European Commission logo
INSPIRE Community Forum

Environmental Noise Directive

I would like to hear from other countries how you regard data that member states are required to produce according to the Environmental Noise Directive 2002/49/EC. All of you out there, do you consider that data as falling within the scope of the Inspire HH-theme?

We have 2 concerns about this.

1. We don't think that it is reasonable that we should be obliged to provide and harmonise data already reported according to another directive and another data specification.

2. We find it diffucult to actually map this data to the Inspire HH-dataspecification. From the Inspire-spec:

HH data model contains a module to address environmental health determinants (envhealth). It offers a possibility to refer to:

- raw environmental health determinant measurement data. This model is based on ISO 19103 on measurements data. This model is extended for localised noise and concentration data, and may be extended in a similar way to other health determinants. EIONET code lists are recommended to describe these measurement data.

- aggregated environmental health determinant measurement data through linking with SU data model

- coverage data resulting from the interpolation of raw measurement data


We don't consider the data reported through 2002/49/EC to be raw measurements. Those data are based on calculations of noise. Or should we consider these calculations as measurements? All data reported through 2002/49/EC don't match any statistical units, unless that statistical unit is supposed to be the national level. Of course this could be prepared if necessary. But that is just our point - is this really necessary?


    Hi Jenny,

    We have the same concerns. First of all, the question is: is the choice of Inspire theme correct? I have done some research in the Inspire geoportal, and a large part of the submitted datasets related to directive 2002/49 (END) are categorized in HH ("Human health & Safety", Annex III.5); this includes the Finnish, Dutch, Belgian, Austrian strategic noise maps. However, the French and English maps are categorized in AM ("Area management", Annex III.11); The Spanish and German maps are mixed (some HH, some AM, some even Environmental monitoring Facilities, EF).

    Let's for now assume that the category HH (human health & safety) is correct. From the HH data specifications v3.0 paragraph 2.2 (informal description), 5 components in HH are considered:

    • Health statistical data on diseases, poisoning, injuries, etc., and data on general health status in a population, such as self-perceived health, people with health problems, smokers, etc.
    • Biomarkers (general health data)
    • Determinants of health (environmental health determinants a.k.a. envhealth)
    • Health care/services data, for example on services provided by hospitals, health care workforce, etc
    • Safety

    We can assume that the directive 2002/49 (END) is situated in the third component, being environmental health determinants, since it is explicitly mentioned in the respective description ["Noise data (Directive 2002/49/EC of the European Parliament and of the Council of 25 June 2002 relating to the assessment and management of environmental noise)."].

    From the EnvHealthDeterminant Diagram, we have 3 options to spatially represent the data (section, :

    Health determinant data are represented as:

    1. Raw measurement data located somewhere.
    2. Statistical aggregation of these raw measurement data reported on some statistical units
    3. (continuous?) Coverages resulting from an interpolation of these raw measurement data.

    In all 3 options, the data is regarded as "measurement data", which is fundamentally wrong in our case since it is modelled data. But let's ignore that for now.

    In our case, where we publish shapefiles of noise contours (5dB intervals), i think we should prefer option 3: if one considers the individual raster cells (=output of the noise propagation model) as individual raw measurements, then the contours can be regarded as an interpolation of these raw data to a continuous coverage surface, as depicted in Figure 13 and 14 of the HH Data Specifications.

    The main problem with option 2 is that it is not really possible to aggregate this to a specific existing statistical unit, as the spatial information will be lost and the primary objective of the data publication (spatial information) will vanish.

    I think it is really important to have a common approach for all memberstates. First of all, the fact that some memberstates have a different Inspire data theme for the strategic noise maps, will already result in large differences of data specifications. But also for the MS that use Human Health as theme, it should be clear how to effectively use these specifications for our case: for instance, some MS will only publish the noise contours in a vector format (as we do), others could very well also make the "raw" model output publicly available, usually in a raster format; this difference would already change the way how to follow the data specifications.

    Please note that the European Commission does not oblige MS to submit spatially explicit results (only graphic and tabular results), but the spatial data that has been produced simply falls under the Inspire regulations and hence needs to comply to it.

    best regards






    Dear colleagues, 

    we would like to revive this discussion. To encourage this we have uploaded the following memo for discussion: 

    We would like to invite reactions through this forum or (if the reaction should be of confidential nature) via the enclosed contact information.  

