From Earth Science Information Partners (ESIP)

Data interoperability is a goal of many programs, both globally (e.g., GEOSS) and within US agencies (e.g., NSF's EarthCube program). In many cases, this is simply because the amount of time it takes for a scientist to find the data they need is difficult, especially when as is increasingly the case, the data comes from outside their discipline and then dealing with that data (assessing its applicability for their uses) and ultimately using that data takes up the majority of their limited research time. What users need is the ability to discover, assess, and ultimately use the data they need, from within their tools of choice. However, in order for this to work in a scalable way, data needs to be automatically discoverable and usable, which in term implies that the data and the services that provide access to that data are machine actionable. In other words the web services need to be described in a machine understandable manner such they are autonomously callable.