Zhang et al. (2025) Development of UI-WRF-Chem (v1.0) for the MAIA satellite mission: case demonstration

Identification

• Journal: Geoscientific model development
• Year: 2025
• Date: 2025-11-26
• Authors: Huanxin Zhang, Jun Wang, Nathan Janechek, Cui Ge, Meng Zhou, Lorena Castro García, Tong Sha, Yanyu Wang, Weizhi Deng, Zhixin Xue, Chengzhe Li, Lakhima Chutia, Yi Wang, Sebastian Val, James McDuffie, Sina Hasheminassab, Scott Gluck, D. J. Diner, Peter R. Colarco, Arlindo da Silva, Jhoon Kim
• DOI: 10.5194/gmd-18-9061-2025

Research Groups

• Department of Chemical and Biochemical Engineering, The University of Iowa, Iowa City, IA, United States
• Center for Global and Regional Environmental Research, The University of Iowa, Iowa City, IA, United States
• South Coast Air Quality Management District (AQMD), Diamond Bar, CA, United States
• Goddard Earth Sciences Technology and Research (GESTAR) II, University of Maryland, Baltimore County, Baltimore, MD, United States
• Global Modeling and Assimilation Office, NASA Goddard Space Flight Center, Greenbelt, MD, United States
• NASA Jet Propulsion Laboratory, California Institute of Technology, Pasadena, CA, United States
• Atmospheric Chemistry and Dynamics Lab, NASA Goddard Space Flight Center, Greenbelt, MD, United States
• Department of Atmospheric Sciences, Yonsei University, Seoul, South Korea

Short Summary

This paper presents the development of the Unified Inputs (of initial and boundary conditions) for WRF-Chem (UI-WRF-Chem) framework to support the Multi-Angle Imager for Aerosols (MAIA) satellite mission. Major updates include improving dust size distribution in chemical boundary conditions, updating land surface properties using recent satellite data, and enhancing the representation of soil NOₓ emissions, with subsequent model improvements demonstrated over several MAIA target areas.

Objective

• To develop and demonstrate an enhanced WRF-Chem modeling framework (UI-WRF-Chem) that accurately simulates meteorological variables and speciated particulate matter (PM) concentrations, thereby supporting the MAIA satellite mission's objective of studying the impacts of different types of PM pollution on human health.

Study Configuration

• Spatial Scale: Two nested domains: an outer domain (D1) with 12 km x 12 km horizontal resolution (nominal dimension ~1080 km x 1000 km) and an inner domain (D2) with 4 km x 4 km horizontal resolution (nominal dimension ~360 km x 480 km). Both domains have 48 vertical levels extending from the surface to 50 hPa.
• Temporal Scale: Case studies for specific months: March 2018 (CHN-Beijing dust), July 2018 (CHN-Beijing land/NOₓ), June 2023 (ITA-Rome dust), July 2018 (USA-LosAngeles dust tuning), and June 2022 (USA-Atlanta cumulus/microphysics). AERONET climatology data from 2000-2020 (CHN-Beijing) and 2000-2023 (ITA-Rome) were used for dust particle size distribution constraints. MODIS land data from 2018-2020 (or 2018 for CHN-Beijing) were used as static inputs.

