CMOP SELFE model from THREDDS is not on IOOS EDS Model Viewer

[emiliom]

From @cseaton, 4/17:

I just checked, and the CMOP model isn't in the EDS model viewer .... I'll mention it to Antonio. Obviously, CMOP attribution would be nice.

I haven't checked what attribution looks like in the EDS Model Viewer. I hope it reuses a lot of the metadata from the THREDDS dataset, and that our improvements to that metadata can lead directly to improved and clearer attribution. For metadata improvements previously identified (and/or implemented, see isssue #5 and possibly the older issue #1)

Note: the OSU ROMS model is in the EDS Model Viewer. Would be good to see how EDS exposes model attribution for OSU ROMS.

[emiliom]

Note: the OSU ROMS model is in the EDS Model Viewer. Would be good to see how EDS exposes model attribution for OSU ROMS.

Adding a follow-up about the attribution exposed by the EDS Model Viewer for OSU ROMS. I expect to have it improved soon.

[emiliom]

EDS Model Viewer Github issue for discussing adding the CMOP SELFE model: https://github.com/asascience-open/IOOS-EDS/issues/73

Also, following up on this earlier comment:

Would be good to see how EDS exposes model attribution for OSU ROMS.

The answer is that we have to be proactive to request a model summary/abstract. We've now done that for OSU ROMS and LiveOcean.

[emiliom]

@cseaton, Kathy Bailey (IOOS) said just now that they'll like to follow up and and CMOP SELFE to the EDS model viewer. We already have OSU ROMS and LiveOcean in there, with reasonably good attribution (OSU ROMS could be improved with PI involvement) and overall presentation.

I'll ping you on that github issue in a bit, so you can engage directly in the discussion.

[emiliom]

@cseaton: The EDS team hasn't notified us that they have the CMOP model on their staging EDS. Still, let's get ahead of them if possible.

Once it's on their staging you'll want to review it before giving the ok to go live.

Besides that review (and answering their questions), here's our to-do's:

• Come up with a brief abstract/description. See what we've put together for the two other NANOOS models on EDS: Go to https://eds.ioos.us, type NANOOS in the Data Layers search box and hit return. Hover over any of the layers for one of the models, and click on the "i" icon that comes up on the right. That'll bring up the abstract. Now do the same thing for the other model to see its abstract. Note that the "NANOOS West Coast ROMS" abstract is overly long, due to an excess of caution (Craig and I adapted from what's on NVS, which is too long ...). I'd shoot for something closer to LiveOcean
• Ensure layer names are consistent with other NANOOS models. For CMOP, a suggested name could be "NANOOS Columbia Estuary (CMOP SELFE)"

I suggest you start drafting the abstract soon. Then share the draft with me on this github issue (using markdown as needed), then once we have the final version we'll share it with the EDS team via their github issue.

[emiliom]

@cseaton, pinging you again to encourage/prod you to draft the brief model asset abstract/description ...

[cseaton]

Here is a draft of a minimal entry:

The experimental nowcast and forecast fields displayed here from the CMOP Virtual Columbia River are produced by a computer model (SELFE) of the Columbia River estuary and the Oregon and Washington coastal ocean circulation, with a primary focus on the Columbia River estuary and the Columbia River plume. These forecasts are updated daily at 9:00 am. They are an experimental product intended for research use only. For more information, please visit the CMOP Virtual Columbia River forecast.

I agree with your suggested layer name.

[emiliom]

Great, thanks! I've passed them on (as you will have seen)