Add CEDS 0.1 x 0.1 degree emissions

[yantosca]

Name and Institution (Required)

Name: Bob Yantosca
Institution: Harvard + GCST

Confirm you have reviewed the following documentation

• Contributing guidelines

Describe the update

This is the companion PR to #1745.

Expected changes

As described in #1745

Related Github Issue(s)

• Closes Adding CEDS at 0.1 degree as a global emission inventory #1745

Tagging @msulprizio

[yantosca]

@1Dandan @Ruijun-Dang: Just wanted to tag #1745 (comment) on this PR. Will await your feedback before proceeding.

[1Dandan]

Hi @yantosca @msulprizio, the seasonality problem as mentioned in #1745 should be fixed in the new version of v2024-06. File names remain the same as those in v2023-04, and thus all v2023-04 can be replaced by v2024-06 for HEMCO_Config.rc and ExtData.rc templates. Let me know if you want other inputs from me.

[1Dandan]

Hi @yantosca, I noticed that the CEDS 0.1 version is still v2023-04, which needs to be changed to v2024-06.
Also, I deleted all files at CEDS/v2023-04 at http://geoschemdata.wustl.edu/ExtData/HEMCO/CEDS/ to avoid any confusion as v2024-06 should be used instead.

[yantosca]

Thanks @1Dandan, yes I just rebased the existing code atop of dev/14.5.0. I'll push another commit to update the directory paths.

[yantosca]

@1Dandan: DIrectory paths updated in commit 91acc1a. Will start integration testing and fix any further issues that arise.

[yantosca]

All GEOS-Chem integration tests passed:

==============================================================================
GEOS-Chem Classic: Execution Test Results

GCClassic #0c22a3a GEOS-Chem submod update: PR #2352 post-merge fix (rebuild KPP files)
GEOS-Chem #769698837 Fix mismatching brackets in HEMCO_Config.rc template
HEMCO     #e68ebae HEMCO 3.9.2 release

Using 24 OpenMP threads
Number of execution tests: 28

Submitted as SLURM job: 43586737
==============================================================================

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%  All execution tests passed!  %%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

[viral211]

Hi @yantosca and @1Dandan, CEDS also identifies their versions with the release date (https://github.com/JGCRI/CEDS/releases) and users can confuse that with the folder names in our CEDS data directory. For example, CEDS had a new version called v_2024_04_01, but I think our folder v2024-06 has CEDS version v_2021_4_21 data. Could we display the CEDS version prominently to avoid any confusion? Thanks.

[1Dandan]

Thanks @viral211 for the suggestion. I agree that it is confusing as our processed version may conflict with the releasing date of the original version. I add a line in the README of This is a seasonality fixed version based on the original released version of v2021_4_21_Release (https://data.pnnl.gov/group/nodes/dataset/13488). Would you @viral211 @yantosca have further suggestions to make it clear?

[yantosca]

Also see issue #2581. There seems to be a newer version of CEDS available. Should we skip this CEDS version and proceed to the newest version?

Tagging @1Dandan @Mashiat25 @eamarais @jaegle

[eamarais]

@yantosca, I'd suggest skipping this CEDS version. @1Dandan is probably more familiar with the different versions, so could comment based on their experience.

[yantosca]

Thanks @eamarais. We can wait for @1Dandan's assessment.

[1Dandan]

Sure I think it is great to include the new version of CEDS (v2024_07_08) as long as they release the gridded dataset.
I had some light email conversations with Steven Smith at PNNL about the CEDS emission development (CEDS-SatEM), which includes some explicit point sources for SO2 (https://zenodo.org/records/7958231). Possibly the new version of CEDS (v2024_07_08) is a step beyond the CEDS-SatEM with point sources of SO2 (?), as said in the releasing note. The time-series comparisons they released against the prior CEDS version (v_2021_04_20 - I guess it is typo and should be the same as v_2021_04_21; Also the one currently implemented in GEOS-Chem) show quite similar trend. I think we need to wait for the gridded dataset to compare the differences for the spatial distribution of emissions especially for SO2 point sources. Or maybe @Mashiat25 you already get the gridded dataset? Hopefully the effects on other aerosol precursors across CEDS versions of v_2021_04_21 and v2024_07_08 are very small.

[randallvmartin]

Hi - my 2cents are that if a 0.1 degree version of v2024_07_08 is not yet available and readily implemented in GC, then we should continue with PR #2171 to enable the 0.1 degree capability needed for fine resolution simulations. Randall

[msulprizio]

Based on discussion with the GCST and @randallvmartin and @djxjacob we think the best way forward is to:

1. Proceed with merging this PR for version 14.6.0 (now in development). This will replace the default CEDSv2 option with CEDS_01x01.
2. When CEDS v2024_07_08 is ready for implementation into GEOS-Chem/HEMCO we can proceed with either:
   a. Replacing CEDS_01x01 with the updated version CEDS_v2024_07_08. For this scenario, we would ideally we would the updated version at 0.1x0.1 resolution but this is not currently availably yet. As mentioned above, @1Dandan is in communication with the CEDS team.
   b. Including CEDS_v2024_07_08 as an option in addition to CEDS_01x01. In that case, users interested in using emissions for more recent years could choose to use CEDS_v2024_07_08, while users interested in high-resolution simulations still have the option to utilize the higher resolution CEDS_01x01 option.

With that said, @yantosca @1Dandan are we ready to change this pull request from a draft PR and mark it as ready for review?

[eamarais]

@msulprizio sounds like a reasonable strategy. I wasn't aware when I suggested skipping to the most recent version that the dataset isn't actually ready to implement.

[yantosca]

@msulprizio: We would have to merge the current 14.5.0 into this version and retest. There have been several emissions updates from the @kelvinhb that went into 14.5.0 that are not reflected yet in this PR.