven by large-scale patterns, as well as a comparable signal-to-noise ratio. However, the global projection seems to be reaching a new stable state towards the end of the XXI century that is not discernible in the regional experiment, a longer run being needed to confirm such a hypothesis. So far, the change in the relative level between the two basins can be estimated to be within centimeters, over an expected sea level increase on the order of tens of centimetres by the end of the century.\u003c/p\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig20\" class=\"InternalRef\"\u003e20\u003c/span\u003e illustrates the spatial distribution of the progressive increase in sea level under RCP8.5 with respect to the historical period 1985\u0026ndash;2005, which qualitatively retains the overall pattern of the dynamic topography under current climate, where the largest increments apparently correspond to the Atlantic signal, mirroring the AW progress into the basin interior. The main notable difference appears to be the sustained amplified signal around the Balearic Islands, presumably itself a consequence of the northward displacement of the Atlantic inflow that numerical models associate with the newly detected Western Mid-Mediterranean Current (WMMC). This current corresponds to the northward meandering stream arising from the bifurcation of the Atlantic inflow after exiting the Alboran Sea and overcoming the Almera-Oran front (Arnone et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e1990\u003c/span\u003e, Pinardi et al., \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). The WMMC branch of the Atlantic inflow has also been detected in the hindcast and historical experiments (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e), yet it has never been well documented in observational datasets. Pinardi et al. (\u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) tentatively interpret it as a residual current resulting from multi-year averaging.\u003c/p\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig21\" class=\"InternalRef\"\u003e21\u003c/span\u003e reports the differences between MED16 and the GCM ensemble mean for the two targeted time horizons. The overall pattern is maintained in time, with localized yet roughly constant negative biases (i.e., CGM estimates are higher) in the Levantine basin, and positive biases in the Ionian Sea and the Western basin, generally not exceeding 10÷12 cm, and an increasingly more pessimistic regional projection for the Balearic region, going from a deviation of 15÷17cm at mid-century, to one of 17÷22 cm in 2100.\u003c/p\u003e \u003c/div\u003e "},{"header":"6. Conclusions","content":" \u003cp\u003eWe have here analysed the results of three climatic simulations of the Mediterranean Sea circulation and sea level, performed with MED16, an updated version of the tide-including, climatic ocean model presented in Sannino et al., (\u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). These consist of a hindcast simulation, covering three recent decades (1981\u0026ndash;2010), a historical run (1981\u0026ndash;2005), and a future climate simulation, under the RCP8.5 scenario, which starts from the end of the historical run and proceeds until the end of the present century (2006\u0026ndash;2100).\u003c/p\u003e \u003cp\u003eThe hindcast simulation has been used to assess the ability of the new model to reproduce the past evolution of the basin circulation, with excellent results. This simulation has a value in itself, because it is the first pluri-decadal, eddy-resolving simulation of the Mediterranean Sea circulation that includes the main tidal effects, together with an accurate treatment of the complex dynamics occurring in the Gibraltar Strait area. We have therefore preliminarily analysed the simulated transports at the Gibraltar, Dardanelles and Bosphorus straits, and the basin hydrology and circulation. We found that the model correctly represents the transport evolution at Gibraltar, the average circulation in the surface and intermediate layers, and the main variability of the Mediterranean hydrologic structure during the simulation period. The simulated SLA was found in good agreement with satellite observations, exhibiting a positive trend in the Western Mediterranean Sea, and a more complex behavior in the Levantine basin, where sea level variability has been strongly influenced by the EMT and its evolution.