Changes requested by Ward Appeltans

[gbif-portal]

Changes requested by Ward Appeltans

changes to this page? https://www.gbif.org/network/2b7c7b4f-4d4f-40d3-94de-c28b6fa054a6

1/ replace Biogeographic to Biodiversity in the name 2/ replace the link to the OBIS website to https://obis.org/about/ (we moved from iobis.org to obis.org)

Some OBIS nodes may ask you to correct their names as well.

I also hope we can get the numbers correct, because 9M records is way too low.

---

Github user: @dschigel User: See in registry System: Chrome 83.0.4103 / Windows 10.0.0 Referer: https://www.gbif.org/network/2b7c7b4f-4d4f-40d3-94de-c28b6fa054a6 Window size: width 1745 - height 970 API log&_a=(columns:!(_source),index:'prod-varnish-',interval:auto,query:(query_string:(analyze_wildcard:!t,query:'response:%3E499')),sort:!('@timestamp',desc))) Site log&_a=(columns:!(_source),index:'prod-portal-',interval:auto,query:(query_string:(analyze_wildcard:!t,query:'response:%3E499')),sort:!('@timestamp',desc))) System health at time of feedback: OPERATIONAL

[dschigel]

Issue as requested by e-mail from Ward @timhirsch @kcopas @ahahn-gbif

[kcopas]

Name changes in Contentful and registry (awaiting refresh from the latter?)

Link corrected in text from Contentful.

[ahahn-gbif]

The network page did catch up on the update now (https://www.gbif.org/network/2b7c7b4f-4d4f-40d3-94de-c28b6fa054a6).

Concerning the record numbers: a few years back, we deleted a large batch of records/datasets on request from OBIS, concerning mostly EurOBIS. As these datasets are re-registered and added to the network entry, we should also see the numbers go up again. A check for newly registered datasets does not show any more recent additions pending though.

[MattBlissett]

The "Node" links ("AntOBIS" etc) go to their old website, and now just redirect to the home page of the new one.

I removed the hidden, old logo, which was causing the security warning on this page.

[pieterprovoost]

@MattBlissett Here's an up-to-date list of OBIS nodes: https://api.obis.org/node

[ahahn-gbif]

No action point, just for some clarity as we have different components here:

• the network page (https://www.gbif.org/network/2b7c7b4f-4d4f-40d3-94de-c28b6fa054a6) is static, authored content (content management system), and as such needs an editor to update information. We recognize it got a bit stale and needs to be updated - checking options for a quick turnover to curate
• the "network" tab behind that page (https://www.gbif.org/network/2b7c7b4f-4d4f-40d3-94de-c28b6fa054a6/dataset) is based on an explicit mapping of datasets to the network, curated in a table in the registry. Again, we would like to find better and more dynamic ways to curate. Due to the concept of a "network" (bracketing individual datasets across a range of publishers, but usually only a subset of their datasets), this mapping has to be done per dataset, not per GBIF publisher (= OBIS node)
• while OBIS nodes can be listed in the authored content of the network page without being registered with GBIF, the linking of datasets to the network requires that the OBIS Node is registered as a publishing institution with GBIF, and publishes the datasets to GBIF. GBIF-external datasets cannot be shown here.

[dnoesgaard]

I can easily import the data on nodes from the API, e.g.

Name	Type	Theme
AfrOBIS	regional	Sub-Saharan African OBIS node
Antarctic OBIS	regional
Arctic OBIS	regional
[Caribbean OBIS]()	regional
ESP OBIS	regional	Tropical and Subtropical Eastern South Pacific OBIS node

but linking to GBIF publishers is manual and time-consuming. Suggestions?

[ahahn-gbif]

For the OBIS network page, the list of Nodes is the relevant information. DP will then need to link datasets (not publishers) to the OBIS network table - if you extracted the list of publishers anyway, we would be happy to take that on from there. Yes - manual and time-consuming.

[dnoesgaard]

In that case, is the above format ok—Name, Type, Theme?

[ahahn-gbif]

Is it possible to continue with the same information / structure that is currently showing under "OBIS nodes in GBIF" at https://www.gbif.org/network/2b7c7b4f-4d4f-40d3-94de-c28b6fa054a6?

[dnoesgaard]

Not if I'm to use the OBIS API. The current list was (painfully) put together manually. Creating one from the API, however, takes 10 seconds...

[ahahn-gbif]

I see, thanks. Can we then update the table based on the OBIS API, please, and maintain backup of the previous version? I agree that the manual edits are not sustainable in the long run

[dnoesgaard]

Done.

