European Commission logo
INSPIRE Community Forum

Series and dataset linking and INSPIRE Geoportal

  • General

    The INSPIRE Community Forum is now part of the INSPIRE helpdesk system, based on GitHub (, so please start there any new discussions.

    The current platform will remain available in read only mode in its current form and URL address until 31/01/2021, then its content will be archived and then published in a different URL which will be announced when it becomes available.

Dear all, I would like to rise a question of how to deal with the linking of dataset series and its datasets, especially from harvesting and the INSPIRE Geoportal point of view.

My colleague and I have already pointed out this problem during a presentation at the MIG-T meeting in Prague:

In the case of the Czech Office for Surveying, Mapping and Cadastre we provide 13 INSPIRE datasets series that have a national spatial scope, but they do not have their pre-defined file. Instead, they consist of almost 75 000 datasets with pre-defined files. Because of the huge amount of datasets, we have decided to send only the series metadata to the INSPIRE Geoportal. But it means the user does not have any information about the datasets, from which the series consists. This situation also cause the data does not to appear downloadable according to the geportal.

In the case of our office we have needed to get at least information about dataset identifiers into the INSPIRE Geoportal to data appears downloadable. We have found a way to get this information into the geoportal (see section below), which works, but from our point of view it would be useful to have a more complex solution.

Our current solution:

We have already successfully tested this solution.

Its schema:


There is only one metadata record for a dataset series and no separate metadata records for datasets on the INSPIRE Geoportal. The identifiers of all datasets are stored in the series metadata as another resource identifier (//gmd:MD_DataIdentification/gmd:citation/gmd:CI_Citation/gmd:identifier). Linking between the dataset series and its view service is done by a coupled resource (srv:operatesOn) in the service metadata and a metadataURL element in the service capabilities. Both are pointing to the dataset series metadata. Connection between the series and its download service is done in the same way as for the view service. But in addition, SpatialDatasetIdentifiers in the capabilities contains series resource identifiers as well as datasets. So the SpatialDatasetIdentifier is multiplied by the number of datasets plus one for the series identifier.

This approach was tested for the following data. For testing purposes we simplified the use case - the series consists of only 10 datasets. Maximum file size of the dataset is about 400 kB.

Title: INSPIRE téma rozšířené parcely (CPX) [in English: Cadastral Parcels Extended (CPX)]
Series identifier: CZ-00025712-CUZK_SERIES_CPX
Datasets identifiers: CPX.SD.697893, CPX.SD.930431 … (they start with CPX.SD.xxxx)
Series metadata:

-    Is it possible to unify provision of series and its datasets across member states?
-    Is it necessary for INSPIRE to harvest metadata for all datasets into the INSPIRE geoportal, if there is metadata for its series? If yes, what should be a linking point between series and its datasets? (A use of some ISO 19115 elements; identifiers...?) If not, how can we integrate the information about datasets into series metadata?
-    How should the INSPIRE geoportal (visually) connect series to their datasets?  
-    Is it necessary (and feasible) to have a standard territorial division of series to datasets? (given that we do not apply this within one organisation)

Thank you in advance for your opinion.

  • Katalin TOTH

    Dear Veronika,

    Robert Tomas drew my attention to your post. I am currently working on tehcnical guidelines for the IACS community that is going to to share the spatial data collected under the Common Agricultural Policy. Our stakeholders ask for a recommendation - what  is "better" to apply datasets or dataset series for the scope of metadata. As there is no simple answer, partly inspired from your post I  put together some aspects in a document - among them is one of linking services and the metadata. Another important aspect is to keep metadata at the highest possible level - this is what ISO 19 115 recommends too. I also summarised the decision workflow that is below.



  • Katharina SCHLEIDT

    By Katharina SCHLEIDT

    Dear Katalin,

    one question to your first condition - are you sure you have the negatives straight on this one?

    I believe it would be more understandable if you reformulated condition 1 to the positive as follows:

    1. The datasets to be published have a common heritage

    To my current logic, the conditions stay correct with this reformulation (as my feeling is telling me that it's currently backwards - if the datasets have a common heritage, they automatically become a series?)