\u003c/p\u003e \u003cp\u003eThe spatial structure of the hindcast average SSH field is well reproduced over the entire basin, especially in the western Mediterranean where results are much closer to that of the MDT, due to the fact that the use of improved sea level information at the Atlantic lateral boundary significantly enhances the reliability of results. A further improvement has been obtained through the adequate treatment of the complex, hydraulically driven dynamics across the Gibraltar Strait.The evolution of the hindcast SSH near coast compares very well with tide gauge observation along most of the basin\u0026rsquo;s coasts. This shows that MED16, if adequately forced, can provide useful information on the sea level variability in coastal areas, which are the more exposed to SLR, even though the current horizontal resolution of the model does not allow to exactly resolve the locations of the tide gauge stations. This also provides an indirect check of the general quality of the atmospheric forcing from the SMHI downscaling. This suggests that further improvement of the forcing may be needed to obtain an even more realistic description of the circulation structure.\u003c/p\u003e \u003cp\u003eThe historical simulation, which sets the reference baseline for the future scenario projections, also correctly reproduces the main mean features of the Mediterranean circulation and hydrology, a result that is all the more satisfactory when considering that no specific constraint or relaxation is prescribed.\u003c/p\u003e \u003cp\u003eUnder the RCP8.5 future scenario, the temperature is projected to generally increase. The spatial pattern of variations in surface salinity is affected by the penetration of the Atlantic stream into the basin interior, with the westernmost regions exhibiting appreciable decreases. The warming of sea waters results in the inhibition or enhanced convection variability in the main deep-water formation sites.\u003c/p\u003e \u003cp\u003eAs regards future sea level, the overall DSL trend estimated by MED16 is comparable to the ensemble mean computed from the Global Circulation Models analysed in Slangen at al. (2014). Nevertheless, our regional dynamical downscaling provides a better characterization of regional patterns arising from ocean dynamics and of their variability, and gives a more pronounced relative sinking of the Mediterranean with respect to the Atlantic. The regional differences in SLR between the presented downscaling and the GCM mean projection are generally constant in time, yet an acceleration emerges in the regional projection for the Balearic area.\u003c/p\u003e \u003cp\u003eIndeed, state-of-the-art numerical estimates show that future sea level rise in the Mediterranean basin will crucially depend on the evolution of the sterodynamic component, as determined by the propagation of the Atlantic signal through the Strait of Gibraltar and by the local circulation. MED16 effectively provides an accurate and reliable representation of the basin\u0026rsquo;s DSL, by accounting for the numerous and complex processes that concur to determine it. Nevertheless, in view of the significant numerical resources it demands, the assessment is recommended of the minimum requirements MED16, or any other similar model, must comply with, in terms of resolution and explicit process representation. In accordance with previous works, we demonstrate that both explicit tidal forcing and an accurate resolution of the Gibraltar Strait are key features that cannot be disregarded in designing numerical tools suitable for the DSL simulation in the Mediterranean Sea.\u003c/p\u003e "},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eConceived the work [GS]\u003c/p\u003e\n\u003cp\u003eImplemented the model and performed the run [GS, AC]\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eCollected data [MP, EN, RI, AC]\u003c/p\u003e\n\u003cp\u003eDesigned the analysis [GS, RI, AC, EN, GP, MVS]\u003c/p\u003e\n\u003cp\u003ePerformed the analysis [RI, AC, MP, EN]\u003c/p\u003e\n\u003cp\u003eWrote the paper [ALL]\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eRegional sea level data from IPCC AR5 distributed in netCDF format by the Integrated Climate Data Center (ICDC, icdc.cen.uni-hamburg.de) University of Hamburg, Hamburg, Germany. SROCC sea-level rise data are available at https://www.ipcc.ch/srocc/download-report/.\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eAblain M, Cazenave A, Larnicol G, Balmaseda M, Cipollini P, Faug\u0026egrave;re Y, Fernandes MJ, Henry O, Johannessen JA, Knudsen P, Andersen O, Legeais J, Meyssignac B, Picot N, Roca M, Rudenko S, Scharffenberg MG, Stammer D, Timms G, and Benveniste J (2015) Improved sea level record over the satellite altimetry era (1993\u0026ndash;2010) from the Climate Change Initiative project. Ocean Sci. 11:67\u0026ndash;82. https://doi.org/10.5194/os-11-67-2015, 2015.