    Best regards


  • Katharina SCHLEIDT

    By Katharina SCHLEIDT

    Hi All,

    I fear before noise data can be provided via INSPIRE, we must first clarify the problems in the data model.

    Problem + possible solution at:





    Dear All,

    I would like to bring up a query with regards to the harmonisation process of statistical data in order to be INSPIRE compliant.

    By statistical data, I mean the statistical information that is reported in excel workbook on EIONET concerning the noise exposure data for each noise source as per Directive 2002/49/EC (END). This includes the estimated number of people exposed to Lden and Lnight noise indicators for each of the 5dB noise contour bands from the different sources for each respective strategic noise map.

    I would like to hear from other countries how member states are required to produce the statistical information according to the INSPIRE Directive. Is there a way how to incorporate such data in the spatial datasets (maybe an attribute to the shapefiles?)

    Any replies are greatly appreciated

    Best Regards,


  • Johanna Ott

    Hi Kathi,

    I think I am currently running into the data model issues you are writing about. Unfortunately the link to the PDF does not work. Is there a thread in the Thematic Cluster on that somewhere? I cannot find any.

    Do you have any information if a change of the model is planned and what should be the workaround to use until then? Should we work with the solution used in your example:

    All the best


  • Katharina SCHLEIDT

    By Katharina SCHLEIDT

    Hi Johanna,

    sorry for not catching this post earlier, but was traveling a lot. Now back in Vienna and just caught your personal message. Sorry for taking the file off of my server, now the link should work again:

    Still havn't found the underlying work, probably got lost as I was leaving UBA, but think I've got the essence nailed in the pdf

    As for status, this doc was sent to JRC when I did it, got a nice thank you, but then just the sound of silence ;) However, as I think the way I remodelled it could actually be OK based on the IRs, it could be worth a 2nd look, and worst case doing a Good Practice to just establish this as a way forward (think I could do a parallel hack on XSD level fairly quickly if you need it)



  • Johanna Ott

    Hi Kathi,

    After finally reading your PDF file, I think my issue is slightly different from yours (or at least it is not explicitely mentioned in the PDF): independently from how the model's associations are changed, there is a discrepancy between the Measure-Type that should be used according to the INSPIRE Data Specification ( and the one used in the INSPIRE xsd ( The Data Specification demands the ISO 19105 Measure-Type (chapter while the xsd is using the gml:Measure-Type.

    Do you know who we could ask for more input on that and the current state of the review of your PDF at JRC?


    All the best


  • Katharina SCHLEIDT

    By Katharina SCHLEIDT

    Hi Johanna,

    fear the HH spec is a bit self-contradictory! It mostly correctly assigns the Measure type to ISO 19103, only mention of 19105 is in the UML diagram in the frame, but I believe this is a typo. In the same UML folder, there is also a specific Measure diagram, that correctly references 19103

    HH measure ISO 19103 extraction

    I fear its always a bit tricky figuring out which ISO spec is encoded in which namespace, at least I'm still looking for this document! My approach tends to be as follows:

    • if its something very specific, it may be in a dedicated namespace, i.e. the O&M stuff, SWE Common
    • if its something closely related to metadata (organizations, citations), I'll look towards the ISO namespaces, gmd, gco, ...
    • otherwise, its probably in GML!

    Use of gml:measure for encoding of the 19103 is correct, see the comment at the start of the gml measures.xsd: "See ISO/DIS 19136 17.3.
    gml:MeasureType is defined in the basicTypes schema.  The measure types defined here correspond with a set of convenience measure types described in ISO/TS 19103. ..."

    As for whom to contact at JRC, I believe that Michael Lutz is the main contact, adding Robert Tomas can't hurt



  • Johanna Ott

    Hi Kathi,

    Thank you for the reply and all your explanations and recommendations.

    From my point of view, using the gml:measure encoding, there is still the issue on how to reference the EIONET codelist with the attributes available there.

    I will try to get input on that and let you all know about the current status once I found out more.


    All the best


  • Jóhannes Birgir Jensson

    By Jóhannes Birgir Jensson

    Iceland is currently working on its Noise Directive Maps (until now they have been PDF-files) so I'm interested in this topic.

    Some data providers claim lack of knowledge of how to export noise calculations from their tools - are any guidelines available?

Statistics & Health

Statistics & Health

Join this group if you would like to share knowledge or ask questions regarding the INSPIRE implementation of Statistical Units [SU], Population Distribution (Demography) [PD] or Human Health and Safety [HH] data themes