An intercomparison of Net Primary Production Copernicus Marine Service products in the Mediterranean Sea

The EU Copernicus Marine Service (CMS) provides a cascade of physical, biogeochemical, and sea-ice state products for the global ocean and European regional seas. Among these, net primary production (NPP) is a key variable representing the process by which phytoplankton convert inorganic carbon into organic matter, thereby forming the base of the food web and regulating biogeochemical cycles in marine ecosystems. Multiple model- and satellite-based NPP products are available through CMS, and understanding to what extent these products agree or disagree is important for linking potential discrepancies to underlying differences in algorithms or model assumptions, guiding product selection, and better interpreting spatiotemporal trends. Here, we intercompare NPP estimates in the Mediterranean Sea from model simulations and satellite observations. Modelled NPP is part of three products provided by CMS, i.e., the nominal modelling products for the Global (GLO), the Atlantic-Iberian Biscay (IBI), and the Mediterranean Sea (MED) biogeochemistry, while two products include NPP based on satellite ocean colour observations, i.e., the Global and the Mediterranean Sea Ocean Colour products. Preliminary results from climatologies calculated over January 1999–December 2022 revealed differences in the timing of NPP peaks. GLO and IBI models (integrated over 200 m) exhibited maxima in February–March, whereas both the MED model (also integrated over 200 m) and satellite products peaked in summer. Monthly NPP magnitudes, however, remained within comparable ranges across all datasets. Consistently, Hovmöller diagrams (models only) clearly showed high NPP in the surface layers during the winter bloom. The GLO hindcast model detected relatively high NPP at the level of the summer deep chlorophyll maxima, while NPP appeared more evenly distributed along the vertical layers across summer months in the MED model. No significant seasonal trends were observed in any of the datasets. Future analyses will include a comparison with the non-Copernicus NPP data from the Carbon, Absorption, and Fluorescence Euphotic-resolving (CAFE) model, as well as a similar intercomparison exercise for CMS chlorophyll products.