Uncertainty checklist

This issue was created to capture the discussion we had on the 13/03/2025 during the Hackathon on what sources of uncertainty may/should be included in observational datasets in the REF. For context, metrics packages do not tend to use much uncertainty information, other than what is called "standard_error" and "number_of_observations" in their metric development, however, the standard_error can include many sources of uncertainty. Obs4MIPS currently requests the submission of a technical note alongside the dataset which would explain all the sources of uncertainty considered in producing the final uncertainty estimate. It would be useful for the REF to have a checklist for dataset providers to look through and help guide them in their preparation.