Here's the previous version for your records:

Region	Node	Host	GBIF publisher
Antarctica	AntOBIS	Biodiversity and Ecosystems Data and Information Centre (BEDIC), Royal Belgian Institute of Natural Sciences	SCAR - AntOBIS
Arctic	Arctic OBIS	Conservation of Arctic Flora and Fauna (CAFF), Iceland	Conservation of Arctic Flora and Fauna
Argentina	ArOBIS	Centro Nacional Patagónico (CENPAT) CONICET	ArOBIS Centro Nacional Patagónico
Australia	OBIS Australia	Commonwealth Scientific and Industrial Research Organisation (CSIRO)	Australian Ocean Biogeographic Information System (OBIS Australia)
Black Sea	OBIS Black Sea	Ukrainian Scientific Centre of Ecology of the Sea	[]()
Canada	OBIS Canada	Centre of Marine Biodiversity, Bedford Institute of Oceanography	Canadian node of the Ocean Biogeographic Information System (OBIS Canada)
Caribbean	Caribbean OBIS	Universidad Simón Bolívar - Centro de Biodiversidad Marina	Caribbean OBIS Node
China	OBIS China	Department of Marine Organism Taxonomy & Phylogeny, Institute of Oceanology, Chinese Academy of Sciences	[]()
Colombia	OBIS Colombia	The Marine and Coastal Research Institute of Colombia (INVEMAR)	[]()
CPPS - South Pacific	OBIS CPPS	Comisión Permanente Pacifico Sur	[]()
Europe	EurOBIS	Vlaams Instituut voor de Zee (VLIZ)	[]()
Indian Ocean	IndOBIS	Centre for Marine Living Resources and the Environment, Ministry of Eart Sciences, India	[]()
Japan	OBIS Japan	Japan agency for Marine-Earth Science and Technology	[]()
Korea	KOBIS	Korea Institute of Ocean Science and Technology	[]()
Malaysia	OBIS Malaysia	Institute of Oceanography and Environment (INOS) of the University of Malaysia, Terengganu	[]()
Mediterranean	MedOBIS	Hellenic Centre for Marine Research	Hellenic Center for Marine Research
Persian Gulf and Gulf of Oman	PEGO-OBIS	Iranian National Institute for Oceanography (INIO)	[]()
Senegal	OBIS Senegal	Centre De Recherches Océanographiques Dakar-Thiaroye (C.R.O.D.T)	[]()
South-East Asia	SEAOBIS	ASEAN Centre for Biodiversity	[]()
South-West Pacific	South Western Pacific OBIS	National Institute of Water & Atmospheric Research	Southwestern Pacific Ocean Biogeographic Information System (OBIS) Node
Sub-Saharan Africa	AfrOBIS	Southern African Data Centre for Oceanography	[]()
Tropical and Subtropical Eastern South Pacific	ESPOBIS	FONDAP COPAS	[]()
Tropical and Subtropical Western South Atlantic	WSAOBIS	University of São Paulo	[]()
UK	OBIS UK	The Marine Biological Association of the United Kingdom	[]()
USA	OBIS-USA	US Geological Survey	[]()

[wardappeltans]

Can GBIF and @pieterprovoost think of an automated way to assign the "OBIS network" to datasets from OBIS nodes?

[ahahn-gbif]

I suspect this will remain semi-automatic, but to make the process smoother, it would be helpful to have a list of those GBIF-registered entities (publishers or installations: name and UUID) under which we can assume all published datasets to be OBIS-datasets. We could then define a process to trawl these entities periodically, and add all datasets to the network that are new since the last check.

For EurOBIS, for example, we would need to manage the information that the GBIF entry corresponding to the OBIS Node is called "Flanders Marine Institute" (https://www.gbif.org/publisher/2089ce96-4fb5-4a20-999c-3ccf45a27a4d), and is of type publisher, but that only those, and all those, datasets behind the "European Ocean Biodiversity Information System (EurOBIS)" installation (https://www.gbif.org/installation/f9df6cff-21d4-4d4b-8da4-eda454f5a086), not the other one, should be hooked up to the network.

Also keep in mind that this kind of trawl would not work for entry points that serve both OBIS and non-OBIS datasets through a single installation, if such cases exist - these would still require individual, explicit network registration.

[wardappeltans]

@ahahn-gbif, i was more thinking like a script that compares the DwC-A URLs in OBIS with GBIF via the OBIS API? But not sure if that is possible...

[MattBlissett]

but linking [OBIS notes] to GBIF publishers is manual and time-consuming. Suggestions?

To do this properly, you'd add an additional identifier to the publishing organization, as I've done here: https://registry.gbif.org/organization/104e9c96-791b-4f14-978c-f581cb214912/identifier . You can then search for organizations by identifier: http://api.gbif.org/v1/organization?identifier=dc6c6ea2-83f5-4b18-985a-9efff6320d69

i.e. there's manual / semi-manual work to link each publishing organization the first time (either GBIF or OBIS must make a link where currently none exists).

---

I was more thinking like a script that compares the DwC-A URLs in OBIS with GBIF via the OBIS API? But not sure if that is possible...

That's certainly possible as a one-off check or a script -- I've sent a list of omissions to the GBIF helpdesk address.

[kcopas]

Just want to flag the arrival of a new Caribbean OBIS dataset here, in the event it helps this process.

h/t GBIFDodo

[ManonGros]

@ahahn-gbif could you check if we can close this issue?

[ahahn-gbif]

The initial issue (update of the network page) has been dealt with. Suggest to open a new issue for outstanding workflow items on the dataset-network association, rather than continuing here (closing).