\u003c/li\u003e\n \u003cli\u003eAdcroft A, Campin JM (2004) Rescaled height coordinates for accurate representation of free-surface flows in ocean circulation models. Ocean Model 7(34):269\u0026ndash;284\u003c/li\u003e\n \u003cli\u003eAdloff F, Jord\u0026agrave; G, Somot S, Sevault F, Arsouze T, Meyssignac B, Li L, and Planton S (2018) Improving sea level simulation in Mediterranean regional climate models. Clim Dyn, 51:1167\u0026ndash;1178\u003c/li\u003e\n \u003cli\u003eAmitai Y, Lehahn Y, Lazar A, Heifetz E (2010) Surface circulation of the eastern Mediterranean Levantine basin: Insights from analyzing 14 years of satellite altimetry data. J. Geophys. Res. 115, C10058.\u003c/li\u003e\n \u003cli\u003eAntonioli F et al (2017) Sea-level rise and potential drowning of the Italian coastal plains. Flooding risk scenarios for 2100. Quaternary Science Review 158:29\u0026ndash;43, doi: 10.1016/j.quascirev.2016.12.021.\u003c/li\u003e\n \u003cli\u003eArnone RA, Wiesenburg DA, Saunders KD (1990) The origin and characteristics of the algerian current. Journal of Geophysical Research 95 (C2):1587\u0026ndash;1598. http://dx.doi.org/10.1029/JC095iC02p01587.\u003c/li\u003e\n \u003cli\u003eBalmaseda MA, Mogensen K, Weaver AT (2013) Evaluation of the ECMWF ocean reanalysis system oras4. QJR Meteorol Soc 139(674):1132\u0026ndash;1161.\u003c/li\u003e\n \u003cli\u003eBignami F, Marullo S, Santoleri R, and Schiano ME (1995) Longwave radiation budget in the Mediterranean Sea. J. Geophys. Res 100:2501\u0026ndash;2514.\u003c/li\u003e\n \u003cli\u003eBonaduce A, Pinardi N, Oddo P, Spada G, and Larnicol G (2016) Sea-level variability in the Mediterranean Sea from altimetry and tide gauges. Climate Dynamics 47(9\u0026ndash;10): 2851\u0026ndash;2866.\u003c/li\u003e\n \u003cli\u003eBougeault P, and Lacarr\u0026egrave;re P (1989) Parameterization of Orography-Induced Turbulence in a Mesobeta-Scale Model. Monthly Weather Review 117:1872-1890. http://dx.doi.org/10.1175/1520-0493(1989)117\u003c/li\u003e\n \u003cli\u003eBoutov D, Peliz A, Miranda PMA, Soares PMM, Cardoso RM (2014) Inter-annual variability and long term predictability of exchanges through the Strait of Gibraltar. Glob Planet Change 114:23\u0026ndash;37. doi:10.1016/j.gloplacha.2013.12.009\u003c/li\u003e\n \u003cli\u003eCalafat FM, Jordà G (2011) A mediterranean sea level reconstruction (1950\u0026ndash;2008) with error budget estimates. Glob. Planet. Change 79:118\u0026ndash;133.\u003c/li\u003e\n \u003cli\u003eCampin JM, Adcroft A, Hill C, Marshall J (2004) Conservation of prop- erties in a free-surface model. Ocean Model 6:221\u0026ndash;244.\u003c/li\u003e\n \u003cli\u003eCandela J (2001) Mediterranean water and global circulation. In: Siedler G, Church J, Gould J (eds) Ocean circulation and climate. Academic, San Diego, pp 419\u0026ndash;429\u003c/li\u003e\n \u003cli\u003eChurch JA, Clark PU, Cazenave A, Gregory J, et al (2013a) Sea level change, in Climate Change 2013: The Physical Science Basis. Stocker TF, Qin D, Plattner GK, et al (ed) Cambridge University Press, Cambridge, UK and New York, NY. USA.\u003c/li\u003e\n \u003cli\u003eChurch JA, Clark PU, Cazenave A, Gregory J, et al (2013b) Sea Level Change Supplementary Material. In: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Available from www.climatechange2013.org and www.ipcc.ch.\u003c/li\u003e\n \u003cli\u003eCiuffardi T, Napolitano E, Iacono R, Reseghetti F, Raiteri G, Bordone A (2016) Analysis of surface circulation structures along a frequently repeated XBT transect crossing the Ligurian and Tyrrhenian seas. Ocean Dyn. 66 (6\u0026ndash;7): 767\u0026ndash;783.\u003c/li\u003e\n \u003cli\u003eCollins WJ, Bellouin N, Doutriaux-Boucher M, Gedney N, Halloran P, Hinton T, Hughes J, Jones CD et al (2011) Development and evaluation of an Earth-System model \u0026ndash; HadGEM2. Geoscient Model Dev 4:1051\u0026ndash;1075. https:// doi.org/10.5194/gmd-4-1051-2011\u003c/li\u003e\n \u003cli\u003eDangendorf S, Hay C, Calafat FM, Marcos M, Piecuch CG, Berk K, and Jensen J (2019) Persistent acceleration in global sea-level rise since the 1960s. Nature Climate Change 9:705-710\u003c/li\u003e\n \u003cli\u003eDee D, Uppala S, Simmons A, Berrisford P, Poli P, Kobayashi S et al (2011) The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Q. J. R. Meteorol. Soc. 137:553\u0026ndash;597. doi: 10.1002/qj.828\u003c/li\u003e\n \u003cli\u003eEgbert GD, Erofeeva SY (2002) Efficient inverse modeling of barotropic ocean tides. Journal of Atmospheric and Oceanic Technology 19:183\u0026ndash;204. doi:10.1175/1520-0426(2002)019\u0026iexcl;0183:EIMOBO\u0026iquest;2.0.CO;2.