DOI - Digital Objective Identifiers for ERDDAP data sets

[John Wilkin]

Does anyone have experience registering a DOI for an ERDDAP dataset? 

JPL PO.DAAC have started creating stable DOI's for many of their datasets, including near-real-time data sets that are being continually added to. This makes it easy for someone to unambiguously reference the data sets they used in a project.

For example, if you google 10.5067/GHGMR-4FJ04 you will get the PO.DAAC end point for the data description page for the MUR SST product. 

My understanding is that NASA/JPL have paid for the privilege of minting their own DOIs. 

I've tried working through my own university's library (via their membership in DataCite.org) to see if they would register a DOI that points to an ERDDAP end point that I maintain, but they won't because they don't control the page it points to. Ideally, I would have control over what service the DOI points to in the event I rename or move the data to a new location. That's kind of the point of a DOI.

I wonder, might NOAA CoastWatch consider doing this for data sets they host? 

[Rob Fuller]

Hi John,

I have no experience with this but doi.org might be an option, I see some organisations are using that one.

http://erddap.com/#search=DOI.org

Kind regards,

Rob.

--
You received this message because you are subscribed to the Google Groups "ERDDAP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to erddap+un...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/erddap/dd3779b5-311b-402d-89a3-355f56b147bcn%40googlegroups.com.

[Bob Simons]

I don't have experience minting DOIs. My understanding is: NOAA's NCEI has taken exclusive control of minting DOIs within NOAA. Other groups in NOAA are not allowed to do it. So NOAA Coastwatch and NOAA NMFS SWFSC ERD are not allowed to do it.

The other problem that you mentioned is the issue of whether the dataset is static / finished / unchanging. The DOI-related people I talked to say the dataset must be completely finished and thus unchanging. They said DOIs must not be minted for datasets where data is still being added, even if the data won't ever change after it has been published. To me, this is an unwarranted requirement. It means some datasets (for example, satellite datasets) won't have a DOI until the mission ends after some number of years, and others (for example, the NDBC buoy data) will never have a DOI since the project will never end. Apparently, JPL has decided to push against this boundary -- but they're NASA so people are unlikely to contradict them. To me, DOI requirement goes against the spirit of DOI's: to facilitate the citation of data. I have suggested to NCEI people before that, if this is the case and can't be changed, it would be nice if there were a new OngoingDOI which could be used for ongoing datasets. I think they don't care partly because they almost exclusively see finished datasets so it isn't an issue for them. It won't be easy to make that happen.

I wish I had a better answer.

Best wishes.

[Cooper Van Vranken]

Hi John,

We got a DOI for the eMOLT fishing vessel bottom temps that you've been working with via SEANOE: https://www.seanoe.org/data/00642/75397/

No cost at all to put it up there, and they pull from the ERDDAP, we just included the link in the setup info word doc they request.

Quite easy and they helped me through the process for the first data sets we put on there.

Hope this helps,

Cooper

--
You received this message because you are subscribed to the Google Groups "ERDDAP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to erddap+un...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/erddap/ef2db448-5efa-4033-81e0-9366d2f19b66n%40googlegroups.com.

[crid...@oceannetworks.ca]

Hi all, 

I just wanted to clarify that, yes, DOIs traditionally were only assigned to static datasets, but Ocean Networks Canada (my organization), just completed a project: MINTED: Making Identifiers Necessary to Track Evolving Data. We have networked ocean observatories in the pacific ocean, and so we are continuously receiving data from deployed instruments. I don't believe that it was a rule, it's just much more complex to assign DOIs to evolving data. Below is a link to our user documentation, to give an idea of our process.

We used DataCite, and I know that might be outside the scope of your question, but I wanted to clarify that there is work being done on dynamic data citations. 

If you have any questions, please don't hesitate.

https://wiki.oceannetworks.ca/display/DP/Data+Citations 

Cheers, 

Chantel

[crid...@oceannetworks.ca]

Hello all, 

My project minted DOIs with DataCite, and now we would like to attach them to the respective ERDDAP records as well. I see that ACDD suggests using the identifier global attribute field, but unfortunately that angle doesn't work for us. We did some additional investigating, but we did not see any conversation or work being undertaken on this subject. We then looked within ERDDAP to see what others are doing with their DOIs and learned that two well-established, and widely respected, organizations (Scripps and Biological and Chemical Oceanographic Data Management Office) have created a new 'doi' global attribute, and included the existing DOI in this field:

• Scripps https://spraydata.ucsd.edu/erddap/info/binnedCUGN80/index.html 
• BCO-DMO https://erddap.bco-dmo.org/erddap/info/bcodmo_dataset_555666/index.html 

Ocean Networks Canada is wondering if anyone knows of any work being done on best practices, or even suggested practices, on this topic: how to incorporate DOIs into a record, should there be a link enabled (i.e., 10.34943.... or https://doi.org/10.34943....), should a link to the DOI landing page be separate from the DOI itself? (i.e., link to the landing page with an 'additional metadata' attribute?), etc. If so, if you could point us the direction of the people and/or organizations undertaking this, we would be very appreciative. If not, we are wondering if this is a discussion that should begin, since the use of DOIs for data is increasing. 

Thank you in advance for your help, 

Chantel

Chantel Ridsdale, MLIS | Data Steward
Ocean Networks Canada | oceannetworks.ca
University of Victoria Queenswood Campus 
#100-2474 Arbutus Road, Victoria, BC V8N 1V8 
A UNIVERSITY OF VICTORIA INITIATIVE  

[felimon....@gmail.com]

Chantel,

 

We also had some internal discussions about this in preparation when we can have DOIs minted. Nothing really stops you from creating another global attribute. However, we think that the full URL should be provided if a global DOI attribute is introduced, instead of simply the DOI. In the BCO-DMO example, it may be better if the full DOI URL, i.e., https://doi.org/10.1575/1912/7227 is provided instead and this should resolve to the registered landing page.

 

My thoughts on the subject matter,

 

Felimon Gayanilo

 

Harte Research Institute for Gulf of Mexico Studies

Texas A&M University-Corpus Christi

Office: 361-825-3454  Cell: 305-335-1584

To view this discussion on the web, visit https://groups.google.com/d/msgid/erddap/d83c0c31-123f-4e96-88ee-32118b2f84f8n%40googlegroups.com.

[crid...@oceannetworks.ca]

Thank you Felimon, 

We were thinking the same thing, but couldn't find an example fo this implemented. Thanks for sharing!

Cheers, 

Chantel