Using BGC-Argo Data to Disentangle the Physical and Biogeochemical Uncertainty in the Quality of Operational Biogeochemical Products
Stefano Salon1, Laura Feudale1, Gianpiero Cossarini1, Anna Teruzzi2, Giorgio Bolzon1 and Emanuela Clementi3, (1)National Institute of Oceanography and Applied Geophysics (OGS), Italy, (2)National Institute of Oceanography and Applied Geophysics (OGS), Oceanography, Trieste, Italy, (3)Euro-Mediterranean Center on Climate Change, Italy
Abstract:
The quality of operational biogeochemical forecast products is commonly assessed against reference data sets, whose availability has recently increased thanks to the BGC-Argo network. The error in biogeochemical model outputs, estimated in terms of root mean square difference (RMSD) between co-located model output and BGC-Argo data, has been shown to be proportional to the local variability, which depends on physical (e.g. vertical mixing, mesoscale dynamics, coastal effects due to strong topographic gradients and terrestrial inputs) and biogeochemical (e.g. chemical/biological interactions) processes. Moreover, the uncertainty in the physical forcing, related both to the physical model itself, the atmospheric forcing, and the data assimilation process, may propagate to the biogeochemical dynamics, thus contributing to the error in biogeochemical variables. Forecast accuracy is thus related to the representativeness errors of physical and biogeochemical models, boundary conditions, inner model parameterisation, and data assimilation schemes.
Through the definition of different model error metrics based on comparing physical and biogeochemical model variables with Argo and BGC-Argo data, our focus is to highlight the impact of the data assimilation of physical quantities over the quality of the biogeochemical ones. In particular, for a 1-year simulation of the Mediterranean Sea biogeochemistry at 1/24 degree resolution, we clustered float profiles characterised by different impacts of physical forcing over the biogeochemical variables quality, quantified according to the RMSD of specific fields (temperature, chlorophyll, nitrate). The clustered profile metrics have been then related to specific area and season, aiming to correlate the presence of peculiar physical processes to an increase or a decrease of RMSD between model and BGC-Argo. Resulting statistics may be translated into feedbacks to physical models in the perspective view of a collaborative evolution of coupled physical–biogeochemical systems, both in terms of process modelling or coupled data assimilation.
