To get a useful definition of EVs it is necessary to clarify the meaning of the terms “essential” and “variable”. The adjective “essential” has a number of different meanings, ranging from absolutely necessary and indispensable to containing an essence of something. A variable that significantly improves the reliability and accuracy of desired results can be considered essential. A variable that provides important information related to a specific goal is essential for this goal, independent of the capability to actually observe the variable. The essentiality of a variable may also depend on the information needs of different communities and target users (science, policy, etc.). In principle any variable can be “essential” for somebody or for achieving something. Therefore, the definition of what an EV is needs to be formalized and endorsed according to the collective interest and preferably in an internationally recognized process.

In the context of the ConnectinGEO project, EVs are “a minimal set of variables that determine the system’s state and developments, are crucial for predicting system developments, and allow us to define metrics that measure the trajectory of the system” (ConnectinGEO, 2015). The spatial and temporal resolution of observations of any given EV depend on the use of the observations, and in some cases, available observations of an EV may not be ready for use for specific applications. Limited knowledge of EVs implies limited predictive capabilities and limited means to measure where the system is heading. The generic nature of this EV definition allows us to identify sets of EVs for whatever system is being considered and for whatever goals are established in relation to this system. It allows for a broad application across different communities, and can be complemented with the specific requirements that each community may have related to the system the community is focusing on. Moreover, in a system of systems approach, sets of EVs can be aggregated into larger sets supporting prioritization of efforts across community and system boundaries.

The concept of “variable” embedded in the definition given above has a certain level of abstraction. Identifying a variable does not imply that observation requirements in terms of spatial and temporal resolution, accuracy, latency, observation interval, etc. are also specified. Nor does it imply that measurement instruments are available to observe the variable. In some cases, variable may not be observable directly and have to be derived from a combination of observations. In such cases, the essential variable may have to be composed of a set of sub-variables that together provide the required information.

One of the most important Workshop’s objectives was to review the different definition of EVs used in the different communities engaged or linked to GEO and to evaluate the level of consistency of these definitions. Although the “landscape” is quite diverse and complex, there seems to be some level of consistency with the system relevance of EVs but also considerable diversity with respect to the process how EVs are identified, validated and accepted.

The definition of the Essential Climate Variables (ECVs) developed by the Global Climate Observing System (GCOS) and endorsed by the United Nation Framework Convention on Climate Change (UNFCCC) is consistent with the definition chosen by ConnectinGEO: “An ECV is a physical, chemical, or biological variable or a group of linked variables that critically contributes to the characterization of Earth’s climate.” Datasets of ECV provide the empirical evidence needed to document, understand and predict the evolution of climate, to guide mitigation and adaptation measures, to assess risks and enable attribution of climatic events to underlying causes, and to underpin climate services (Bojinski et al., 2014). ECVs are widely accepted and used at international level.