Layer names in view services for Population distribution?
The INSPIRE Community Forum is now part of the INSPIRE helpdesk system, based on GitHub (https://github.com/INSPIRE-MIF), 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.
Tags:
Hello, we are trying to configure INSPIRE compliant view services for population distribution (population in 1 x 1 km grids) and we are struggeling to understand the naming conventions to be used for the layers in the service. When we validate our view services, the ETF validator complains about the name element of the layers in the WMS because according to the validator the names does not follow specs. But really, we cannot find any guidance in the TG or in the INSPIRE layer register under PD. Hence, it’s a bit of catch 22 here, our services does not follow the specs but there seems to be no specs to follow? Could anyone better informed help out here? We need to set up services for total population, population by sex and 5year groups.
Hello Jerker!
My guess would be that the lack of a layer name for PD is the result of the PD theme not carrying geometry itself but deriving it from the SU (statistical units) theme. Therefore a valid geographical reference for PD data would be a SU.VectorStatisticalUnit or SU.GridStatisticalUnit layer.
The human health theme (somewhat similar to PD) has a HH.HealthStatisticalData layer which is of HealthStatisticalUnit type, so I would carry the analogy to the SU-PD tandem.
Miro
Hello Jerker and Miro,
The suggestion of Miro might be a good one, although it could be confusing with the publication of the already existing
SU.VectorStatisticalUnit or SU.GridStatisticalUnit layers.
We have chosen our own layer naming in our recent PD and HH publications because we did not know of a standard.
We recently published the following layers for PD:
The last one is the same as publication for Geostat3, but will be replaced by census 2021. Metadata in English can be found here: https://www.nationaalgeoregister.nl/geonetwork/srv/dut/catalog.search#/metadata/76541ebb-a554-4540-a017-399d4bb4a860
And we published the following layers for HH:
Metadata link: https://www.nationaalgeoregister.nl/geonetwork/srv/dut/catalog.search#/metadata/12bb57cd-73e0-41ba-a1e9-6d986b9109d3
The NUTS2 publications are the same ones as published at the Eurostat SDMX services. We joined them with SU.VectorStatisticalUnit to create a WMS and WFS
Pieter
Hi!
we also use our own layer naming, imitating the naming of the other themes: PD.StatisticalDistribution
We have also asked our national contact person (validator/JRC) to give feedback to the JRC (errors in validator when layer naming not defined)
For the health theme we use the HH.HealthStatisticalData layer.
Sibylle
Thank you very much for your input and suggestions! We are probably going for a solution similar to what is suggested by Pieter and Sibylle. This seems to be the only reasonable solution considering the conflict Pieter described with the already published SU services.
Jerker
Regarding the PD:StatisticalDistribution: this is actually the only featureType in the model but according to the model, its geometry is the geometry of the whole distribution. This means that if you supply data for the whole country, the outer country boundary should be the only geometry provided for the StatisticalDistribution featureType, regardless of the number of values distributed.
It's the biggest cognitive problem with the PD model since the very beginning and JRC is aware of it (it has been repeatedly reported), however no solution has been produced to date.
Following up on Pieter's comment, I think it's a good idea to use your own naming convention for PD. As mentioned before - PD initially has been designed as a theme without geometry, so this might be the reason for not having an official layer name.
Miro