INSPIRE Thematic Clusters

Point Time Series Observation

  • Public

Point Time Series Observation

Started by Elena Flórez García Replies (2)

Hello everybody,

My name is Elena Flórez and I am a member of the TWG that is working with the guidelines of themes 13-14 in Spain.

We have a doubt with the time in the results values in pointTimeSeriesObservations.

Following the document “d2.9. Guidelines for the use of Observations & Measurements” (D2.9.O&M-swe-3.0) there’s an example of our doubt. ( TimeSeries Observation, page 12)

For example, we have an air quality monitoring station providing ozone concentration measurements. As this design pattern usually provides a time series result, the explicit phenomenonTime and/or resultTime will often be provided together with the result values.

In that example, the result is:

2010-03-01T01:00, 40.87

2010-03-01T02:00, 50.24

We have a doubt with the parameter values and the time. Is this time the beginning of the time period (in this case, hourly period) or is it the end? I mean, the 40.87 value is referred to time period from T01:00 to T02:00 or is it referred to period from T00:00 to T01:00?

Thank you very much.




    • Public

    By Ilkka RINNE

    Hi Elena,

    I'm replying to you based on the O&M model experience in general, but I'm guessing this would apply to the INSPIRE times series as well:

    The phenomenonTime is the instance or period of time for which the result of the observation (event) reflects the value of the observed property of the feature-of-interest. The resultTime is the time when the observation value has been analysed to produce the result.

    In your case I would guess that the air quality samplers could be collecting/averaging one result for the some time, like for an hour, 10 minutes etc. Depending on the insight you have on the observation process, there would be two options: If the result value reflects the observed property value during the whole of the particular time period (an average of many samplings), then I would use the time period for the phenomenonTime. If on the other hand the result reflects the air quality property value only after the collecting has been going on for the collecting time (one statistical sample is created for the collected particles), then I would use a single time instant for the phenomenonTime (the end of the collecting period).

    When the produced result is recorded (at the end of the hour or so), the value of the resultTime would be recorded with it. In some cases the raw measurement values are analysed much later, say the next day, or even re-analysed considerable later to fix errors in the original analysis. In these cases the resultTime is very convenient, as it provides a way for all the different observation results based no the same sampling to co-exist with traceable changes in the results produced by the different analysis.

    • Public

    By Elena Flórez García

    Hi Ilkka,

    Thank your very much. I'm sorry for my late reply. I really appreciate your explanation, the subject is clearer for me now. Thank you!