Skip to main content
NCBI home page
EPA Author Manuscripts logoLink to EPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Aug 28.
Published in final edited form as: J Geophys Res Atmos. 2025 Aug 28;130(16):e2025JD044021. doi: 10.1029/2025JD044021

Improved Simulation of Particle Number Concentrations Over the US: Integrating a Size-Resolved Advanced Particle Microphysics Model Into CMAQ

Jingbo Mao 1, Fangqun Yu 1, Benjamin N Murphy 2, Jingyu An 3, Yan Zhang 3, Gan Luo 1, Shao Lin 4, A Gannet Hallar 5
PMCID: PMC12466536  NIHMSID: NIHMS2107041  PMID: 41018119

Abstract

Accurate representation of aerosol size distributions and total number concentrations is critical for evaluating particle impacts on climate, clouds, and public health. An Advanced Particle Microphysics model (APM) has been integrated into the USEPA’s Community Multiscale Air Quality Modeling System (CMAQ) to improve its representation of aerosol size distribution. CMAQ is a state-of-the-science air quality model simulating the emission, transport, formation, evolution, and deposition of air pollutants. Unlike the original CMAQ’s simplified modal approach, the sectional APM model distinguishes primary and secondary particles, keeps track of secondary species coated on each type of primary particle (black carbon, primary organic carbon, dust, and sea salt), and uses flexible binning for different particle types. This introduces 116 new APM-related tracers increasing computational cost by 83% but enabling high-size resolution aerosol simulations. The updated CMAQ-APM incorporates a ternary H2SO4-H2O-NH3 ion-mediated nucleation scheme to investigate particle formation and growth over the United States. Predicted monthly mean nucleation rates, particles larger than 10 nm (CN10), and cloud condensation nuclei (CCN) at 0.4% supersaturation (CCN0.4) in the surface in summer and winter of 2013 range from ~0.001 to ~2 cm−3 s−1, ~2,000–15,000 cm−3, and ~300–5,000 cm−3, respectively. Modeled CN10 and CCN0.4 generally align well with observations with normalized mean bias in the range of −6.42%–192.32% and 11.44%–36.06%, respectively. The incorporation of APM into CMAQ substantially improves the capability of CMAQ in representing size-resolved particles, especially ultrafine particles and CCN activation thresholds, hence offering a robust tool for quantifying aerosol-climate interactions and exposure risks.

Plain Language Summary

Airborne particles can harm human health and influence cloud formation, but accurately modeling their complex behavior remains challenging. We enhanced the capability of Community Multiscale Air Quality (CMAQ) model to address size-dependent aerosol properties by integrating an advanced particle microphysics (APM) module tracking different types of particles, their coatings, and their size evolution with higher size resolution. We employed the CMAQ-APM model utilizing a ternary H2SO4-H2O-NH3 ion-mediated nucleation (TIMN) scheme to investigate particle formation and growth over the United States. By comparing results from the improved model to real-world measurements at seven U.S. locations, we found it reliably predicts both the number of pollution particles and their ability to form cloud condensation nuclei (CCN). Although the upgrade increases computing time by 83%, CMAQ-APM helps scientists better understand air pollution, climate change, and the health risks associated with tiny particles in the air.

1. Introduction

The concentration of airborne particles significantly impacts both climate and human health with elevated levels of ultrafine particles posing serious health risks (Han et al., 2016; Knibbs et al., 2011; Kwon et al., 2020; Lin et al., 2022; Qi et al., 2024). The troposphere is populated with aerosol particles that have either been directly emitted or have formed through in situ nucleation. Although new particle formation (NPF) contributes minimally to aerosol mass near nucleation sites, it often dominates ultrafine particle production and strongly modulates cloud condensation nuclei (CCN) abundance (Pierce & Adams, 2009; Spracklen et al., 2008; F. Yu & Luo, 2009). These secondary particles exhibit disproportionately large health and climate impacts owing to their smaller size, deeper respiratory tract penetration, and greater surface-area-to-volume ratios compared to PM2.5 (HEI, 2013; Moreno-Ríos et al., 2022; Nair et al., 2023; Oliveira et al., 2019). Traditional chemistry-transport models (CTMs) are frequently used to estimate pollutant concentrations, and they have long focused on representing the mass of PM2.5 and PM10 with diameters smaller than 2.5 and 10 μm, respectively.

