GEOS-Chem model simulations with iodine chemistry sensitivity studies
Data files
May 05, 2026 version files 8.54 GB
-
moon_GEOS-Chem_model_output.zip
8.54 GB
-
README.md
6.91 KB
Abstract
Background
Tropospheric reactive halogens influence the oxidizing capacity of the atmosphere. Although previous studies have shown that iodine exerts the strongest impact on tropospheric ozone concentrations compared to chlorine and bromine, the impact of the recycling of aerosol iodide back to the gas phase on oxidants has not been estimated prior to this work. Here, we explicitly represent aerosol iodine speciation in a chemical transport model, including soluble organic iodine (SOI), iodate, and iodide, and allow for interconversion among these species. We find that aerosol iodine speciation, interconversion, and recycling substantially affect modeled oxidant abundances. Across model sensitivity studies, tropospheric Ox (the odd-oxygen family for ozone) and Oz (the odd-oxygen family for the HOx family, including OH) burdens vary by up to 10% and 5%, respectively. These ranges arise from uncertainties in heterogeneous iodine, demonstrating a need for future laboratory experiments and field observations in this area.
Dryad dataset DOI: 10.5061/dryad.3ffbg7b0c
Description of the data and file structure
Model input is included here from nine 1-year GEOS-Chem simulations, with output files in a separate directory for each simulation.
See Moon et al. (2026), ACP for details of base model, new iodine chemistry, and no HIO3 NPF simulations https://doi.org/10.5194/acp-26-2353-2026. More details on sensitivity studies may be found in a follow-up paper to be submitted, which does not yet have a DOI.
GEOS-Chem Classic simulations (v14.4.0, 10.5281/zenodo.11403303) in this study were conducted at 4° × 5° resolution with 72 vertical levels driven by MERRA-2 meteorology. Model runs were conducted for the year 2022 with 1-year spin-up.
We test model sensitivity to aerosol iodine speciation by turning off SOI (soluble organic iodine) and iodate chemistry in the “No SOI” and “No HIO3/IO3” simulations. We test model sensitivity to aerosol iodine interconversion rates by altering the rates (e.g., SOI to iodide, iodide to SOI, and iodate to iodide) with the “fast convert” simulation on the order of minutes, the “new iodine chemistry” simulation on the order of hours, and the “slow convert” simulation on the order of days. We test model sensitivity to the abundance and distribution of HIO3 by turning HIO3 new particle formation off (“No HIO3 NPF”), keeping HIO3 NPF off but increasing the reactive uptake coefficient (𝛾) to unity (“HIO3 𝛾= 1 + No HIO3 NPF), and increasing the reactive uptake of HIO3 to unity with HIO3 NPF turned on (“HIO3 𝛾= 1).
Files and variables
moon_GEOS-Chem_model_output.zip
Base_model : StateMet, SpeciesConc, ProdLoss, DryDep, WetDepLS, and WetDepConv for 2022.
new_iodine_chemistry : StateMet, SpeciesConc, ProdLoss, DryDep, WetDepLS, and WetDepConv for 2022.
no_HIO3_NPF : StateMet, SpeciesConc, ProdLoss, DryDep, WetDepLS, and WetDepConv for 2022.
HIO3_gamma_1_no_HIO3_NPF : StateMet, SpeciesConc, ProdLoss, DryDep, WetDepLS, and WetDepConv for 2022.
HIO3_gamma_1 : StateMet, SpeciesConc, ProdLoss, DryDep, WetDepLS, and WetDepConv for 2022.
fast_convert_minutes : StateMet, SpeciesConc, ProdLoss, DryDep, WetDepLS, and WetDepConv for 2022.
slow_convert_days : StateMet, SpeciesConc, ProdLoss, DryDep, WetDepLS, and WetDepConv for 2022.
no_SOI : StateMet, SpeciesConc, ProdLoss, DryDep, WetDepLS, and WetDepConv for 2022.
no_IO3 : StateMet, SpeciesConc, ProdLoss, DryDep, WetDepLS, and WetDepConv for 2022.
