Novel metrics based on Biogeochemical Argo data to improve the model uncertainty evaluation of the CMEMS Mediterranean marine ecosystem forecasts

Stefano Salon,Anna Teruzzi,Laura Feudale,Paolo Lazzari,Cosimo Solidoro,Gianpiero Cossarini,Alessandro Crise,Giorgio Bolzon

doi:10.5194/os-15-997-2019

Abstract

Abstract. The quality of the upgraded version of the Copernicus Marine Environment Monitoring Service (CMEMS) biogeochemical operational system of the Mediterranean Sea (MedBFM) is assessed in terms of consistency and forecast skill, following a mixed validation protocol that exploits different reference data from satellite, oceanographic databases, Biogeochemical Argo floats, and literature. We show that the quality of the MedBFM system has been improved in the previous 10 years. We demonstrate that a set of metrics based on the GODAE (Global Ocean Data Assimilation Experiment) paradigm can be efficiently applied to validate an operational model system for biogeochemical and ecosystem forecasts. The accuracy of the CMEMS biogeochemical products for the Mediterranean Sea can be achieved from basin-wide and seasonal scales to mesoscale and weekly scales, and its level depends on the specific variable and the availability of reference data, the latter being an important prerequisite to build robust statistics. In particular, the use of the Biogeochemical Argo floats data proved to significantly enhance the validation framework of operational biogeochemical models. New skill metrics, aimed to assess key biogeochemical processes and dynamics (e.g. deep chlorophyll maximum depth, nitracline depth), can be easily implemented to routinely monitor the quality of the products and highlight possible anomalies through the comparison of near-real-time (NRT) forecasts skill with pre-operationally defined seasonal benchmarks. Feedbacks to the observing autonomous systems in terms of quality control and deployment strategy are also discussed.

Highlights

Operational ocean forecasting systems integrate remote observations, in situ measurements and modelling systems, and have been widely recognized as important assets for ocean state monitoring and the development of the blue economy (She et al, 2016)
To evaluate the model consistency (GODAE Class 1 metrics) with the general features of the biogeochemistry of the Mediterranean Sea in terms of chlorophyll, nutrients, dissolved oxygen, carbonate system variables (DIC, ALK, pressure of CO2 (pCO2), pH), and primary production, model mean fields are compared with different reference datasets
Modelled late-winter to early-spring surface chlorophyll maxima in nwm appear anticipated of 2–3 weeks with respect to the satellite ones: this is related to a possible mismatch of the spatial patterns which characterize the temporal succession of deep convection and subsequent stratification and bloom, known to have very highly patchy dynamics in this area (Estrada et al, 2014; Mayot et al, 2017; Severin et al, 2017)

Summary

Introduction

Operational ocean forecasting systems integrate remote observations, in situ measurements and modelling systems, and have been widely recognized as important assets for ocean state monitoring (von Schuckmann et al, 2016) and the development of the blue economy (She et al, 2016) In such a framework, the operational monitoring and forecasting of marine biogeochemistry and ecosystem dynamics is based on biogeochemical models designed to represent the low-trophic-level ecosystem (i.e. from phytoplankton to zooplankton). The operational monitoring and forecasting of marine biogeochemistry and ecosystem dynamics is based on biogeochemical models designed to represent the low-trophic-level ecosystem (i.e. from phytoplankton to zooplankton) The improvement of their predictive capability on weekly and seasonal timescales mostly required by users is strongly related to the development of data assimilation capacity, while their quality assessment is constrained by the availability of reference data, both remote and in situ (Gehlen et al, 2015), and possibly independent (i.e. not assimilated; Gregg et al, 2009). As a user-driven service based on a “continuous improvement” philosophy, CMEMS is com-

Results

Discussion

Conclusion