Access impact of observations

The accuracy of the Copernicus Marine Environment and Monitoring Service (CMEMS) ocean analysis and forecasts highly depend on the availability and quality of observations to be assimilated. In situ observations are complementary to satellite observations that are restricted to the ocean surface. Higher resolution model forecasts are required by users of the CMEMS global and regional ocean analysis and forecasts. To support this with an efficient observational constrain of the model forecast via data assimilation, an increase observation coverage is needed, associated with an improved usage of the available ocean observations. This work exploits the capabilities of operational systems to provide comprehensive information for the evolution of the GOOS. In this report, we analyse the use and the efficiency of the in-situ observations to constrain regional and global Mercator Ocean systems. Physical and biogeochemical variables are considered. The in-situ observations are used either to estimate physical ocean state at global and regional scale via data assimilation or to estimate BGC model parameters. The impact of the physical in situ observations assimilated in open ocean and coastal areas is assessed with numerical data assimilation experiments. The experiments are conducted with the regional 1/36° resolution and global 1/12° resolution systems operated by Mercator Ocean for the Copernicus Marine Service. For the global physical ocean, the focus is on the tropical ocean to better understand how the tropical mooring observations constrain the intraseasonal to daily variability and the complementarity with satellite observations and the deep ocean. The tropical moorings provide unique high frequency observations at different depth, but they are far away from each other, so part of the signal in the observation are decorrelated from one mooring to the others. It is only via an integrated approach, as data assimilation into a dynamical model and complementarity with other observing networks that those observations can efficiently constrain the different scales of variability of the tropical ocean circulation. As the satellite observations brings higher spatial resolution between the tropical moorings but for the ocean surface, we show that the tropical mooring and Argo profile data assimilation constrain the larger scale ocean thermohaline vertical structure (EuroSea D2.2; Gasparin et al., 2023). The representation of the high frequency signals observed at mooring location is also significantly improved in the model analysis compared to a non-assimilative simulation. The ocean below 2000 m depth is still largely under constrained as very few observations exist. Some deep ocean basins, as the Antarctic deep ocean, shows significant trend over the past decade but they are still not accurately monitored. Based on the spread of four deep ocean reanalysis estimates, large uncertainties were estimated in representing local heat and freshwater content in the deep ocean. Additionally, temperature and salinity field comparison with deep Argo observations demonstrates that reanalysis errors in the deep ocean are of the same size as or even stronger than the observed deep ocean signal. OSSE already suggested that the deployment of a global deep Argo array will significantly constrain the deep ocean in reanalysis to be closer to the observations (Gasparin et al., 2020). At regional and coastal scales, the physical ocean circulation is dominated by higher frequency, smaller scale processes than the open ocean which requires different observation strategy to be well monitor. The impact of assimilating high frequency and high-resolution observations provided by gliders on European shelves is analysed with the regional Iberic Biscay and Irish (IBI) system. It was found that repetitive glider sections can efficiently help to constrain the transport of water masses flowing across those sections. BGC ocean models are less mature than physical ocean models and some variable dependencies are still based on empirical functions. In this task, Argo BGC profile observations were used to optimize the parameters of the global CMEMS biogeochemical model, PISCES. A particle filter algorithm was chosen to optimize a 1D configuration of PISCES in the North Atlantic. The optimization of the PISCES 1D model significantly improves the model's ability to reproduce the North Atlantic bloom Recommendations on the in-situ network extensions for real time ocean monitoring are given based on those results, and the one also obtained in the WP2, Task 2.2 where data assimilation experiments but with simulated observations where conducted. Argo extension and the complementarity with satellite altimetry was also extensively studied. (EuroSea Deliverable ; D4.8)