Correct the issue that injection height has been artifically decreased by HEMCO

[jinlx]

Name and Institution (Required)

Name: Lixu Jin
Institution: University of Montana

Describe the update

This update reads injection heights to account for grid-dependent issues in GFAS, as detailed in Jin et al. (2023).

This update also requires that users create the corrected injection height data first, using the code in #2330. A better approach would be to have the data created and placed in the GEOS-Chem data portals for community use.

Expected changes

Correct the issue that injection height has been artifically decreased by HEMCO.

The independent injection height could be further used for other fire inventories.

Reference(s)

Jin, Lixu, et al. "Constraining emissions of volatile organic compounds from western US wildfires with WE-CAN and FIREX-AQ airborne observations." Atmospheric Chemistry and Physics 23.10 (2023): 5969-5991.

Related Github Issue

• Closes Correcting Injection height in biomass burning inventories #2330

[yantosca]

@jinlx: Do you have the injection height files at 0.1 x 0.1 already made?

[jinlx]

@yantosca Apologies for my delayed response—I was traveling last month and wasn’t able to monitor my personal emails effectively. For the inj files, I did not use 0.1x0.1 because HEMCO regrids these as emissions, which could potentially reduce the values artificially. Instead, I generated emission-corrected inj at a nested resolution (25km) for my personal use. I can extend this adjustment to other GEOS-Chem resolutions, such as 0.5x0.65, 2x2.5, and 4x5. This modification would only require a simple edit in the HEMCO_Config.rc.

Change

300 GFAS_EMITL        $ROOT/GFAS/v2018-09/$YYYY/GFAS_$YYYY$MM.nc mami            2003-2023/1-12/1-31/0 C xy m 1

into

300 GFAS_EMITL_weight $ROOT/GFAS/v2024-04/$YYYY/GFAS_$YYYY$MM_hgts_$RES.nc mami_avg_weight 2003-2023/1-12/1-31/0 C xy m 1

[yantosca]

Thanks @jinlx. I was wondering which file to use with this PR. It would be good if we could use the highest resolution file for all cases, if that is possible, as that is the HEMCO philosophy.

[yantosca]

@jinlx: Just checking in on this PR again. We were unsure which resolution file(s) to use with this PR. Can you provide us with the proper file?

[yantosca]

NOTE: This PR affects the GFAS biomass burning which is not enabled in fullchem benchmarks. So we can label this as "no-diff-to-benchmark".

[lizziel]

@jinlx, will be you be able to work on this? We would like to bring this into the model but are waiting for a response (see earlier post from @yantosca).

[jinlx]

I was bad at monitoring github in the past months. Back to questions Bob asked:

I was wondering which file to use with this PR. It would be good if we could use the highest resolution file for all cases, if that is possible, as that is the HEMCO philosophy.

The approach I’ve been using involves having injection files of varying resolutions preloaded in the project storage with resolution-specific filenames. In HEMCO config, I reference a universal variable $RES, which allows flexibility without the need to manually edit configurations for each resolution.

I agree that it will be better to use the highest resolution files. But I am not sure if HEMCO will artificially decrease the inj value again when regriding from high to resolutions. May need a little efforts from people understand both injection and hemco well (?).

Can you provide us with the proper file?

I will provide corrected inj files in standard resolutions, i.g., 2x2.5, 4x5, 0.25x0.3125, and 0.5x0.625 here after AGU, so that people can use it whether if it is implemetned in the future versions.

Again, sorry for not staying on notifications well. I will set up notifications for my school email moving forward so I can respond more promptly.

[yantosca]

@jinlx do you have the files for this PR?

[jinlx]

Hi @lizziel and @yantosca,

Thanks for the nudge and for your patience! I extended the GFAS files to support multiple horizontal resolutions for 2018 and 2019, which should be easy to use for benchmarking. You can find them under:

.../ExtData/HEMCO/GFAS/v2025-07

I also modularized my old python code so users can easily customize the year and resolution (provided they have /HEMCO/GFAS downloaded). Please see GFAS_injs.py and README.md included in the archive for more details.

To enable GEOS-Chem to dynamically select resolution-specific files, please update the HEMCO_Config.rc as follows:

To make the GEOS-Chem dynamically select the right resolution. We need to make corresponding changes in the HEMCO_Config.rc.

Old line:

300 GFAS_EMITL        $ROOT/GFAS/v2018-09/$YYYY/GFAS_$YYYY$MM.nc mami            2003-2023/1-12/1-31/0 C xy m 1

New line:

300 GFAS_EMITL_weight $ROOT/GFAS/v2025-07/$YYYY/GFAS_$YYYY$MM_hgts_$RES.nc mami_avg_weight 2003-2023/1-12/1-31/0 C xy m 1

The archived data and code: Inj4GCST.zip

Let me know if you have any questions.

Best,
Lixu

[yantosca]

Thanks @jinlx! I'll take a look at this shortly.

[yantosca]

Thanks @jinlx, this looks good. I just have a couple of updates:

1. Can you add the corresponding change for scale factor 300 to the run/GCHP/HEMCO_Config.rc.templates/HEMCO_Config.rc.fullchem file?
2. We would also need a quick sentence describing the update in the CHANGELOG.md file. This could go under the ### Added section

[jinlx]

Hi @yantosca,

The updates in HEMCO_Config.rc.templates apply equally to both GCHP and Classic. So, we’ll need to make the corresponding changes in:

• run/GCHP/HEMCO_Config.rc.templates/HEMCO_Config.rc.fullchem
• run/GCClassic/HEMCO_Config.rc.templates/HEMCO_Config.rc.fullchem

Old line:

300 GFAS_EMITL        $ROOT/GFAS/v2018-09/$YYYY/GFAS_$YYYY$MM.nc mami            2003-2023/1-12/1-31/0 C xy m 1

New line:

300 GFAS_EMITL_weight $ROOT/GFAS/v2025-07/$YYYY/GFAS_$YYYY$MM_hgts_$RES.nc mami_avg_weight 2003-2023/1-12/1-31/0 C xy m 1

For CHANGELOG.md, how about:

- Added support for resolution-specific GFAS injection height files in HEMCO_Config.rc

Let me know if that works!

Best,
Lixu

[msulprizio]

@jinlx Do you know if/how this update differs from the implementation in Singh et al. 2025?

Also tagging @inder-math and @yuanjianz for comment.

[yuanjianz]

@msulprizio Yes, Jin et al. 2023 and Singh et al. 2025 are different implementations. In the current HEMCO folders, Singh et al. 2025 is v2025-05 and Jin et al. 2023 is v2025-07.