How to handle RCP scenarios & Scenarios Convention Discussion

[rasherman]

We need a convention for how publications (e.g. images) connect to RCP scenarios through provenance. Both the structure of the connection (e.g. does it require an activity or reference?) and the right semantic link need to be determined.

Amrutha should come up with the initial proposed convention for discussion.

[lomky]

We are going to use the GCIS Scenario for this.
We acknowledge it may be valid in the future for Scenarios to encompass other Scenarios.
For the provenance connection Figure/Finding/Table/Chapter/etc -> Scenario will be prov:wasDerivedFrom

Take away todos:

• create the two scenarios (@amruelama)
• plan how to connect them for NCA4 (@lomky create ticket)

[rasherman]

Scenarios are only created when used in USGCRP reports. These will be agreed-upon scenarios for use in the assessment process, so should be known ahead of time. Can be used for a sub-scenario (topic-specific projection) or comprehensive umbrella scenario.

Convention for internal metadata:

• Identifier: GCIS creates the identifier. Use Scenario name/abbreviation as possible. For example, sres_b1 for the B1 Scenario from the SRES set of scenarios, or rcp_4_5 for Representative Concentration Pathway 4.5. If no abbreviations are commonly used, use full name as in organization. If scenario has multiple versions, make sure version is noted in identifier and that record is unique to one version.
• Description: Should be copied from official source as used by USGCRP, and should match what is found at the link in Description Attribution at time of record creation. Can choose to only excerpt introductory paragraph as appropriate.
• Name: List complete name as used by USGCRP and listed at the Description Attribution.
• Description Attribution: URL of website describing the scenario hosted by the organization responsible for the scenario.

In future, we probably want other fields similar to those in Dataset, namely "Date Published" and "URL".

[rasherman]

Other conventions:

• We are not concerned with the provenance of the scenarios.
• Distinction between a projection dataset and a scenario is that a dataset has probability ascribed to possible futures (e.g. error bars). A scenario is a projected future taken as an assumption.
• If the scenario source url has a representative image that can be captured and attached to record as a file, do so. If no image exists, leaving file blank is accepted.

[amruelama]

Here is the link to the RCP scenarios created - https://data-stage.globalchange.gov/watch?limit=4&t=scenario for review @rasherman