\u003c/li\u003e\n \u003cli\u003eEscudier R, Clementi E, Omar M et al (2020) Mediterranean Sea Physical Reanalysis (CMEMS MED-Currents) (Version 1).Copernicus Monitoring Environment Marine Service (CMEMS). https://doi.org/10.25423/CMCC/MEDSEA_MULTIYEAR_PHY_006_004_E3R1\u003c/li\u003e\n \u003cli\u003eFrederikse T, Landerer F, Caron L, Adhikari S et al (2020) The causes of sea-level rise since 1900. Nature 584:393\u0026mdash;397\u003c/li\u003e\n \u003cli\u003eFox-Kemper B, Menemenlis D (2008) Can large eddy simulation techniques improve mesoscale rich ocean models? Geophys Monogr:319\u0026ndash;337\u003c/li\u003e\n \u003cli\u003eGarcía-Lafuente J, Alvarez-Fanjul E, Vargas J, Ratsimandresy A (2002) Subinertial variability through the Strait of Gibraltar. J Geophys Res 107(C10):3168. doi:10.1029/2001JC001104\u003c/li\u003e\n \u003cli\u003eGarcía-Lafuente J, Sammartino S, Huertas IE, Flecha S, Sánchez-Leal RF, Naranjo C, Nadal I and Bellanco MJ (2021) Hotter and Weaker Mediterranean Outflow as a Response to Basin-Wide Alterations. Front. Mar. Sci. 8:613444. doi: 10.3389/fmars.2021.613444\u003c/li\u003e\n \u003cli\u003eGaspar P, Gregoris Y, Lefevre JM (1990) A simple eddy kinetic energy model for simulations of the oceanic vertical mixing: Tests at station Papa and long- term upper ocean study site. Journal of Geophysical Research 95 16179. doi:10.1029/jc095ic09p16179.\u003c/li\u003e\n \u003cli\u003eGill AE, and Niller PP (1973) The theory of the seasonal variability in the ocean. Deep-Sea Res. Oceanogr. Abstr. 20: 141\u0026ndash; 177, doi:10.1016/0011-7471(73)90049-1.\u003c/li\u003e\n \u003cli\u003eGiorgi F, Jones C, Asrar GR (2009) Addressing climate information needs at the regional level: the CORDEX framework. WMO Bull 58(3)\u003c/li\u003e\n \u003cli\u003eGökaşan E, Tur H, Ecevitoğlu B, Görüm T, Türker A, Tok B, Çağlak F, Birkan H, Şimşek M (2005) Evidence and implications of massive erosion along the Strait of İstanbul (Bosphorus). Geo-Mar Lett 25: 324\u0026ndash;342\u003c/li\u003e\n \u003cli\u003eGökaşan E, Ergin M et al (2007) Factors controlling the morphological evolution of the Çanakkale Strait (Dardanelles, Turkey). Geo-Mar Lett 28:107\u0026ndash;129\u003c/li\u003e\n \u003cli\u003eGreatbatch RJ (1994) A note on the representation of steric sea level in models that conserve volume rather than mass. J. Geophys. Res. 99(C6):12767\u0026ndash;12771.\u003c/li\u003e\n \u003cli\u003eGregory JM, Griffies SM, Hughes CW et al (2019) Concepts and terminology for sea level: mean, variability and change, both local and global. Surv Geophys 40:1251\u0026ndash;1289 https://doi.org/10.1007/s10712-019-09525-z\u003c/li\u003e\n \u003cli\u003eGriffies SM, Greatbatch RJ (2012) Physical processes that impact the evolution of global mean sea level in ocean climate models. Ocean Modell. 51:37\u0026ndash;72, doi:10.1016/ j.ocemod.2012.04.003\u003c/li\u003e\n \u003cli\u003eHamlington BD, Thompson P, Hammond WC, Blewitt G, and Ray RD (2016) Assessing the impact of vertical land motion on twentieth century global mean sea level estimates, J. Geophys. Res. Oceans 121:4980\u0026ndash; 4993, doi:10.1002/2016JC011747.\u003c/li\u003e\n \u003cli\u003eHermans THJ, Tinker J, Palmer MD et al (2020) Improving sea-level projections on the Northwestern European shelf using dynamical downscaling. Clim Dyn 54:1987\u0026ndash;2011. https://doi.org/10.1007/s00382-019-05104-5\u003c/li\u003e\n \u003cli\u003eHolgate SJ, Matthews A, Woodworth PL, Rickards LJ, Tamisiea ME, Bradshaw E, Foden PR, Gordon KM, Jevrejeva S, and Pugh J (2013) New data systems and products at the Permanent Service for Mean Sea Level. Journal of Coastal Research 29(3):493\u0026ndash;504 (DOI: 10.2112/JCOASTRES-D-12-00175.1).\u003c/li\u003e\n \u003cli\u003eIacono R, Napolitano E (2020) Aspects of the summer circulation in the eastern Ligurian Sea, Deep Sea Research Part I: Oceanographic Research Papers, Volume 166, 103407, ISSN 0967-0637, https://doi.org/10.1016/j.dsr.2020.103407\u003c/li\u003e\n \u003cli\u003eIPCC (2013) Summary for Policymakers. In: Climate Change 2013: The Physical Science Basis. Contribution of Working Group II to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press: Cambridge and New York, USA and UK, 2014, available online: https://www.ipcc.ch/site/assets/uploads/2018/02/WG1AR5_all_final.pdf.\u003c/li\u003e\n \u003cli\u003eIPCC (2019) Technical Summary [H.-O. Pörtner, D.C. Roberts, V. Masson-Delmotte, P. Zhai, E. Poloczanska, K. Mintenbeck, M. Tignor, A. Alegría, M. Nicolai, A. Okem, J. Petzold, B. Rama, N.M. Weyer (eds.)]. In: \u003cem\u003eIPCC Special Report on the Ocean and Cryosphere in a Changing Climate\u0026nbsp;\u003c/em\u003e[H.- O. Pörtner, D.C. Roberts, V. Masson-Delmotte, P. Zhai, M. Tignor, E. Poloczanska, K. Mintenbeck, A. Alegría, M. Nicolai, A. Okem, J. Petzold, B. Rama, N.M. Weyer (eds.)], https://www.ipcc.ch/srocc/\u003c/li\u003e\n \u003cli\u003eJeftic L, Milliman JD and Sestini, G (1992) Climatic change and the Mediterranean: environmental and societal impacts of climatic change and sea-level rise in the Mediterranean region. United Nations Environment Programme, http://hdl.handle.net/20.500.11822/29285\u003c/li\u003e\n \u003cli\u003eJord\u0026agrave; G, Gomis D (2013) On the interpretation of the steric and mass components of sea level variability: The case of the Mediterranean basin. J. Geophys. Res. Oceans 118:953\u0026ndash; 963, doi:10.1002/jgrc.20060.\u003c/li\u003e\n \u003cli\u003eJord\u0026agrave; G, Von Schuckmann K, Josey SA, Caniaux G et al (2017a) The Mediterranean Sea heat and mass budgets: Estimates, uncertainties and perspectives. Progress in Oceanography 156:174\u0026ndash;208. doi:10.1016/j.pocean.2017.07.001\u003c/li\u003e\n \u003cli\u003eJord\u0026agrave; G, S\u0026aacute;nchez-Rom\u0026aacute;n A, Gomis D (2017b) Reconstruction of transports through the Strait of Gibraltar from limited observations. Clim Dyn 48:851\u0026ndash;865, https://doi.org/10.1007/s00382-016-3113-8\u003c/li\u003e\n \u003cli\u003eKara AB, Wallcraft AJ, Hulbert HE, Stanev EV (2008) Air‐sea fluxes and river discharges in the Black Sea with a focus on the Danube and Bosphorus. J. Mar. Syst. 74:74\u0026ndash;95.\u003c/li\u003e\n \u003cli\u003eKing MD, Howat IM, Candela SG, Noh MJ et al (2020) Dynamic ice loss from the Greenland Ice Sheet driven by sustained glacier retreat. Commun Earth Environ 1, 1. https://doi.org/10.1038/s43247-020-0001-2\u003c/li\u003e\n \u003cli\u003eKlein B, Roether W, Manca BB, Bregant D, Beitzel V, Kovacevic V, Luchetta A (1999) The large deep water transient in the Eastern Mediterranean. Deep Sea Research Part I 46: 371-414.\u003c/li\u003e\n \u003cli\u003eKnysh, VV, Demyshev SG, Korotaev GK (2002) Method of reconstructing the climatic seasonal circulation of the Black Sea on the basis of assimilation of hydrologic data in models. Marine Hydrophys. J. 2:36\u0026ndash;52, (in Russian).\u003c/li\u003e\n \u003cli\u003eKubryakov AA, Stanichny SV (2011) Mean Dynamic Topography of the Black Sea, computed from altimetry, drifter measurements and hydrology data. Ocean Sci. 7:745\u0026ndash;753. https://doi.org/10.5194/os-7-745-2011\u003c/li\u003e\n \u003cli\u003eLambeck K, Smither C, Johnston P (1998) Sea-level change, glacial rebound and mantle viscosity for northern Europe. Geophys. J. Int. 134:102\u0026ndash;144.\u003c/li\u003e\n \u003cli\u003eLambeck K, Yokoyama Y, Purcell A, Johnston P (2002) Reply to the comment by WR Peltier. Quaternary Science Reviews, 21:415-418.\u003c/li\u003e\n \u003cli\u003eLanderer FW, Volkov DL (2013) The anatomy of recent large sea level fluctuations in the Mediterranean Sea. Geophys. Res. Lett. 40:553\u0026ndash;557\u003c/li\u003e\n \u003cli\u003eLarge WG, Yeager SG (2004) Diurnal to decadal global forcing for ocean and sea-ice models: the data sets and flux climatologies. National Center for Atmospheric Research Boulder. http://dx.doi.org/10.5065/D6KK98Q6\u003c/li\u003e\n \u003cli\u003eLeith C (1968) Parameterization of vertical mixing in numerical models of tropical oceans. Phys Fluids 10:1409\u0026ndash;1416\u003c/li\u003e\n \u003cli\u003eLlasses J, Jordà G, Gomis D, Adloff F et al (2018) Heat and salt redistribution within the Mediterranean Sea in the Med‐CORDEX model ensemble. Clim. Dyn. 51:1119-1143.\u003c/li\u003e\n \u003cli\u003eLosch M, Adcroft A, Campin JM (2004) How sensitive are coarse general circulation models to fundamental approximations in the equations of motion? J. Phys. Oceanogr., 34:306\u0026ndash;319.\u003c/li\u003e\n \u003cli\u003eMarullo S, Buongiorno Nardelli B, Guarracino M, Santoleri R (2007) Observing the Mediterranean Sea from space: 21 years of Pathfinder-AVHRR sea surface temperatures (1985 to 2005): re-analysis and validation. Ocean Sci. 3: 299-310\u003c/li\u003e\n \u003cli\u003eMeinshausen M, Smith SJ, Calvin K, Daniel JS, Kainuma MLT, Lamarque JF, Matsumoto K, Montzka SA, Raper SCB, Riahi K, Thomson A, Velders GJM, van Vuuren DPP (2011) The RCP greenhouse gas concentrations and their extensions from 1765 to 2300. Clim Change 109:210\u0026ndash;241. https://doi.org/10.1007/s1058 4-011-0156-z\u003c/li\u003e\n \u003cli\u003eMarcos M, Tsimplis MN (2008) Coastal sea level trends in Southern Europe. Geophys. J. Int. 175:70\u0026ndash;82.\u003c/li\u003e\n \u003cli\u003eMarshall J, Hill C, Perelman L, Adcroft A (1997a) Hydrostatic, quasi- hydrostatic, and nonhydrostatic ocean modeling. J Geophys Res 102:5733\u0026ndash;5752\u003c/li\u003e\n \u003cli\u003eMarshall J, Adcroft A, Hill C, Perelman L, Heisey C (1997b) A finite- volume, incompressible Navier Stokes model for studies of the ocean on parallel computers. J Geophys Res 102:5753\u0026ndash;5766\u003c/li\u003e\n \u003cli\u003eMeyssignac B, Calafat FM, Somot S, Rupolo V, Stocchi P, Llovel W, Cazenave A (2011) Two-dimensional reconstruction of the mediterranean sea level over 1970\u0026ndash;2006 from tide gage data and regional ocean circulation model outputs. Glob. Planet. Change 77(1\u0026ndash;2): 49\u0026ndash;61.\u003c/li\u003e\n \u003cli\u003eMohamed B, Mohamed A et al (2019) Inter-Annual Variability and Trends of Sea Level and Sea Surface Temperature in the Mediterranean Sea over the Last 25 Years. Pure Appl. Geophys. 176:3787\u0026ndash;3810.\u003c/li\u003e\n \u003cli\u003eNaranjo C, Garcia-Lafuente J, Sannino G, Sanchez-Garrido JC (2014) How much do tides affect the circulation of the Mediterranean Sea? From local processes in the Strait of Gibraltar to basin-scale effects, Progress in Oceanography 127:108-116 /doi.org/10.1016/j.pocean.2014.06.005.\u003c/li\u003e\n \u003cli\u003ePalma M, Iacono R, Sannino G, Bargagli A, Carillo A, Fekete BM, Lombardi E, Napolitano E, Pisacane G, Struglia MV (2020) Short-term, linear and non-linear local effects of the tides on the surface dynamics in a new, high-resolution model of the Mediterranean Sea circulation. Ocean Dyn. 70:935-963 https://doi.org/10.1007/s10236-020-01364-6.\u003c/li\u003e\n \u003cli\u003ePeltier WR (2002) Comments on the paper of Yokoyama et al. (2000) entitled \u0026lsquo;\u0026lsquo;Timing of the Last Glacial Maximum from observed sea level minima\u0026rsquo;\u0026rsquo;. Quaternary Science Reviews 21:409-414\u003c/li\u003e\n \u003cli\u003ePeltier WR (2004) Global glacial isostasy and the surface of the ice-age earth: The ICE-5G (VM2) model and GRACE. Annu. Rev. Earth Planet. Sci. 32:111\u0026ndash;149 doi:10.1146/annurev.earth.32.082503.144359.\u003c/li\u003e\n \u003cli\u003ePeneva E, Stanev E, Belokopytov V, Le Traon PY (2001) Water transport in the Bosphorus Strait estimated from hydro‐meteorological and altimeter data: Seasonal and decadal variability. J. Mar. Syst. 31:21\u0026ndash;33.\u003c/li\u003e\n \u003cli\u003ePinardi N, Zavatarelli M, Adani M, Coppini G, Fratianni C, Oddo P, Simoncelli S, Tonani M, Lyubartsev V, Dobricic S, Bonaduce A (2015) Mediterranean Sea large-scale low-frequency ocean variability and water mass formation rates from 1987 to 2007: a retrospective analysis. Prog Oceanogr., 132: 318-332. https://doi.org/10.1016/j.pocean.2013.11.003.\u003c/li\u003e\n \u003cli\u003ePonce VM (1994) Engineering hydrology: principles and practices. Prentice Hall\u003c/li\u003e\n \u003cli\u003ePoulain PM, Mauri E, Gerin R, Chiggiato J et al (2020) On the dynamics in the southeastern Ligurian Sea in summer 2010. Cont. Shelf Res. 196:104083\u003c/li\u003e\n \u003cli\u003eRahmstorf S (2007) A semi-empirical approach to projecting future sea-level rise, Science 315(8010): 368-370, doi: 10.1126/science.1135456.\u003c/li\u003e\n \u003cli\u003eRio MH, Pascual A, Poulain PM, Menna M, Barcelo\u0026rsquo; B, Tintor\u0026eacute; J (2014) Computation of a new mean dynamic topography for the Mediterranean Sea from model outputs, altimeter measurements and oceanographic in situ data. Ocean Sci. 10:731-744.\u003c/li\u003e\n \u003cli\u003eRoether W, Manca BB, Klein B et al (1996) Recent changes in eastern Mediterranean deep waters. Science 271: 333\u0026ndash;335. doi:10.1126/science.271.5247.333.\u003c/li\u003e\n \u003cli\u003eS\u0026aacute;nchez Rom\u0026aacute;n A, Sannino G, Garc\u0026iacute;a Lafuente J et al (2009) Transport estimates at the western section of the Strait of Gibraltar: A combined experimental and numerical modeling study. Journal of Geophysical Research: Oceans 114, C06002. https://doi.org/10.1029/2008JC005023.\u003c/li\u003e\n \u003cli\u003eSannino G, Sözer A, Özsoy E (2017) A high-resolution modelling study of the Turkish Straits System. Ocean Dynamics 67(3):397\u0026ndash;432\u003c/li\u003e\n \u003cli\u003eSannino G, Carillo A, Pisacane G, Naranjo C (2015) On the relevance of tidal forcing in modelling the Mediterranean thermohaline circulation. Prog Oceanogr 134:304\u0026ndash;329\u003c/li\u003e\n \u003cli\u003eSannino G, Herrmann M, Carillo A, Rupolo V, Ruggiero V, Artale V, Heimbach P (2009) An eddy-permitting model of the Mediterranean sea with a two-way grid refinement at the Strait of Gibraltar. Ocean Modelling 30 (1): 56\u0026ndash;72\u003c/li\u003e\n \u003cli\u003eSannino G, Carillo A, Pratt L (2009) Hydraulic criticality of the exchange flow through the strait of Gibraltar. Journal of Physical Oceanography 39:2779\u0026ndash;2799. doi:10.1175/2009JPO4075.1\u003c/li\u003e\n \u003cli\u003eSantamar\u0026iacute;a-G\u0026oacute;mez, A, Gravelle M, Dangendorf S, Marcos M, Spada G, W\u0026ouml;ppelmann G (2017) Uncertainty of the 20th century sea-level rise due to vertical land motion errors. Earth and Planetary Science Letters 473:24-32. https://doi.org/10.1016/j.epsl.2017.05.038.