Date

24 Oct 2019

Attendees

•  Nathan James, Elli Pauli  Lalit Wanchoo 

• ASDC -  user-99280            

  GES - Dana Ostrenga,

          ORNL - Daine Wright, Tammy Walker

Goals

Background

• Discuss ideas of DOI evolution system to hep map out a path forward 

Discussion items

Agenda

1. Work flows of DOI at the DAAC at a high level. 
   • ORNL, ASDC, and GES DISC presented their workflow for handling DOIs and discussed the differences between them
     • ORNL -
       • Reviewed their workflow - DOI_Registration_Process_at_ORNLDAAC.pptx
       • Key point - Dataset landing page and DOI landing page are the same
     • ASDC 
       • CAMP - science provider creates a product record
       • approvse record
       • Workflow starts for directory
       • Science team put metadata in to record for testing 
       • Make world visible - create LP and DOI registered as one of the data services for quality data product
       • DOIS go to product Landing page for older satellite products
       • But for airborne data products - not necessarily the same
       • CAMP will be creating dynamic landing page
       • decommissioning process - turning of LPs for DOI
       • Using delete capability to clean up reserve DOIs
       • Some data products for complete reprocessing with TES being decommission
       • Diff from ORNL who starts the mission process when the data is available - so identifier is tied to the DOI-name - so opaque ID vs name that has meaning.
       • Science team members like intelligent filenames
     • GES DISC
       • Reserve through DOI system prior to getting the data
       • Program creates opaque DOI and sends to PI to their mission dataset w shortname
       • Sometime they’d like intelligent DOI names
       • Once reserved - create collection level metadata then dataset LP are created from CMR in collection-level metadata so you can have resolving LP but not created into dataset published to CMR (LPs created dynamically)
       • DPS is SAuS like but moving to have metadata inside an internal tool.
       • Would like to have a more automated process like ORNL to the DOI - very manual right now.
       • When we deprecate or decommission, we update collection level metadata to capture prior version and update DOI to point to latest version.
       • Single-point of contact for DOIs for 35 missions - need more automation or more folks to reserve and register DOIs.
       • GES MISSION LEADS SHOULD HAVE DAAC BLESSING AND our system will allow submissions.
       • If not doing automated fashion, make sure you have page - through tool, PI could state how they want page formatted.
   • Each DAAC has some very unique and useful approaches to processing DOIs
     
     • Is there a way to combine the various processes to make one global, highly sophisticated method?
       • Concern is pushback from the scientists, depending on whether the data is opaque or non
2. What are some show-stoppers in adopting other DAACs DOI processes?
   1. Opaque DOIs would be a game changer for ASDC - 6 new missions - 2 missions are not familiar with processes but 4 new ones (would be easier to get them on board)
   2. GES does both ways so they RESERVE ahead of time so DOIs can be available ahead of time for publishing research.
   3. ORNL Question for Lalit - metrics pulled from DOI database or DataCite - Ans: Database  - Have you compared with DataCite metrics? Not yet.
   4. ORNL - Different DAACs do the same thing at different points in the process - need some sort of reservation process early on with unique DOI string.
3.  CAn EarthdataPub be used to automate some this process?
   • DOI are not part of EDP and is not a requirement
   • However, it does support modules such that DOIs can be plugged into the Framework
   • ORNL: Common DOI Registration tool could be used outside of EDPub
4. Anything from each other's process that might cause an issue for anyone?
   • Appears they all do the same thing, just at different points in the process - so - no