Start making RRFS_A Ensemble graphics.

[Brian-Jamison]

----- oct 25

• Fixed the colorbar error. Created another error, but figured it out with Craig on Slack. Indexing issues with <9 members.
• Need to fix that error.
• Will follow up with whether there should be a PR for the changes.

---- Aug 30

• Need help with a weird error on the color bars.
• Already deployed in real time. Still needs a PR to pygraf to add the changes.

---- Aug 16

• Nick K. will pull and put on public, make a soft link to ctl directory to link to member 0.
• Now, can wire in the new directory path.
• 60 hours -- 43 hours successful. Memory error crash at 44 hours. Split in workflow 0-29, 30-60.
• Need to add a couple of workflow things for the live data and shorter forecast length.
• Only 5 single-physics members are currently running.

----- July 19

Testing to see if it can go out to 80 hours. Filled up home space.

Moved the clone and logs over to lfs and linked back to home.

One issue: needs forecast hour variable. Had to include fcst_hr in coord_dims variable. It doesn't have lifted parcel level, so cape products were unstable.

---- Jun 7

Setting up the Rocoto XML for the data transfer of RRFS_A. Files are being transferred to Jet from the cloud -- Trevor and Nick K. are working on that transfer.

---

Slack exchange:

Terra Ladwig
1:33 PM
Are we making any RRFS_A ensemble graphics?

Curtis Alexander
1:35 PM
The GSL DSG just got the RRFS_A ensemble grids flowing onto /public yesterday afternoon from AWS, so I’m guessing not yet, but they are now in position to do so.
Here’s the path according to them: /lfs4/BMC/public/data/grids/rrfs_a/conus/{natlev,prslev}/{single_physics,multi_physics}/mem000N
Verification and graphics teams should be able to start ingesting/viewing these

Terra Ladwig
Okay, cool. The HWT pages only show the 0 and 12UTC data. I asked for them to add 6/18 UTC, but not sure if they will or not.

Curtis Alexander
Yeah, we should try to plot 0/6/12/18 to 60 hrs if possible 😉
Hopefully we have the graphics templates from the last year when we ran ensemble forecasts that can be re-deployed here.

[terraladwig]

Nick (DSG) saved a set of ensemble data. See below message pasted from email on June 15.
Please use this example data to determine the minimum number/type of files required to make ensemble graphics. Then we can report that info back to Nick, which will enable real-time RRFS_A ensemble graphics. Note: the example is CONUS. RRFS_A will be on North American domain soon (mid-July).

"The full 6/13 00Z ensemble is available at /lfs4/BMC/public/retro/rrfs_a/rrfs_a.20230613/00/{single_physics,multi_physics}/mem000{1..9} for the graphics/verification teams to explore."