INSPIRE Thematic Clusters

Inspire OF v EF - Which to Choose?


Inspire OF v EF - Which to Choose?

Started by Keiran MILLARD Replies (1)

There has been some discussion offline about when a Member State should publish oceanographic data as OF and when to use EF. So, here are some general principles that I use.

Some background.  All the Inspire data specifications related to observational coverages  (time series, grids, grid series) use the same common data models and this is specified in the Inspire General Conceptual Model.  That means that OF, EF and AC/MF all use the same approach to describe the coverage element of their data.  So far, so good.

EF describes the observational network/activity/facility (the thing doing the measuring)  and optionally the results of this network.  If your dataset is describing an observation process and the results it delivers then EF is normally a better publishing option.  This is because it allows all the coverages to be described related to the measurement process and not individual Inspire Themes.  For example if I have an observation program that is collecting oceanographic physical and chemical parameters along with atmospheric parameters and marine biological parameters then it makes sense to publish this as EF describing observation process and its associated results.  It does not make sense to publish all of these data sets under separate Inspire themes as the integrity of the observation process is reduced.

OF simply describes the physical or chemical properties of the marine environment as a coverage, separate from a particular observation activity.  This may be the case when you have a derived dataset, or a dataset comprised of the analysis of several observations.  An OF dataset would identify any observation datasets through its metadata.


    • Public

    By Katharina SCHLEIDT

    Hi all,

    a few further thoughts from my (O&M&EF centric) perspective:

    first off, while Keiran only mentions coverage observations, there are more out there in the INSPIRE Observational Model in the GCM. The observation specializations available within INSPIRE also cover time series, single one off measurements or observations (need not be a sensor, could be a person determining ocean color, or also reporting the presence (or absence) of a specific species).

    secondly, while the OF model provides a list of possible observation types to be provided together with their measurement metadata (WHAT was measured WHERE using what process (HOW)), the EF model pulls this measurement metadata together (it references exactly the same objects describing the WHAT, WHERE and HOW, but pulls them together and provides context. In addition, EF can also be used for the description of the aggregation (or other types of data derivation) methodology.

    As to the question which theme what parts should be served under, this is a bit tricky as the INSPIRE Dataset concept doesn't well align to the Themes; there are many examples of logically consistent datasets spanning themes. EF is especially prone to this problem as some of the environmental media covered also have their own theme (i.e. OF, AC, Soil, ...) while others don't.

    Where it gets especially tricky is the question of which service provides which dataset. The basic EF part logically goes to a WFS, the measurement aspects (as covered in OF) are far better suited to an SOS; how to deal with the overlapping bits. For the Austrian Air Quality Reporting Services I tried an interesting approach of linking these services, see my presentation on Austrian AQD Reporting

    TL;DR; Although an organisation is providing OF data, as this usually stems from an EF, this must also be provided