A High Resolution Reanalysis for the Mediterranean Sea

In order to be able to forecast the weather and estimate future climate changes in the
ocean, it is crucial to understand the past and the mechanisms responsible for the ocean
variability. This is particularly true in a complex area such as the Mediterranean Sea with
diverse dynamics like deep convection and overturning circulation. To this end, effective
tools are ocean reanalyses or reconstructions of the past ocean state. Here we present a
new physical reanalysis of the Mediterranean Sea at high resolution, developed in the
Copernicus Marine Environment Monitoring Service (CMEMS) framework. The
hydrodynamic model is based on the Nucleus for European Modelling of the Ocean
(NEMO) combined with a variational data assimilation scheme (OceanVar). The model has
a horizontal resolution of 1/24° and 141 unevenly distributed vertical z* levels. It provides
daily and monthly temperature, salinity, current, sea level and mixed layer depth as well as
hourly fields for surface velocities and sea level. ECMWF ERA-5 atmospheric fields force
the model and daily boundary conditions in the Atlantic are taken from a global reanalysis.
The reanalysis covers the 33 years from 1987 to 2019. Initialized from SeaDataNet
climatology in January 1985, it reaches a nominal state after a 2-years spin-up. In-situ
data from CTD, ARGO floats and XBT are assimilated into the model in combination with
satellite altimetry observations. This reanalysis has been validated and assessed through
comparison to in-situ and satellite observations as well as literature climatologies. The
results show an overall improvement of the comparison with observations and a better
representation of the main dynamics of the region compared to a previous, lower resolution
(1/16°), reanalysis. Temperature and salinity RMSD are decreased by respectively 14 and
18%. The salinity biases at depth of the previous version are corrected. Climate signals
show continuous increase of the temperature and salinity, confirming estimates from
observations and other reanalysis. The new reanalysis will allow the study of physical
processes at multi-scales, from the large scale to the transient small mesoscale structures
and the selection of climate indicators for the basin.