Particle mass concentrations are predominantly affected by larger-sized particles (diameter Dp > ~100 nm) due to the cubic relationship between particle diameter and mass. Conversely, number concentrations are driven by smaller particles (Dp < 100 nm). Moreover, the particle size distributions (PSDs) have been shown to explain over 80% of the variability in CCN formation with chemical composition playing a lesser role (Antony & Vladimir, 2010; Dusek et al., 2006). There are two typical approaches to modeling PSDs in CTMs, that is, the sectional approach (e.g., Debry & Sportisse, 2007; Gelbard et al., 1980; Jacobson, 1997) and the modal approach (e.g., Binkowski & Roselle, 2003; Binkowski & Shankar, 1995; Sartelet et al., 2006; Vignati et al., 2004; K. T. Whitby, 1978; E. R. Whitby & McMurry, 1997). A sectional aerosol microphysical model (Community Aerosol and Radiation Model for Atmospheres, CARMA) and a modal aerosol microphysical model (Modal Aerosol Model, MAM4) were integrated into the same framework of the Community Earth System Model Version 2 (CESM2). The findings indicated a superior depiction of the aerosol size distribution following the Mount Pinatubo volcanic eruption in CARMA relative to MAM4 (Tilmes et al., 2023). Blichner et al. (2021) utilized the OsloAeroSec model and showed that the sectional technique mitigated biases in modeling particle numbers ranging from 50 to 100 nm. The Community Multiscale Air Quality (CMAQ) model (www.epa.gov/cmaq, last access: 12 February 2025) developed and distributed by the United States Environmental Protection Agency (U.S. EPA) is a state-of-the-science numerical air quality model with comprehensive representation of the emission, transport, formation, destruction, and deposition of many air pollutants, including PM2.5 (Appel et al., 2021). A modal representation, which includes Aitken, accumulation, and coarse modes with the mean diameter (standard deviation) of 15 nm (1.7), 80 nm (2.0), and 600 nm (2.2), respectively, is used for the aerosol in CMAQ. In addition to a coarse mode for large particles, fine particles are represented by two modes (Binkowski & Roselle, 2003; Binkowski & Shankar, 1995; K. T. Whitby, 1978) called the Aitken and accumulation modes. Numerous studies have demonstrated that CMAQ simulations generally meet rigorous evaluation criteria for modeling aerosol mass concentrations (e.g., Appel et al., 2021; L. Chen et al., 2023; Lee et al., 2024) and the size distribution of fine and coarse particles (Nolte et al., 2015). Additionally, prior studies have integrated CMAQ with detailed aerosol sectional models for specific research applications including simulations of regional aerosols at Big Bend National Park and urban aerosols in Tampa, Florida (Nolte et al., 2008; Pun et al., 2006). However, the centrally maintained CMAQ model still has limited capacity in modeling aerosol number concentrations.

The publicly released CMAQ model considers one nucleation pathway, H2SO4-H2O binary homogeneous nucleation (BHN) (Vehkamäki et al., 2002), which deviates significantly from well-constrained laboratory measurements (F. Yu, Nadykto, et al., 2020; F. Yu, 2007; F. Yu, Luo, et al., 2020) and inadequately explains nucleation events in the real atmosphere (Semeniuk & Dastoor, 2018; Zhao et al., 2024). A further limitation of CMAQ is its size distribution using three lognormal modes, which is inadequate for robust simulation of the number concentration of ultrafine particles and CCN that are strongly influenced by size distributions (Jacquot & Sartelet, 2024). Furthermore, aerosols in CMAQ are assumed to be internally mixed, which is not sufficient for capturing variations in hygroscopic properties of multiple particles (Zheng et al., 2021).

The integration of an advanced particle microphysics (APM) model has enabled detailed investigations of particle formation and growth mechanisms in GEOS-Chem (F. Yu, 2011; F. Yu & Luo, 2009) and WRF-Chem (Luo & Yu, 2011) models. The APM model differentiates between secondary and primary particles, follows secondary species adhered to each category of primary particles including black carbon, primary organic carbon, dust, and sea salt, and uses variable numbers of size bins for particles of different types. APM, designed to better predict particle number concentrations and other key properties important for their public health and climatic effects, represents secondary particles (which generally dominate ultrafine particles and CCN) with 30 size bins for those between 1.2 and 120 nm and 10 bins for larger particles from 120 nm to 12 μm (Luo & Yu, 2011; F. Yu & Luo, 2009). The model utilizes an up-to-date version of the H2SO4–H2O–NH3 ternary ion-mediated nucleation (TIMN) theory (F. Yu et al., 2018, 2020). The removal of nascent particles via coagulation with primary particles is specifically simulated. The APM model in GEOS-Chem and WRF-Chem has undergone a comprehensive evaluation of new particle formation and particle number concentrations using a diverse set of measurements from land, ship, and aircraft observations (e.g., An et al., 2024; Luo & Yu, 2011; Mao et al., 2025; F. Yu, Nadykto, et al., 2020; F. Yu & Luo, 2009; F. Yu, Luo, et al., 2020; F. Yu et al., 2015). The APM model has also been implemented in the Nested Air Quality Prediction Modeling System (NAQPMS) (X. Chen et al., 2014), the Atmospheric Aerosol and Chemistry Model developed by the Institute of Atmospheric Physics, Chinese Academy of Sciences (IAP-AACM) (X. Chen et al., 2021), and the Whole Atmosphere Community Climate Model version 6 (WACCM6) in the Community Earth System Model version 2 (CESM2) (Williamson et al., 2019).

