Hg simulation restart file

[angothelene]

Name and Institution (Required)

Name: Helene Angot
Institution: CNRS/IGE

Confirm you have reviewed the following documentation

• [ y] Support guidelines
• [y ] User manuals
• [ y] Debugging GEOS-Chem and HEMCO errors
• [y ] Current and past Github issues

Description of your issue or question

@jennyfisher and I have run Hg simulations using v14.2.2 (out of the box). We obtained similar results independently, suggesting that we followed standard procedures. It seems that the provided restart file provided (from http://geoschemdata.wustl.edu/ExtData/GEOSCHEM_RESTARTS/v2021-12/) is problematic. While we could continue running for a bit to allow for a proper spin-up, it might be beneficial to correct this on the server. Are there any other Hg folks, such as @arifein or @viral211, who might have a properly spun-up restart file that we could use instead?

[yantosca]

Thanks @angothelene. Happy to add a new restart file to the data directory if you can provide us one.

Also note: We just released 14.2..3, which now makes sure that the same restart file that is used in the run directory creation is also the one that is used in the dry-run data download. This fixed an issue where the restart files paths could have been inconsistently defined in the two operations.

[viral211]

Thanks for pointing this out, @angothelene. Here is a link to a spun-up (15 yrs or so) Hg restart file. I would still recommend at least a one-year spin-up for your runs.

[yantosca]

Thanks @Viral, @angothelene. I'll submit a PR to add the new Hg restart file into the rundir creation.

[yantosca]

Also note issue #2057 that will affect the Hg simulations in 14.3.0 and later (once Cloud-J photolysis replaces FAST-JX). You can simply keep using FAST-JX until #2057 is resolved.

[angothelene]

@yantosca @viral211 thank you both!!

[jennyfisher]

Hi @viral211, @yantosca

Thanks for the new restart file. I tried to run with it, but the run dies because the species names are not quite the same. In the restart all the species names are fully capitalised (e.g. SpeciesRst_HGBR) whereas it seems this version of the code now expects them without unnecessary capitalisation (e.g. SpeciesRst_HgBr).

What is the easiest way to fix this? Should I rename every species in the restart file using ncrename?

FYI @yantosca whatever fix recommended here will need to go into the public code, too.

Cheers,
Jenny

[jennyfisher]

@yantosca I just read your pull request and saw that you have already further spun up to 2019 from Viral's restart file, so you must have resolved the issue and presumably have an updated file, but it doesn't appear to be on WUSTL yet (http://geoschemdata.wustl.edu/ExtData/GEOSCHEM_RESTARTS/ - no v2023-12/). Could you put the udpated restart on the WUSTL server or share here please? Thanks!

[arifein]

@jennyfisher @angothelene I'm sending here my restart file for v14.1. Let me know if this could work for you? If not, @emroy00 might have a more recent restart. Restart_file

[yantosca]

Hi @jennyfisher. Our Cannon cluster has not been playing nice so it took me a couple of days to get the spinup done (my runs kept timing out due to high latency on the scratch disk). I should have the restart files for GCClassic and GCHP ready shortly.

[jennyfisher]

Thanks @yantosca ! Can you post an update here when you've uploaded the new restart files to WUSTL? I think it would be good for me to use the updated restart for my runs to make sure there are no further issues with the newer file (but if it will take a long time, I will use @arifein 's restart in the meantime - thanks Ari!).

[yantosca]

Hi @jennyfisher. thanks for your patience. We have uploaded the new restart file onto the GEOS-Chem data (WashU) Globus endpoint, in folder ExtData/GEOSCHEM_RESTARTS/v2023-12.

There is an issue syncing that to the FTP site that we hope to resolve soon. For now the best thing is to grab the file with Globus.

[jennyfisher]

Thanks @yantosca - unfortunately I don't have access to Globus on my system here, so I'll have to wait until it is on the FTP site. Can you let me know when it is there? Thanks! :)

[msulprizio]

Hi @jennyfisher. The restart files should now be accessible at http://geoschemdata.wustl.edu/ExtData/GEOSCHEM_RESTARTS/v2023-12/.

[yantosca]

I will close out this issue now. Thanks.

[jennyfisher]

Just a note to say thanks, and having tested this new restart it all looks good! :)