\u003c/li\u003e\n \u003cli\u003eSanz JL, Acosta J, Esteras M, Herranz P, Palomo C, Sandoval N (1991) Prospección geofísica del Estrecho de Gibraltar: resultados del Programa Hércules (1980\u0026ndash;1983). Publ. espec. Inst. Esp. Oceanogr. Span. Inst. of Oceanogr., Madrid\u003c/li\u003e\n \u003cli\u003eSen Gupta A, Jourdain NC, Brown JN, Monselesan D (2013) Climate drift in the CMIP5 models. J. Climate 26:8597\u0026ndash; 8615. doi:10.1175/JCLI-D-12-00521.1.\u003c/li\u003e\n \u003cli\u003eShugar DH, Burr A, Haritashya UK et al (2020) Rapid worldwide growth of glacial lakes since 1990. Nature Climate Change 10:939-945. https://doi.org/10.1038/s41558-020-0855-4\u003c/li\u003e\n \u003cli\u003eSimonov AI, Altman EN (1991) Hydrometeorology and hydrochemistry of the USSR seas, in Sensitivity to Change: Black Sea, Baltic Sea and North Sea. NATO/ASI Ser., vol. IV, edited by E. \u0026Ouml;zsoy and A. Mikaelyan, pp. 11\u0026ndash;25, Kluwer Acad., Dordrecht, Netherlands.\u003c/li\u003e\n \u003cli\u003eSlangen ABA, Adloff F, Jevrejeva S et al (2017a) Review of Recent Updates of Sea-Level Projections at Global and Regional Scales. Surv. Geophys., 38:385\u0026ndash;406.\u003c/li\u003e\n \u003cli\u003eSlangen ABA, Carson M, Katsman CA et al (2014) Projecting twenty-first century regional sea-level changes. Climatic Change 124:317\u0026ndash;332. https://doi.org/10.1007/s10584-014-1080-9\u003c/li\u003e\n \u003cli\u003eSmagorinsky J (1963) General circulation experiments with the primitive equations. Mon Weather Rev 91:99\u0026ndash;164\u003c/li\u003e\n \u003cli\u003eSoto-Navarro J, Criado‐Aldeanueva F, Lafuente JG, Sanchez-Roman A (2010) Estimation of the Atlantic inflow through the Strait of Gibraltar from climatological and in situ data. J. Geophys. Res. 115. https://doi.org/10.1029/2010JC006302\u003c/li\u003e\n \u003cli\u003eSoto-Navarro J, Jord\u0026aacute;, G, Amores A et al (2020) Evolution of Mediterranean Sea water properties under climate change scenarios in the Med-CORDEX ensemble. Clim Dyn 54:2135\u0026ndash;2165. https://doi.org/10.1007/s00382-019-05105-4\u003c/li\u003e\n \u003cli\u003eSpada G, Stocchi P (2006) The Sea Level Equation, Theory and Numerical Examples, Aracne, Roma, p. 96, ISBN: 88\u0026ndash;548\u0026ndash;0384\u0026ndash;7.\u003c/li\u003e\n \u003cli\u003eSpada G, Stocchi P (2007) SELEN: a Fortran 90 program for solving the \u0026lsquo;Sea Level Equation\u0026rsquo;. Comput. Geosci. 33(4): 538\u0026ndash;562. doi:10.1016/j.cageo.2006.08.006.\u003c/li\u003e\n \u003cli\u003eSpada G (2017) Glacial Isostatic Adjustment and Contemporary Sea Level Rise: An Overview. Surv Geophys 38:153\u0026ndash;185. https://doi.org/10.1007/s10712-016-9379-x\u003c/li\u003e\n \u003cli\u003eStrandberg G, B\u0026auml;rring L, Hansson U, Jansson C, Jones C (2014) CORDEX scenarios for Europe from the Rossby Centre regional climate model RCA4.\u003c/li\u003e\n \u003cli\u003eTheocharis A, Klein B, Nittis K, Roether W (2002) Evolution and status of the Eastern Mediterranean Transient (1997\u0026ndash;1999). J. Mar. Syst. 33-34: 91-116.\u003c/li\u003e\n \u003cli\u003eThi\u0026eacute;blemont R, Le Cozannet G, Toimil A, Meyssignac B, Losada IJ (2019) Likely and High-End Impacts of Regional Sea-Level Rise on the Shoreline Change of European Sandy Coasts Under a High Greenhouse Gas Emissions Scenario. Water 2019, 11, 2607.\u003c/li\u003e\n \u003cli\u003eTsimplis MN, Josey SA (2001) Forcing of the Mediterranean Sea by atmospheric oscillations over the North Atlantic. Geophysical Research Letters 28:803-806. https://doi.org/10.1029/2000GL012098\u003c/li\u003e\n \u003cli\u003eTsimplis MN, Bryden HL (2000) Estimation of the transports through the strait of Gibraltar. Deep-Sea Res I 47:2219\u0026ndash;2242\u003c/li\u003e\n \u003cli\u003e\u0026Uuml;nl\u0026uuml;ata \u0026Uuml;, Oğuz T, Latif MA, \u0026Ouml;zsoy E (1990) On the physical oceanography of the Turkish Straits, in The Physical Oceanography of Sea Straits. NATO/ASI Ser., edited by L. J. Pratt, pp. 25\u0026ndash;60, Kluwer Acad., Dordrecht, Netherlands.