Methodology and Data

• Models used:
  • UI-WRF-Chem (v1.0) based on WRF-Chem (v3.8.1)
  • Gas-phase chemistry: Regional Acid Deposition Model, Version 2 (RADM2)
  • Aerosol module: Modal Aerosol Dynamics model for Europe (MADE) and Secondary ORGanic Aerosol Model (SORGAM) (modified to MADE/SORGAM-DustSS, including 5 dust bins and 4 sea salt bins)
  • Dust emission scheme: GOCART with Air Force Weather Agency (AFWA) modifications
  • Soil NOₓ emission scheme: Berkeley Dalhousie Iowa Soil NO Parameterization (BDISNP, newly developed)
  • Biogenic emissions: Model of Emissions of Gases and Aerosols from Nature (MEGAN)
  • Fire emissions: Fire Locating and Modeling of Burning Emissions inventory (FLAMBE)
  • Photolysis scheme: Madronich Fast Tropospheric UV and Visible Radiation Model (F-TUV)
  • Land Surface Model: Noah land model (with MODIS land data updates)
  • Cumulus schemes tested: Grell 3D ensemble (G3D, turned off for D2), Grell-Freitas (GF)
  • Microphysics schemes tested: Lin, WRF Single-Moment 6-Class Microphysics Scheme (WSM6), Morrison
  • Radiation schemes tested: Goddard (shortwave), Rapid Radiative Transfer Model for GCMs (RRTMG, shortwave/longwave), Rapid Radiative Transfer Model (RRTM, longwave)
  • Planetary Boundary Layer (PBL) schemes tested: Yonsei University (YSU), Mellor-Yamada-Janjic (MYJ), Mellor-Yamada-Nakanishi-Niino level 2.5 (MYNN2.5)
  • Emission Preprocessor: WRF-Chem Emission Preprocessing System (WEPS, newly developed)
  • Input/Boundary Condition Processors: GEOS2WRF, LDAS2WRF, GEOSBC (modified mozbc)
• Data sources:
  • Satellite: MODIS (AOD, land cover type MCD12Q1, NDVI MOD13A3, FPAR/LAI MCD15A2H, surface albedo MCD43A3/MCD43C3, Land Surface Temperature LST), VIIRS (AOD), TROPOMI (NO₂ Vertical Column Density VCD), CALIOP (aerosol extinction coefficient, aerosol type), Global Precipitation Measurement (GPM) mission (precipitation).
  • Observation (Ground-based): Aerosol Robotic Network (AERONET) (AOD, aerosol volume size distribution PSD), surface PM₂.₅ and PM₁₀ mass concentrations (China Ministry of Environmental Protection, U.S. EPA, ARPA Lazio, ARPAE Emilia Romagna), surface meteorological variables (MICAPS China, U.S. EPA, NCEI Integrated Surface Database ISD), Interagency Monitoring of Protected Visual Environments (IMPROVE) (speciated PM₂.₅), Chemical Speciation Network (CSN) (speciated PM₂.₅).
  • Reanalysis/Assimilation: NASA Goddard Earth Observing System Forward Processing (GEOS FP) (meteorological and chemical initial and boundary conditions), Modern-Era Retrospective analysis for Research and Application, version 2 (MERRA-2) (meteorological and chemical initial and boundary conditions, AOD, dust AOD), Global Land Data Assimilation System (GLDAS) (soil properties), North American Land Data Assimilation System (NLDAS) (soil properties).
  • Emission Inventories: HTAPv2.2, HTAPv3 (global anthropogenic), EDGARv5.0, EDGARv6.1, U.S. EPA National Emissions Inventory (NEI) 2017 (regional anthropogenic), Multi-resolution Emission Inventory model for Climate and air pollution research (MEIC) 2016 (China anthropogenic), FLAMBE, FINN v1.01, GFED v3.1, FEER-SEVIRI v1.0, GFAS v1.0, NESDIS Global Biomass Burning Emissions Product (GBBEP-Geo), Quick Fire Emissions Dataset version 2.4 (QFED v2.4), FIre Light Detection Algorithm (FILDA-2) (fire emissions).

