Uncertainty Information in Climate Data Records
Merchant, Christopher1; Corlett, Gary2; Rayner, Nick3; Donlon, Craig4
1University of Edinburgh/Reading, UNITED KINGDOM; 2University of Leicester, UNITED KINGDOM; 3UK Met Office, UNITED KINGDOM; 4ESA, NETHERLANDS

The European Space Agency created the Climate Change Initiative (CCI) to maximize the usefulness of Earth Observations to climate science. An important aspect of developing climate data records (CDRs) is to provide uncertainty information to users. In this presentation, we argue that uncertainty information is an important part of the dataset that must be validated in its own right. Uncertainty is rarely uniform within a dataset, and in the context of climate model evaluation or assimilation, it is important to identify which are the more or less certain data. Climate users have diverse needs for data in terms of spatiotemporal resolution, which implies that uncertainty information must be provided in a form that supports proper propagation of uncertainty estimates when data are averaged to other resolutions. This requires description of the spatio-temporal correlations of errors.