\u003c/li\u003e\n \u003cli\u003eVörösmarty CJ, Federer CA, Schloss AL (1998) Potential evaporation functions compared on US watersheds: implications for global-scale water balance and terrestrial ecosystem modeling. J Hydrol 207: 147\u0026ndash;169\u003c/li\u003e\n \u003cli\u003eVousdoukas MI, Mentaschi L, Voukouvalas E, Verlaan M, Feyen L (2017) Extreme sea levels on the rise along Europe\u0026rsquo;s coasts. Earth\u0026rsquo;s Future 5:304-323. https://doi.org/10.1002/2016EF000505\u003c/li\u003e\n \u003cli\u003eWisser D, Frolking S, Douglas EM, Fekete BM, Vörösmarty CJ, Schumann AH (2008) Global irrigation water demand: variability and uncertainties arising from agricultural and climate data sets. Geophys Res Lett 35:L24408\u003c/li\u003e\n \u003cli\u003eWisser D, Fekete BM, Vörösmarty CJ, Schumann AH (2010) Reconstructing 20th century global hydrography: a contribution to the Global Terrestrial Network- Hydrology (GTN-H). Hydrol Earth Syst Sci 14:1\u0026ndash;24\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"climate-dynamics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"cldy","sideBox":"Learn more about [Climate Dynamics](https://www.springer.com/journal/382)","snPcode":"382","submissionUrl":"https://submission.nature.com/new-submission/382/3","title":"Climate Dynamics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Mediterranean Sea climate, MED16, altimetric observations, RCP8.5","lastPublishedDoi":"10.21203/rs.3.rs-653703/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-653703/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eWe present results of three simulations of the Mediterranean Sea climate: a hindcast, a historical run, and a RCP8.5 scenario simulation reaching the year 2100. The simulations are performed with MED16, a new, tide-including implementation of the MITgcm model, which covers the Mediterranean - Black Sea system with a resolution of 1/16°, further increased at the Gibraltar and Turkish Straits. \u003c/p\u003e\u003cp\u003eValidation of the hindcast simulation against observations and numerical reanalyses has given excellent results, proving that the model is also capable of reproducing near-shore sea level variations. Moreover, the spatial structure of the elevation field compares well with altimetric observations, especially in the Western basin, due to the use of improved sea level information at the Atlantic lateral boundary and to the adequate treatment of the complex, hydraulically driven dynamics across the Gibraltar Strait.\u003c/p\u003e\u003cp\u003eUnder the RCP8.5 future scenario, the temperature is projected to generally increase while the surface salinity decreases in the portion of the Mediterranean affected by the penetration of the Atlantic stream, and increases elsewhere. The warming of sea waters results in the partial inhibition of deep-water formation.\u003c/p\u003e\u003cp\u003eThe scenario simulation allows for a detailed characterization of the regional patterns of future sea level, arising from ocean dynamics, and indicates a relative sinking of the Mediterranean with respect to the Atlantic more pronounced than the current one. Explicit tidal forcing and an accurate resolution of the Gibraltar Strait are proved to be key features in the designing of numerical simulations for the Mediterranean Sea.\u003c/p\u003e","manuscriptTitle":"Modelling Present and Future Climate in the Mediterranean Sea: A Focus on Sea-Level Change","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2021-07-01 21:57:37","doi":"10.21203/rs.3.rs-653703/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Major Revision","date":"2021-09-27T06:21:36+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2021-08-16T08:20:12+00:00","index":0,"fulltext":""},{"type":"reviewersInvited","content":"","date":"2021-06-28T00:00:00+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2021-06-24T00:00:00+00:00","index":"","fulltext":""},{"type":"submitted","content":"Climate Dynamics","date":"2021-06-23T12:04:56+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"climate-dynamics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"cldy","sideBox":"Learn more about [Climate Dynamics](https://www.springer.com/journal/382)","snPcode":"382","submissionUrl":"https://submission.nature.com/new-submission/382/3","title":"Climate Dynamics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"75db0625-ead7-4ad1-8c6d-b3680f9a3d49","owner":[],"postedDate":"July 1st, 2021","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":5411649,"name":"Climatology"},{"id":5411650,"name":"Climate Analysis and Modeling"}],"tags":[],"updatedAt":"2022-01-21T06:17:26+00:00","versionOfRecord":{"articleIdentity":"rs-653703","link":"https://doi.org/10.1007/s00382-021-06132-w","journal":{"identity":"climate-dynamics","isVorOnly":false,"title":"Climate Dynamics"},"publishedOn":"2022-01-21 06:17:26","publishedOnDateReadable":"January 21st, 2022"},"versionCreatedAt":"2021-07-01 21:57:37","video":"","vorDoi":"10.1007/s00382-021-06132-w","vorDoiUrl":"https://doi.org/10.1007/s00382-021-06132-w","workflowStages":[]},"version":"v1","identity":"rs-653703","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-653703","identity":"rs-653703","version":["v1"]},"buildId":"WrCJVZZCHTDjtuVLN7oU0","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}