Assessing the impact of multiple altimeter missions and Argo in a global eddy-permitting data assimilation system

Abstract

Abstract. A series of observing system simulation experiments (OSSEs) is carried out with a global data assimilation system at 1/4 • resolution using simulated data derived from a 1/12 • resolution free-run simulation. The objective is to not only quantify how well multiple altimeter missions and Argo profiling floats can constrain the global ocean analysis and 7-day forecast at 1/4 • resolution but also to better understand the sensitivity of results to data assimilation techniques used in Mercator Ocean operational systems. The impact of multiple altimeter data is clearly evidenced even at a 1/4 • resolution. Seven-day forecasts of sea level and ocean currents are significantly improved when moving from one altimeter to two altimeters not only on the sea level, but also on the 3-D thermohaline structure and currents. In high-eddy-energy regions, sea level and surface current 7-day forecast errors when assimilating one altimeter data set are respectively 20 and 45 % of the error of the simulation without assimilation. Seven-day forecasts of sea level and ocean currents continue to be improved when moving from one altimeter to two altimeters with a relative error reduction of almost 30 %. The addition of a third altimeter still improves the 7-day forecasts even at this medium 1/4 • resolution and brings an additional relative error reduction of about 10 %. The error level of the analysis with one altimeter is close to the 7-day forecast error level when two or three altimeter data sets are assimilated. Assimilating altimeter data also improves the representation of the 3-D ocean fields. The addition of Argo has a major impact on improving temperature and demonstrates the essential role of Argo together with altimetry in constraining a global data assimilation system. Salinity fields are only marginally improved. Results derived from these OSSEs are consistent with those derived from experiments with real data (observing system evaluations, OSEs) but they allow for more detailed characterisation of errors on analyses and 7-day forecasts. Both OSEs and OSSEs should be systematically used and intercompared to test data assimilation systems and quantify the impact of existing observing systems.

From Observation to Information and Users: The Copernicus Marine Service Perspective

“…CMEMS modeling and data assimilation systems highly depend on the status of the altimeter constellation (Le Traon et al, 2017b). Both OSEs (e.g., Hamon et al, in press) and OSSEs (e.g., Verrier et al, 2017) demonstrate the major contribution of altimetry. At least three and preferably four altimeters are required to constrain modeling and data assimilation systems (Figure 4).…”

Contribution of future wide-swath altimetry missions to ocean analysis and forecasting

“…As the atmospheric pressure forcing is not considered in the global reanalysis, the inverse barometer effects (e.g. Wunsch and Stammer, 1997) are computed from the ECMWF pressure fields and applied along the boundaries. Tidal harmonics were obtained from a 10-year free run simulation, performed using the same OGCM configuration (AR).…”

Requirements for an Integrated in situ Atlantic Ocean Observing System From Coordinated Observing System Simulation Experiments

“…Several coordinated initiatives are currently handled in the framework of the Global Ocean Data Assimilation Experiment (GODAE) Ocean View (Bell et al, 2015), such as the inter-comparison and validation approaches of the forecasting systems (e.g., Ryan et al, 2015) and reanalysis (e.g., Balmaseda et al, 2015). Over the last two decades, assessments of the impact of observations on ocean forecasts and reanalysis have regularly been made in order (i) to optimize the use of observation information in the analysis step and to improve the assimilation component (e.g., Li et al, 2008), (ii) to quantify the impact of the present observation network on ocean analyses and forecasts (e.g., Storto et al, 2013;Oke et al, 2015;Turpin et al, 2015), (iii) to demonstrate the value of an observation network for operational ocean analyses and forecasts (e.g., Lea et al, 2014;Gasparin et al, 2018), and (iv) to help define and test future observing systems from an integrated system view, involving satellite and in situ observations and numerical models (e.g., Verrier et al, 2017). However, the evolution of observing networks and operational forecasting and analysis systems requires an updated and refined understanding of observations' impact on numerical models.…”