iLAMB with zppy generated no figures #523
Comments
xuezhengllnl
changed the title
|
The output file begins with: => I think you would need to have a => Add |
|
Thank you for trouble shooting @forsyth2 ! |
Thanks @forsyth2, I added these settings to the cfg file (see the updated cfg below). The generated ilamb link still has no result. |
|
Thanks @xuezhengllnl, I will try to look into this more this afternoon. Sometimes coming up with the right parameter settings can be tricky. |
|
@xuezhengllnl Hmm you're actually making it further than me. I'm running the following, edited-down version of your linked cfg: And getting this error on |
|
@xuezhengllnl I tried running again with exactly your same cfg (except I will try to keep debugging. |
|
@xuezhengllnl Sorry, I haven't had a chance to do more debugging yet, due to the upcoming E3SM Unified patch release. Have you had a chance to check if you did link your latest cfg? I'm not sure why I'm getting different results than you... |
|
@mahf708 @chengzhuzhang If one of you has spare time, can you try running the cfg above (#523 (comment), you'll need to change the |
|
I have submitted a zppy run with Xue's config, will update the my finding aft the run finish... |
|
@forsyth2 the error you reported is caused by missing land variables that This would by pass the error you had, but new error occur: The tri-grid dataset does not have |
|
I think we should add some standard r05 maps (e.g. map_r05_cmip6_180x360_aave) to our standard maps here: https://web.lcrc.anl.gov/public/e3sm/diagnostics/maps/ to better support upcomming tri-grids output processing. I'm not sure if it is worthwhile to add |
|
Looked up the cf-convention. In general, it is recommended to supply cell bounds whenever possible. But it should be fairly straightforward to add bounds during post-processing. |
|
There are a few considerations regarding bounds variables. As you note above, the regridder adds the bounds variables to the regridded output, so long as the destination grid is FV (finite volume), including dual grid representations of SE (spectral element), so that's one option. It is also possible to place bounds variables in the native ELM output, i.e., to add that functionality to EAM itself. However, it's ill defined what the bounds would be in SE mode (not really an issue since v2, v3 are FV). Also, bi-grid runs use the pg2 grid for ELM, and pg2 grids are not rectangular, so the bounds variables would need to be four (lat,lon) vertices per gridcell. That means a timeseries file with a single field, temperature say, would use 10x the storage for the grid information as it does for the field information. Tri-grid runs usually use r05 grids, which are rectangular, so the bounds variables can be simple 1D arrays for lat and lon, and the grid overhead is minimal. Let me know if you'd like me to add any r05-to-cmip map-files to https://web.lcrc.anl.gov/public/e3sm/diagnostics/maps. |
|
@czender thank you for weighing in! I still think it is worthwhile to add In the mean time I added a new issue in |
|
Sounds good to me. Note that your new maps are missing the word "to" and thus do not comply with the old naming convention. (The v3 naming convention is different still, see here). There are multiple types of monotonic maps and the one you want can be obtained by invoking ncremap with - |
|
Thank you for reviewing and pointing out the new naming convention for map files..Though I'm still confused about the algorithm types/tools, and not sure if I provide the right command. Does below look right?: |
|
Almost, but that will get you (due to back-compatibility) ESMF weights with an NCO name. Use these: |
|
Thank you @czender! Now I get the recommended changes for v3 mapping file naming convention, it aligns with ncremap's -a argument values!. I will update the maps once /lcrc/group/e3sm back on line. Update: I have generated 3 files and saved at /lcrc/group/e3sm/public_html/diagnostics/maps/, following #523 (comment). |
|
I agree to not bother with cell bounds if the grids are not rectangular (and only the river grids in E3SM are rectangular). |
|
@chengzhuzhang @czender Thanks for looking into this further.
That problem sounds familiar, but it still doesn't answer why @xuezhengllnl's run got further than mine, when we had the same configuration.
Should I try to rerun now and see if the ILAMB figures generate? |
Yes! it looks like we have two sets of river/land grids, r0125 and r05, which will be relevant to v3 simulation campaigns. |
|
@xuezhengllnl and @forsyth2 By extending the variable list, and specifying one of the r05_to_cmip mapping files. The ilamb figures generated. |
To be clear, these are all new algorithm names that were recently agreed to in the new convention. Once we agreed to the new naming convention, I added these algorithm names as synonyms to the existing (and new, for TR bilinear) algorithms that ncremap supports. The first version of NCO that supports invocation by these algorithm names is 5.1.9, which is in the E3SMU-1.9.1 "emergency" release. Once I have time I will verify and support the new convention for MBTR as well, and expect that to be in NCO 5.2.0 in ~December. |
|
@chengzhuzhang @czender Thank you! I was able to generate the ILAMB plots using this cfg: |
|
I'm providing an updated .cfg for I think one of the issue is that if a mapping file |
|
@chengzhuzhang @forsyth2, Thank you so much for your help on this! I generated the ILAMB plots with your cfg settings. Here is the link: [https://web.lcrc.anl.gov/public/e3sm/diagnostic_output/ac.xzheng/E3SMv3_dev/20231117.v3b02.piControl.chrysalis/ilamb/_0351-0400/] I'm closing this issue now. |
github-project-automation
bot
moved this from Need to reply
to Done
in forsyth2 current tasks
|
@xuezhengllnl thank you for reporting back, in the ilamb results you recently generated, it still seems like only atmosphere variables are created. Maybe try the new config here would give the land diagnostics plots: #523 (comment) |
github-project-automation
bot
moved this from Done
to In Progress
in forsyth2 current tasks
|
@xuezhengllnl To follow up on supporting more bgc variables. Below is my updated variable list: |
@chengzhuzhang Thank you so much for your suggestions. I generated more ilamb diagnostics with your updated setting. Here is the ILAMB link |
|
Note for myself: Test |
|
Update: with updated v2.7 of ILAMB, more results are generated: https://web.lcrc.anl.gov/public/e3sm/diagnostic_output/ac.zhang40/tests/E3SMv3_dev/ilamb_bgc_eu1.9.2rc2/20231012.v3alpha04_trigrid_bgc.piControl.chrysalis/ilamb/land_monthly_0051-0100/, i.e. Leaf aera index related plots generated. There are some remaining issues, (e.g. only a subset of biomass diagnostics shown up). Those should be discussed with land bgc developers. |
github-project-automation
bot
moved this from In Progress
to Done
in forsyth2 current tasks
I added an ilamb section in my zppy cfg for v3alpha04.trigrid.bgc run
https://github.com/E3SM-Project/SimulationScripts/blob/master/archive/CoupledGroup/v3_dev/post.20231012.v3alpha04_trigrid_bgc.piControl.chrysalis.cfg
It ran and created a web link https://web.lcrc.anl.gov/public/e3sm/diagnostic_output/ac.xzheng/E3SMv3_dev/20231012.v3alpha04_trigrid_bgc.piControl.chrysalis/ilamb/180x360_aave_0091-0100/
But there is no figure or result in the link to me.
The zppy ilamb log file is /lcrc/group/e3sm2/ac.xzheng/E3SMv3_dev/20231012.v3alpha04_trigrid_bgc.piControl.chrysalis/post/scripts/ilamb_0091-0100.o410331
Thanks,
The text was updated successfully, but these errors were encountered: