On 25-26 November 2025, ObsSea4Clim held an internal project workshop aimed at synchronizing the work across project’s work packages and outputs. The focus was on the EOV/ECVs requirements and indicators, as they are both key elements of the Rolling Review of Requirements (RRR), the core principle of the ObsSea4Cim operations. 

Moreover, the workshop provided valuable input for refining the stakeholder groups and anticipating a dialogue and for planning for dissemination and uptake of results.

Selected outcomes of the workshop:

• Key elements of EOVs in sustained ocean observing, namely quality specifications and observing requirements, were jointly presented and discussed for the three types of data that are used for describing the ocean (physics) climate state: 1) observational data, 2) reanalysis and gridded data, and 3) ocean hindcast/forecast data.
• As expected, a considerable heterogeneity in pathways towards a common framework for quality specifications is shown in the presentations:
  • For observational data, it is acknowledged that approaches exist for some parameters by using reference materials (e.g., Standard Seawater, fixed-point calibrations, blackbody furnace calibrations) and standards (e.g., ITS-90, TEOS-10, ISO 2020-08). However, for other parameters, standardised approaches are missing (e.g., current vector).  
  • For gridded data products, uncertainties are more complex because they arise from the underlying observational (or remotely sensed) data and from extrapolation between data points (e.g., gridding error). 
  • For reanalysis data, a variety of sources of uncertainty have been identified (Storto and Masiaa 2017), including availability of underlying observational data (also dependent on the parameter, e.g. current speed and direction are almost always a result of the model and not assimilated), model resolution and model physics (incl. parametrisations), data assimilation method used for the reanalysis. The variety of sources of uncertainty makes a coherent quantification challenging. 
  • For numerical model data, uncertainty estimates mainly arise from model resolution, model physics, parametrisation, uncertainties in forcing and uncertainties in the initial conditions (incl. stochastic processes), and numerical constraints. One approach to provide a measure of simulation stability is via Ensemble forecasting, which may create a range of possible outcomes by running multiple simulations with varied initial conditions and/or perturbed model physics to provide a probabilistic prediction, rather than a single deterministic one.
• The requirements for EOVs came through the indicator sessions during the workshop, as indicators should ideally address the need for knowledge of specific states of ocean processes, such as local warming, extreme sea level, trends in sea-ice cover loss, and trends in stratification. Selected indicators have been presented at the workshop for each WP4 task, which also gives a breadth of the nation’s interest in ocean observing. The most mature and developed indicators were chosen by WP3 as pilots.  
• The workshop provided a great summary of ObsSea4Clim joint operations and identified themes for policy briefs, stakeholder dialogue, and broader dissemination plans.