Coordination Action for the integration of


Solar System Infrastructures and Science

Observational Metadata

The Observational Metadata describes the observations and hence what the file contains – in particular, it should contain information about how the observations were made.

It is important that the observational metadata should be as complete as possible so that it stands by itself with minimal need to refer to auxiliary metadata. Note that it may not be possible to use the observations in some circumstances if the observational metadata are not properly formed.

Even if the data are from very different domains, there are certain pieces of information that always need to be conveyed as part of the observational metadata – for example, date and time of the observation, name of the observatory and instrument, observing domain, target, etc. There also parts of the observational metadata that are specific to the instrument or domain.

The key to improved interoperability is to maximize the number of common parameters used in the observational metadata – that is, push the boundary between the common part and domain-specific part as far down as possible.

The observational metadata is largely defined when a data system is established for a new instrument or observatory. It is extremely important that a lot of thought goes into the design of this since making changes can be difficult once the data train has started rolling.

It should be noted that a move to use standards to describe the observational matadata also has implication on the Derived Metadata since some of this is based on the Observational Metadata.

Common Part

The common area contains certain pieces of information that always need to be conveyed – date and time of the observation, name of the observatory and instrument, observing domain, location of the observatory and the target of the observations, etc.

In order to maximize interoperability the names, meaning and possible contents of parameters used in the common area need to be agreed by as wide a community as possible – i.e. a standard needs to be established.

Domain-specific Part

There will always be large parts of the metadata in a file containing observations that are specific to the instrument or domain.

Any attempt to standardize this over too wide a community would be counter-productive but within a domain it is important that discussions take place and a set of parameter names and their meaning agreed – these should be described within some form of data model.