Grantee Research Project Results
Final Report: Improving chemical mechanisms for regional/global models in support of US air quality management: application to the GEOS-Chem model
EPA Grant Number: R840014Title: Improving chemical mechanisms for regional/global models in support of US air quality management: application to the GEOS-Chem model
Investigators: Jacob, Daniel J.
Institution: Harvard University
EPA Project Officer: Chung, Serena
Project Period: August 1, 2020 through July 31, 2023 (Extended to July 31, 2024)
Project Amount: $785,010
RFA: Chemical Mechanisms to Address New Challenges in Air Quality Modeling (2019) RFA Text | Recipients Lists
Research Category: Air Quality and Air Toxics , Air
Objective:
The goal of this project was to unify, improve, and condense the chemical mechanisms used in the GEOS-Chem global/regional model for air quality applications. Specific objectives included: (a) implementation of a common model framework for the different GEOS-Chem chemical mechanisms to facilitate their merging, subsetting, and updating; (b) improvement of current mechanisms for halogen, mercury, and volatile organic compounds (VOCs) chemistry; (c) speed-up of the chemical calculation through adaptive mechanism reduction and machine learning methods; (d) updated estimates of background influences on U.S. air quality.
Summary/Accomplishments (Outputs/Outcomes):
We delivered on all four of these objectives and produced 15 peer-reviewed publications acknowledging funding from this grant. Specifically:
- We developed a new adaptive implementation of the Kinetic Pre-Processor (KPP) numerical solver for chemical mechanisms that allows chemical transport models such as GEOS-Chem to select an appropriately compact submechanism locally and on the fly [Shen et al., 2022; Lin et al., 2023]. This feature along with unification of gas- and aerosol chemistry as well as a number of performance improvements have been implemented into a new standard version KPP 3.0 hosted by Harvard as an independent GitHub repository and available for distribution to any model [Lin et al., 2023]
- We developed and implemented into GEOS-Chem new model mechanisms for halogen chemistry [Wang et al., 2021], mercury chemistry [Shah et al., 2021], aromatic chemistry [Bates et al., 2021a], methanol chemistry [Bates et al., 2021b], cloud acidity [Shah et al., 2020], NOx chemistry [Shah et al., 2023], and isoprene SOA [Bates et al., 2023]. We also updated GEOS-Chem mechanism kinetics with the latest recommendations from JPL and IUPAC [Bates et al., 2024] All of these mechanisms and updates have by now been brought into the standard version of the GEOS-Chem model.
- We developed a prototype neural-network machine learning algorithm for numerical solution of chemical mechanisms [Kelp et al., 2020] and implemented it successfully in GEOS-Chem using the SuperFast chemical mechanism [Kelp et al., 2022]. Results highlighted some important limitations of machine-learning algorithms applied to atmospheric chemistry mechanisms.
- We evaluated the impact of these updates on background NOx concentrations over the US and the implications for the interpretation of satellite NO2 data [Dang et al., 2023], and on background ozone affected by the combination of halogen chemistry and nitrate photolysis [Shah et al., 2024].
Conclusions:
- We significantly sped up the KPP 3.0 solver widely used in atmospheric chemistry models, and our adaptive auto-reduction capability further increases its performance by 30%. We benchmarked KPP 3.0 and the auto-reduction capability following the standard GEOS-Chem benchmarking procedures (https://geos-chem.org). We also added a number of diagnostics to KPP 3.0 (such as total production and loss of species and families) and enabled it to compute aqueous-phase chemistry coupled to the gas phase.
- Our new tropospheric halogen chemistry mechanism in GEOS-Chem is considered state-of-the-art. It has led to a model underestimate of background ozone observations which we then corrected by nitrate photolysis. That chemistry, together with the VOC chemistry mechanism updates developed through this project, is now carried in the standard GEOS-Chem mechanism and will be subject to further scrutiny by the GEOS-Chem community. Our new redox mercury mechanism is transformative in its treatment of elementary reactions and speciated forms of Hg(0)/Hg(II). It is now carried in the standard GEOS-Chem. We also implemented a new computation of cloudwater acidity that corrected a previous error revealed by an EPA-led model intercomparison. All these chemical updates were subject to the standard GEOS-Chem benchmarking procedure.
- We explored the possibility of using machine learning (ML) to speed up the chemical solver by an order of magnitude in year-long GEOS-Chem simulations. We advanced significantly on previous attempts to apply ML to atmospheric chemistry mechanisms but in the end we were not successful in avoiding drift of the solution over year-long simulations. This was due to the large dimensionality of the problem as well as the accumulation of errors over time (unlike in a regular ODE solver, where errors tend to correct themselves). ML is still of value for short-term chemical forecasts.
- We examined the implications of our work for background concentrations of NOx, ozone, and mercury over the US. We confirmed the presence of a large free tropospheric pool of NO2 that would affect the interpretation of satellite NO2 measurements.
