Fix missing objects: 5 objects in layer nrm2017_qld2019 (cl10947) are present in geoserver but not in the spatial portal api

[cofiem]

Hello,

I maintain an instance of biocollect and ecodata which uses the ALA's spatial server.

I'm trying to resolve an issue with missing objects in the layer nrm2017_qld2019 (cl10947). The objects all show up on the ALA's spatial portal, but in biocollect there are some objects missing on the 'globe' dropdown in the maps when creating or editing a project or site.

I've looked into this, and the issue seems to be that the geoserver contains the 61 expected objects, but the spatial server api is missing 5 objects. Biocollect uses the spatial service api, so the items are missing in biocollect.

I've compared these endpoints:

• https://spatial.ala.org.au/layers-service/objects/cl10947
• https://spatial.ala.org.au/ws/field/cl10947?start=0&pageSize=500
• https://spatial.ala.org.au/geoserver/wfs?service=wfs&version=2.0.0&request=GetFeature&typeNames=ALA:nrm2017_qld2019&outputFormat=application/json&propertyName=NRM_REGION

The missing items in the two spatial server API, but present in the geoserver API are:

• Cape York (id 7)
• Norfolk Island (id 27)
• South Australian Arid Lands (id 43)
• South Australian Murray Darling Basin (id 44)
• South NRM Region - Macquarie Islands (id 42)

I've compared this to the older NRM region nrm_regions_2017 (cl10930). It has 63 items in all the three endpoints, which is what I expect.

Could you please look into this? I think the three APIs should return the same results, which should include the missing 5 items. There should be 61 objects in total.

[nickdos]

@adam-collins could you take a look at this please?

[ansell]

For reference, biocache is showing the expected 61 objects in facet queries for cl10947:

https://biocache-ws.ala.org.au/ws/occurrences/facets/download?q=cl10947%3A*&facets=cl10947&count=true&lookup=true

[adam-collins]

61 objects now in

• https://spatial.ala.org.au/layers-service/objects/cl10947
• https://spatial.ala.org.au/ws/field/cl10947?start=0&pageSize=500

The problem was that these areas had self intersecting polygons and were excluded by the validity test.

[cofiem]

I've tested this out, and it looks to be fixed. Thanks very much! Is there any chance of the areas being excluded again in the future?

[adam-collins]

I am currently reviewing the process for this issue.

This issue will remain open until the underlying issue is resolved.

[adam-collins]

This will not be an issue for future layers.

The CAPAD 2010 Terrestrial was the only other layer with an issue. It is now consistent.

• The aggregation on PA_ID instead of NAME in the web services other than geoserver caused 5 areas to merge into 1.
• The aggregation on TYPE was case insensitive in the web services other than geoserver causing 2 areas to merge into 1.