Assessment of ocean models in the Mediterranean Sea and Black Sea against altimetry and gravity measurements
Event: 2017 Ocean Surface Topography Science Team Meeting
Session: Science I: Climate data records for understanding the causes of global and regional sea level variability and change
Presentation type: Type Poster
Contribution: not provided
We compare in the Mediterranean and Black Sea the sea level change over the basin with its mass and steric components estimated separately. Contamination of land signal in geodetic data is accounted for. The biggest challenge is for the ocean models, due to unrealistic boundary conditions at the Gibraltar and Dardanelli and Bosphorus straits and uncertainties in the air-sea freshwater fluxes and river-runoff. The small sea level trend of the models is implicitly related to the Boussinesq assumption, which implies conservation of volume rather than mass.
The geodetic data are from the gridded multi-mission altimeter dataset of the ESA Sea Level Climate Change Initiative and from GRACE monthly solutions. We account for Glacial Isostatic Adjustment (GIA) and correct the leakage of land signals using hydrological models.
In the Mediterranean Sea we consider two ocean simulations (RMCS, ENEA) and one reanalysis (CMEMS) assimilating satellite altimetry. The models differ mostly in annual amplitude and halosteric trend. Best agreement in trend, with 2.2 +/- 0.5 mm/yr in 1993-2016, is between altimetry and sum of modelled sea level and thermo-steric component.
In the Black Sea we consider the BS-CMEMS multi-year reanalysis based on the Nucleus for European Modelling of the Ocean v.3.6 (NEMO) hydrodynamic model. The data assimilation system ingests hydrographic profiles, altimeter sea level anomaly and space-based sea surface temperature. The trend of thermo-steric component is small over 1993-2016 (0.45 ± 0.01 mm/yr) and the halosteric component highly inaccurate, due to model freshwater forcing and scarcity of salinity data.
The synergy between altimeter data and model simulations could be used to overcome the errors of mass balances. Using tide gauge data, we discuss approaches for separating natural variability and long-term signal in models and data sets.