Not for merge - Preview CMIP6 TAS Kerchunk

[abarciauskas-bgse]

No description provided.

[netlify]

✅ Deploy Preview for veda-ui ready!

Name	Link
🔨 Latest commit	3aafd38
🔍 Latest deploy log	https://app.netlify.com/sites/veda-ui/deploys/651ef730bfb0260008369d81
😎 Deploy Preview	https://deploy-preview-667--veda-ui.netlify.app/
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify site configuration.

[abarciauskas-bgse]

@j08lue @anayeaye @slesaad @brianmfreitag this is a preview of the CMIP6 visualization via Kerchunk/NetCDF for feedback.

Questions:

• I've only included the historical data. There are 4 SSPs (ssp126, ssp245, ssp370, ssp585) available for 2015-2100. Do we want to include one or more of those? If we want to include multiple SSPs we can easily publish separate collections. It will be slightly more complex if we want to publish multiple zarr assets within the same STAC collection. The simplest thing to do for now is determine if we want to include SSP data at all and if so, which SSP.
• Could you review text (dataset description and layer description) content or should we have someone from the CMIP6 NEX GDDP approve the content?

NOTEs:

• I'm trying to determine if we can mask out the no data values (i.e. water bodies), but as per usual the data encoding is giving some trouble
• I'm using near-surface air temperature and I've used the jet colormap with scale from 200,325. The data encoding issue below is blocking me from determining the absolute min,max of the data but this looks ok for now.

Let me know if you have other thoughts, but once we're happy with the dataset (STAC and MDX) configuration, I think the next step would be to publish to the staging STAC and add to veda-config.

[anayeaye]

@abarciauskas-bgse this preview is awesome!

RE

I've only included the historical data. There are 4 SSPs (ssp126, ssp245, ssp370, ssp585) available for 2015-2100. Do we want to include one or more of those? If we want to include multiple SSPs we can easily publish separate collections. It will be slightly more complex if we want to publish multiple zarr assets within the same STAC collection. The simplest thing to do for now is determine if we want to include SSP data at all and if so, which SSP.

Can we pilot the simplest case first (no SSPs)? It would be great if defer the SSP decisions another week so that we can discuss organization strategies for the incoming series of nex-gddp data which include many models and SSPs?

[abarciauskas-bgse]

Thanks @anayeaye. I have updated the preview with a new colormap and I think we are good to go since, thanks to @vincentsarago the nodata values have been masked. I think the next steps would be to:

1. Publish the CMIP6 dataset to the staging STAC
2. Add the notebook for how to generate CMIP6 kerchunk reference to veda-data
3. Add the notebook for how to generate CMIP6 STAC record to veda-docs
4. Add the data MDX file to veda-ui

@anayeaye @j08lue @slesaad do you agree we should put this demo in the dashboard?

[j08lue]

I'm all for adding this dataset to gain a high-value data source and a nice Zarr demo case.

Let's flesh out the details, though, please:

• Add more info about the dataset to the Dataset page - no need to have loads of flow text, but some explanation of the origin of this dataset, maybe some scientific references. Pls see the GHG Center dataset pages for content/formatting ideas.

  

• Add units (I guess it's K)

  

Can you provide this information, please, @abarciauskas-bgse?

[j08lue]

Looks great, @abarciauskas-bgse!

I made a few suggestions to edit the text to

1. clarify the use of the word "dataset" - both used for the original source ("NEX-GDDP-CMIP6") and for the VEDA copy of it
2. highlight more that this is just a demo subset - so users do not get surprised to find that it only covers 1950-1951.

Another question: I see you now point at the dev API - Is that necessary / do we not have the collection in staging? dev is probably a bit risky to use for anything user-facing, because we break it every now and then, right @slesaad?

[j08lue]

Suggested change

	This dataset represents a subset of the NEX-GDDP-CMIP6 data. This dataset was constructed to demonstrate a method for visualizing the data without creating any data copies just for visualization.
	VEDA hosts a subset of the NEX-GDDP-CMIP6 data, primarily for demo purposes. With this dataset, we demonstrate a method for visualizing the data without creating any data copies just for visualization.

[abarciauskas-bgse]

I made a small change, since we are not hosting any of the data files themselves on VEDA

[abarciauskas-bgse]

Thanks for looking @j08lue I think I have addressed your comments but I also think I could/should reach out to the dataset contacts Dr. Bridget Thrasher and Dr. Ian Brosnan, NASA NEX listed on https://www.nccs.nasa.gov/services/data-collections/land-based-products/nex-gddp-cmip6 to get their review and approval?

[slesaad]

@j08lue that's right. the next step, (also mentioned by aimee in the comment above) is to publish it to staging and create a PR in the veda-config repo. This is probably just an experimental PR.

[abarciauskas-bgse]

@slesaad is correct, this PR was so that we could easily preview what the data would look like before I go about publishing it to the staging API.

So the next steps are:

• I noticed that the spatial extent values in the dev collection are not what I expect them to be so I'm going to address that
• publish to staging STAC
• Add the STAC documentation + data processing to veda-docs and/or veda-data
• open a PR in veda-config that looks almost exactly like this PR

[abarciauskas-bgse]

closing via NASA-IMPACT/veda-config#323