File descriptions
The list below summarizes the different types of files. Variable names, descriptions, and units are embedded within the netCDF files.
StateMet : Meteorological and physical state fields used to drive transport, chemistry, clouds, and deposition in GEOS-Chem (from the MERRA-2 meteorological archive and/or derived diagnostics).
SpeciesConc : Modeled concentrations of chemical tracers (gas-phase and aerosol-phase species) on the GEOS-Chem grid.
ProdLoss : Diagnostics that quantify chemical production and loss rates for a variety of reactions. Tags for individual reactions start with RXN, have g, h, or p for gas-phase, heterogeneous, or photolysis reactions, and then the two reactants. If there are multiple reactions with the same reactants, the product is listed as well. Examples: the gas-phase reaction IO + HO2 --> HOI + O2 has the tag RXN_g_IO_HO2. The heterogeneous reaction HOI + SO2 --> SO4 + HI has the tag RXN_h_HOI_SO2. The photolysis reaction HOI + hv --> I + OH has the tag RXN_p_HOI.
DryDep: Diagnostics for dry deposition rates of chemical species via removal at the surface without precipitation due to graviational setting.
WetDepLS : Wet deposition rates from large-scale (stratiform) precipitation processes, representing scavenging in and below stratiform clouds as implemented in GEOS-Chem.
WetDepConv : Wet deposition rates from convective precipitation processes via scavenging associated with convective updrafts/downdrafts and convective rainout/washout parameterizations.
Code/software
New iodine chemistry patches for GEOS-Chem 14.6.3
We include model patches for 14.6.3 for the new iodine chemistry in the "Patches for new iodine chemistry in version 14.6.3" directory. See "new iodine chemistry" simulation in Moon et al. (2026) https://doi.org/10.5194/acp-26-2353-2026.
There are four patches:
-
HEMCO_new_halogen_chemistry.diff - Added coarse primary marine organic aerosol and fine/coarse soluble organic iodine, iodide, and iodate emissions from the sea surface.
-
KPP_new_halogen_chemistry.diff - New iodine chemistry described in Moon et al. (2026) in ACP https://doi.org/10.5194/acp-26-2353-2026
Changed AERI, ISALA, and ISALC to AERIA/AERIC(iodide), IO3A/IO3C(iodate) and SOIA/SOIC (soluble organic iodine). Added HIO3 formation, aerosol uptake, and new particle formation. Added more halide recycling via hypohalous acids and halogen nitrates see Moon et al. (2026) for details.
Note that these additions continue to evolve. Contact allicatmoon@gmail.com for the most recent updates to GEOS-Chem halogen chemistry as it continues to be improved.
-
KPP_halogens_reaction_rate_tags.diff - Tags for all halogen chemistry and oxidant budgets. Note that these substantially slow down the simulation run time.
-
rundir_tags_new_halogen_chemistry.diff - run directory set up.
Usage notes
All model output is provided as NetCDF-4 files (*.nc4). These files can be opened and analyzed with free/open tools, including python (xarray + netCDF4). The different sensitivity studies are separated into different folders within the .zip file (see files and variables section).
Variable definitions and units (embedded in files)
Variable names, descriptions, and units are embedded within each NetCDF-4 (.nc4) file as standard NetCDF metadata attributes (e.g., variable name, long_name, units). Because the dataset contains hundreds of variables, we do not reproduce a full variable-by-variable table in this README. Users can list all variables and their units directly from the files using common NetCDF tools. Within the command line: ncdump -h file.nc4 (prints header with variable definitions and attributes). Also, within python xarray, one can define a dataset and list its variables, descriptions, and units with the code below
ds = xr.open_dataset("PATH/TO/OUTPUT/file.nc4).
ds
See Moon et al. (2026), ACP for details of base model, new iodine chemistry, and no HIO3 NPF simulations https://doi.org/10.5194/acp-26-2353-2026.
More details on sensitivity studies may be found in a follow-up paper, which does not yet have a DOI.