Independent Quality Assessment of Essential Climate Variables: Lessons Learned from the Copernicus Climate Change Service

Abstract

If climate services are to lead to effective use of climate information in decision-making to enable the transition to a climate-smart, climate-ready world, then the question of trust in the products and services is of paramount importance. The Copernicus Climate Change Service (C3S) has been actively grappling with how to build such trust; provision of demonstrably independent assessments of the quality of products, which was deemed an important element in such trust-building processes. C3S provides access to Essential Climate Variables (ECVs) from multiple sources to a broad set of users ranging from scientists to private companies and decision-makers. Here we outline the approach undertaken to coherently assess the quality of a suite of observation- and reanalysis-based ECV products covering the atmosphere, ocean, land and cryosphere. The assessment is based on four pillars: basic data checks, maturity of the datasets, fitness for purpose (scientific use cases and climate studies), and guidance to users. It is undertaken independently by scientific experts and presented alongside the datasets in a fully traceable, replicable and transparent manner. The methodology deployed is detailed, and example assessments are given. These independent scientific quality assessments are intended to guide users to ensure they use tools and datasets that are fit for purpose to answer their specific needs rather than simply use the first product they alight on. This is the first such effort to develop and apply an assessment framework consistently to all ECVs. Lessons learnt and future perspectives are outlined to potentially improve future assessment activities and thus climate services.

The Copernicus Climate Change Service: Climate Science in Action

“…Accordingly, C3S is implementing a comprehensive quality assurance framework for CDS data, tools and applications (e.g. Lacagnina et al, 2022;Yang et al, 2022) which includes end-to-end management of user requirements (Fig. 10).…”