Main Results

• Improved Dust Transport Simulation: Incorporating MERRA-2 chemical boundary conditions, especially with AERONET-constrained dust particle size distribution, significantly improved UI-WRF-Chem's ability to capture long-range dust transport events. For CHN-Beijing (March 2018 dust event), this reduced PM₂.₅ mean bias from -66.4 µg/m³ to -24 µg/m³ and increased correlation from 0.19 to 0.54. Similar improvements were observed for ITA-Rome (June 2023 Saharan dust), with significant reductions in mean absolute error for PM₂.₅, PM₁₀, and AOD.
• Enhanced Land Surface Representation: Updating land surface properties with recent MODIS data improved the simulation of surface skin temperature (TSK) over CHN-Beijing (July 2018), better capturing the urban heat island (UHI) phenomenon and reducing relative bias compared to MODIS Land Surface Temperature (LST).
• Refined Soil NOₓ Emissions: The newly developed BDISNP soil NOₓ emission scheme increased simulated NO₂ Vertical Column Density (VCD) over croplands in CHN-Beijing (July 2018), leading to better agreement with TROPOMI NO₂ VCD and a statistically significant reduction in mean absolute error. This also resulted in an increase of surface nitrate concentrations by up to 30% in rural areas.
• Optimized Dust Emissions: Tuning dust emission parameters (gamma=1, alpha=0.3) in the GOCART-AFWA scheme significantly improved UI-WRF-Chem's simulation of daily surface PM₁₀ and hourly AOD over the dust-prone region of USA-LosAngeles (July 2018), reducing PM₁₀ mean absolute error from 46.7 µg/m³ to 17.9 µg/m³.
• Cumulus Scheme Impact on PM₂.₅: For the USA-Atlanta target area (June 2022), turning off the traditional Grell 3D ensemble (G3D) cumulus scheme in the 4 km inner domain (D2) resulted in better performance for precipitation and surface total PM₂.₅ concentrations compared to keeping it on, with a lower mean absolute error for PM₂.₅ (3.7 µg/m³ vs. 5.7 µg/m³).
• Speciated PM₂.₅ Evaluation: Over USA-Atlanta (June 2022), UI-WRF-Chem simulated daily Organic Carbon (OC), Elemental Carbon (EC), and dust showed higher correlations (0.45-0.72) with ground observations, while combined sulfate+nitrate showed lower correlation (-0.03 to 0.23). The model consistently overestimated dust and underestimated other speciated PM₂.₅ components.

Contributions

• Development of UI-WRF-Chem (v1.0), a comprehensive and enhanced WRF-Chem framework specifically tailored for the MAIA satellite mission, integrating multiple new modules and significant modifications.
• Implementation of a novel approach for providing unified, self-consistent meteorological and chemical initial and boundary conditions using NASA GEOS FP and MERRA-2 data, which effectively captures long-range transport events and reduces computational costs.
• Development of a method to constrain MERRA-2 dust particle size distribution using AERONET climatology data, leading to improved accuracy in dust transport simulations.
• Integration of up-to-date land surface properties from MODIS satellite data, enhancing the model's ability to represent land cover changes and their impacts on local meteorology, such as the urban heat island effect.
• Creation and integration of the Berkeley Dalhousie Iowa Soil NO Parameterization (BDISNP) scheme, improving the accuracy of soil NOₓ emissions and subsequent NO₂ and nitrate aerosol simulations.
• Development of the WRF-Chem Emission Preprocessing System (WEPS) for flexible ingestion and processing of diverse global and regional anthropogenic and fire emission inventories.
• Comprehensive demonstration and evaluation of these UI-WRF-Chem enhancements across four diverse MAIA target areas (Beijing, Rome, Los Angeles, Atlanta), providing a robust foundation for MAIA's PM health studies and future air quality research.

Funding

• NASA Jet Propulsion Laboratory (JPL) (subcontract no. 1583456)
• NASA (contract no. 80NM0018D0004)

Citation

@article{Zhang2025Development,
  author = {Zhang, Huanxin and Wang, Jun and Janechek, Nathan and Ge, Cui and Zhou, Meng and García, Lorena Castro and Sha, Tong and Wang, Yanyu and Deng, Weizhi and Xue, Zhixin and Li, Chengzhe and Chutia, Lakhima and Wang, Yi and Val, Sebastian and McDuffie, James and Hasheminassab, Sina and Gluck, Scott and Diner, D. J. and Colarco, Peter R. and Silva, Arlindo da and Kim, Jhoon},
  title = {Development of UI-WRF-Chem (v1.0) for the MAIA satellite mission: case demonstration},
  journal = {Geoscientific model development},
  year = {2025},
  doi = {10.5194/gmd-18-9061-2025},
  url = {https://doi.org/10.5194/gmd-18-9061-2025}
}