Journal Articles on this Report : 15 Displayed | Download in RIS Format
| Other project views: | All 16 publications | 16 publications in selected types | All 16 journal articles |
|---|
| Type | Citation | ||
|---|---|---|---|
|
|
Bates K, Jacob D, Li K, Ivatt P, Evans M, Yan Y, Lin J. Development and evaluation of a new compact mechanism for aromatic oxidation in atmospheric models. Atmospheric Chemistry and Physics 2021;21(24):18351-18374. |
R840014 (2021) R840014 (2023) R840014 (Final) |
Exit Exit |
|
|
Kelp M, Jacob D, Lin H, Sulprizio M. An online-learned neural network chemical solver for stable long-term gobal simulations of atmospheric chemistry. Journal of Advances in Modeling Earth Systems 2022;14(6). |
R840014 (2021) R840014 (2023) R840014 (Final) |
Exit Exit |
|
|
Shah V, Jacob DJ, Moch JM, Wang X, Zhai S. Global modeling of cloud water acidity, precipitation acidity, and acid inputs to ecosystems. Atmospheric Chemistry and Physics 2020;20:12223–45. |
R840014 (2023) R840014 (Final) |
Exit Exit |
|
|
Shah V, Keller CA, Knowland KE, Christiansen A, Hu L, Wang H, et al. Particulate nitrate photolysis as a possible driver of rising tropospheric ozone. Geophysical Research Letters. 2024;51(5):e2023GL107980. |
R840014 (Final) |
Exit |
|
|
Kelp MM, Jacob DJ, Kutz JN, Marshall JD, Tessum CW. Toward stable, general machine-learned models of the atmospheric chemical system. Journal of Geophysical Research-Atmospheres 2020;125:e2020JD032759. |
R840014 (2023) R840014 (Final) R835873 (2020) R835873 (Final) R840012 (2021) R840012 (2022) R840012 (2023) R840012 (Final) |
Exit Exit |
|
|
Shen L, Jacob DJ, Santillana M, Wang X, Chen W. An adaptive method for speeding up the numerical integration of chemical mechanisms in atmospheric chemistry models:Application to GEOS-Chem version 12.0.0. Geoscientific Model Development 2020;13:2475–86. doi:10.5194/gmd-13-2475-2020. |
R840014 (Final) |
Exit Exit |
|
|
Shah V, Jacob DJ, Thackray CP, Wang X, Sunderland EM, Dibble TS, Saiz-Lopez A, Černušák I, Kellö V, Castro PJ, Wu R, Wang C. Improved mechanistic model of the atmospheric redox chemistry of mercury. Environmental Science & Technology 2021; doi:10.1021/acs.est.1c03160. |
R840014 (2023) R840014 (Final) |
Exit Exit |
|
|
Bates KH, Jacob DJ, Wang S, Hornbrook RS, Apel EC, Kim MJ, Millet DB, Wells KC, Chen X, Brewer JF, Ray EA, Commane R, Diskin GS, Wofsy SC. The global budget of atmospheric methanol:New constraints on secondary, oceanic, and terrestrial Sources. Journal of Geophysical Research-Atmosphere 2021;126:e2020JD033439. |
R840014 (2023) R840014 (Final) |
Exit Exit |
|
|
Wang X, Jacob DJ, Downs W, Zhai S, Zhu L, Shah V, Holmes CD, Sherwen T, Alexander B, Evans MJ, Eastham SD, Neuman JA, Veres P, Koenig TK, Volkamer R, Huey LG, et al. Global tropospheric halogen (Cl, Br, I) chemistry and its impact on oxidants. Atmospheric Chemistry and Physics 2021:1–34. |
R840014 (2023) R840014 (Final) |
Exit Exit |
|
|
Dang R, Jacob DJ, Shah V, Eastham SD, Fritz TM, Mickley LJ, Liu T, Wang Y, Wang J. Background nitrogen dioxide (NO2) over the United States and its implications for satellite observations and trends:effects of nitrate photolysis, aircraft, and open fires. Atmospheric Chemistry and Physics 2023;23:6271–84. doi:10.5194/acp-23-6271-2023. |
R840014 (2023) R840014 (Final) |
Exit |
|
|
Lin H, Long MS, Sander R, Sandu A, Yantosca RM, Estrada LA, Shen L, Jacob DJ. An adaptive auto-reduction solver for speeding up integration of chemical kinetics in atmospheric chemistry models:Implementation and evaluation in the Kinetic Pre-Processor (KPP) Version 3.0.0. Journal of Advances in Modeling Earth Systems 2023;15:e2022MS003293. doi:10.1029/2022MS003293. |
R840014 (2023) R840014 (Final) |
Exit |
|
|
Shah V, Jacob DJ, Dang R, Lamsal LN, Strode SA, Steenrod SD, Boersma KF, Eastham SD, Fritz TM, Thompson C, Peischl J, Bourgeois I, Pollack IB, Nault BA, Cohen RC, Campuzano-Jost P, et al. Nitrogen oxides in the free troposphere:Implications for tropospheric oxidants and the interpretation of satellite NO2 measurements. Atmospheric Chemistry and Physics 2023;23:1227–57. doi:10.5194/acp-23-1227-2023. |
R840014 (2023) R840014 (Final) |
Exit |
|
|
Shen L, Jacob DJ, Santillana M, Bates K, Zhuang J, Chen W. A machine-learning-guided adaptive algorithm to reduce the computational cost of integrating kinetics in global atmospheric chemistry models: Application to GEOS-Chem versions 12.0.0 and 12.9.1. Geoscientific Model Development 2022;15:1677–87. doi:10.5194/gmd-15-1677-2022. |
R840014 (2023) R840014 (Final) |
Exit |
|
|
Bates KH, Jacob DJ, Cope JD, Chen X, Millet DB, Nguyen TB. Emerging investigator series:aqueous oxidation of isoprene-derived organic aerosol species as a source of atmospheric formic and acetic acids. Environmental Science:Atmospheres. 2023;3(11):1651-64. |
R840014 (Final) |
Exit |
|
|
Bates KH, Evans MJ, Henderson BH, Jacob DJ. Impacts of updated reaction kinetics on the global GEOS-Chem simulation of atmospheric chemistry. Geoscientific Model Development. 2024;17(4):1511-24. |
R840014 (2023) R840014 (Final) |
Exit |
Supplemental Keywords:
Air quality modeling, GEOS-Chem, halogen chemistry, mercury chemistry, isoprene chemistry, numerical methods, chemical solvers, machine learning methods, backgroundRelevant Websites:
Atmospheric Chemistry Modeling Group - Harvard University Exit
Progress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.