The Copernicus ocean forecasting system for the Mediterranean Sea: description and quality assessment of recent evolutions

<p>The Mediterranean Monitoring and Forecast Center of the Copernicus Marine Service (Med-MFC) provides operational, regular and systematic reference information for the blue (Physics -Med-PHY- and Wave -Med-WAV) and green (Biogeochemistry -Med-BGC) state of the Mediterranean Sea. Based on state of the art modelling developments, the Med-MFC delivers Near Real Time (NRT) analysis and short-term (10 days) forecast and consistent Multi-Year (MY) Reanalysis reconstructions and their Interim extensions from 1987 (Med-PHY), 1993 (med-WAV) and 1999 (Med-BGC) till month minus one.</p> <p>This work aims at providing a detailed description and a quality assessment of recent modelling upgrades which have been implemented in the latest operational systems since November 2022.</p> <p>In particular, the major modelling advancements for each system are the following:</p> <ul> <li>Med-PHY NRT system: improvements in both the hydrodynamic model with a better tidal representation and data assimilation components including a new Mean Dynamic Topography, assimilation of 7km filtered altimeter data and ingestion of newly available altimeter data (Hy-2A/B and Sentinel-6A); delivery of a new variable: vertical velocity.</li> <li>Med-BGC NRT system includes the novel bio-optical configuration of the Biogeochemical Flux Model (BFM) and its coupling with the atmospheric light spectral model OASIM. Further, the 3DVarBio assimilation system is upgraded to include oxygen profiles from BGC-Argo.</li> <li>Med-WAV NRT system upgrades include: (1) implementation and tuning of WAM Cycle6, (2) implementation of Charnock parameter reduction for strong winds and (3) ingestion of newly available altimeter data into the data assimilation system (Sentinel-6A).</li> <li>All the 3 Reanalysis time series have been extended until June 2021.</li> </ul> <p>The model evolutions have been extensively qualified by comparing model results from a series of sensitivity numerical experiments with respect to best available satellite and <em>insitu</em> observations in order to provide a reliable validation assessment. All the evolutions have provided, to a different extent, an overall product quality increase by means of a decreased error and bias with respect to the previous version of the systems when comparing modelling data to observations and previous literature.</p>