Abstract
This paper demonstrates the value of Observing System Evaluation (OS-Eval) efforts which have been made or are ongoing to contribute to observing system review and design with the support of Ocean Data Assimilation and Prediction (ODAP) communities such as GODAE OceanView and CLIVAR-GSOP, by highlighting examples that illustrate the potential of the related OS-Eval methodologies and recent achievements. For instance, Observing System Experiment (OSE) studies illustrate the impacts of the severe decrease in the number of TAO buoys during 2012-2014 and TRITON buoys since 2013 on ODAP system performance. Multi-system evaluation of the impacts of assimilating satellite sea surface salinity data based on OSEs has been performed to demonstrate the need to continue and enhance satellite salinity missions. Impacts of underwater gliders have been assessed using Observing System Simulation Experiments (OSSEs) to provide guidance on effective coordination of the western North Atlantic observing system elements. OSSEs are also being performed under H2020 AtlantOS project with the goal to enhance and optimize the Atlantic in-situ networks. Potential of future satellite missions of wide-swash altimetry and surface ocean currents monitoring is explored through OSSEs and evaluation of Degrees of Freedom for Signal (DFS). Forecast Sensitivity Observation Impacts (FSOI) are routinely evaluated for monitoring the ocean observation impacts in the US Navy’s ODAP system. Perspective on the extension of OS-Eval to the deep ocean, polar regions, coupled data assimilation, and biogeochemical applications are also presented. Based on the examples above, we identify the limitations of OS-Eval, indicating that the most significant limitation is reduction of robustness and reliability of the results due to their system-dependency. Inability of performing evaluation in near real time is also critical. A strategy to mitigate the limitation and to strengthen the impact of evaluations is discussed. In particular, we emphasize the importance of collaboration within the ODAP community for multi-system evaluation and communication with ocean observational communities on the design of OS-Eval, required resources, and effective distribution of the results. Finally, we recommend to further develop OS-Eval activities at international level with the support of the international ODAP (e.g., OceanPredict and CLIVAR-GSOP) and observational communities.
Highlights
Ocean Data Assimilation and Prediction (ODAP; see Table 1 for summary of essential acronyms) systems, which include ocean reanalysis systems for seasonal forecasting and longterm ocean state estimation systems, are used in a large range of oceanic applications and weather and climate forecasting services as an essential tool for integrating ocean observations and numerical forecasting models (e.g., Davidson et al, 2009; Brassington et al, 2015; Le Traon et al, 2017)
Ocean color satellites routinely provide global observations of optical properties and chlorophyll concentration for over two decades. This has proved an invaluable tool for reanalysis and forecasting (Gehlen et al, 2015), but the coverage is restricted to the near-surface and cloud-free conditions, and limited information can be obtained about other variables such as nutrient concentrations
This study showed the potential of such an approach by explicitly simulating the joint effects of uncertain biological parameters and unresolved scales using a stochastic model to simulate an ensemble of 60 members in a 1/4◦ resolution North Atlantic configuration
Summary
Ocean Data Assimilation and Prediction (ODAP; see Table 1 for summary of essential acronyms) systems, which include ocean reanalysis systems for seasonal forecasting and longterm ocean state estimation systems, are used in a large range of oceanic applications and weather and climate forecasting services as an essential tool for integrating ocean observations and numerical forecasting models (e.g., Davidson et al, 2009; Brassington et al, 2015; Le Traon et al, 2017). Potential future evolutions of the in-situ observing system, including Argo floats, drifting buoys, and mooring arrays, are defined based on strong interactions between observation agencies and forecasting centers To our knowledge, this is the first time that such a coordinated effort is made using OSSEs. The multi-model and multi-approach feature is a critical point to ensure the robustness of the results (see section Multi-System Evaluation). The DFS can be conveniently computed in an EnKF setting as a by-product of the calculation of the Kalman Gain without additional computing costs (Sakov et al, 2012) It is relevant in the planning phase of a new observing system when the actual data are not yet available, FIGURE 9 | FSOI evaluated for the ODAP system based on the global HYCOM. Adjoint codes of a coupled model are required for calculating the FSOI of a coupled data assimilation system
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