The present work aims to enhance the capability of CMAQ in simulating aerosol nucleation and growth processes by integrating the APM model, previously implemented in GEOS-Chem and WRF-Chem (Luo & Yu, 2011; F. Yu & Luo, 2009), into the CMAQ framework. The coupled model is then used to examine the evolution of sectional particle characteristics across the United States. The coupled model’s architecture and salient characteristics are thoroughly outlined in Section 2. Section 3 details the model setup, validation against empirical measurements, and derived outcomes. Section 4 summarizes and discusses the findings.

2. Methodology

2.1. The CMAQ Model

The Community Multiscale Air Quality (CMAQ) model (www.epa.gov/cmaq, last access: 12 February 2025) developed and distributed by U.S. EPA is an offline, multi-scale, state-of-the-science numerical air quality model. The CMAQ model encompasses a detailed simulation of O3 and PM2.5 together with their precursors. A thorough introduction of the model, encompassing emissions, transport, chemistry, deposition, cloud, and aerosol processes, is available in the User’s Guide (https://github.com/USEPA/CMAQ/tree/main/DOCS/Users_Guide, last access: 29 November 2024).

Meteorological data used in CMAQ is generated by the Weather Research and Forecasting (WRF) model (Skamarock & Klemp, 2008). The CMAQ model version 5.4, as implemented in the study, features four chemical mechanisms Carbon-Bond version 6 (CB6) (Luecken et al., 2019), Statewide Air Pollution Research Center (SAPRC) (Xie et al., 2013), the Regional Atmospheric Chemical Mechanism (RACM) (Sarwar et al., 2013), and the Community Regional Atmospheric Chemistry Multiphase Mechanism (CRACMM) (Place et al., 2023; Pye et al., 2023). Among these mechanisms, CB6 is historically the most commonly used (Appel et al., 2021). The AERO7 aerosol microphysics module includes both direct emissions of primary organic aerosol (POA) and secondary organic aerosol (SOA) formed from oxidation of several gas-phase precursor systems including isoprene, terpenes, aromatics and alkanes and uptake of water onto hydrophilic organics (Pye et al., 2015, 2017; Qin et al., 2021). Inorganic aerosol species include sulfate, nitrate, ammonium, BC, and metals. For emissions, the Detailed Emission Scaling, Isolation, and Diagnostic (DESID; Murphy et al., 2021) module maps species from emission inputs generated by the Sparse Matrix Operator Kernel Emissions (SMOKE) to CMAQ model species. For aerosol calculations, the CMAQ AERO7 module models a range of essential microphysical processes: emission, dry and wet deposition, radiative scattering, nucleation, condensation, evaporation, coagulation, and chemistry. The consideration of aerosols in the original CMAQ is shown in Figure 1a. The module represents particle size distribution as a combination of three log-normal modes: Aitken, accumulation, and coarse. The magnitude and shape of these modes are predicted in time by tracking the magnitude of three moments: 0th (equivalent to number), 2nd (proportional to surface area), and 3rd (proportional to volume and mass). All particle components, such as sulfate, nitrate, ammonium, BC, and organic compounds, are assumed to be internally mixed (Appel et al., 2021; Binkowski & Roselle, 2003; Elleman & Covert, 2009; USEPA, 2017). Several prior studies integrated external aerosol modules with CMAQ to enhance aerosol representation. Karamchandani et al. (2010) integrated an advanced Plume-in-Grid module, the Advanced Modeling System for Transport, Emissions, Reactions, and Deposition of Atmospheric Matter (AMSTERDAM) to CMAQ to resolve sub-grid scale processes from elevated point sources. The University of California at Davis (UCD) aerosol module, a sectional, internally mixed model with dynamic gas-particle mass transfer, was integrated into the CMAQ model (Nolte et al., 2008; K. M. Zhang & Wexler, 2008). The coupled CMAQ-UCD model enhanced the performance of coarse particles with diameters above 2.5 nm; nonetheless, it continued to adhere to the BHN nucleation method and did not account for thermodynamic equilibrium.

Figure 1.

Figure 1.

Open in a new tab

The treatments of aerosols in (a) CMAQv5.4 and (b) APM; (c) the flowchart of CMAQ-APM. The inorganics in (a) mainly include sulfate, nitrate, ammonium, Cl−1, Na+. In (c), P_LVSOGs represents the production rate of low volatile SOGs; P_H2SO4 and APMSO4CLD are the production of H2SO4 by the reaction of SO2 with OH and via in-cloud chemistry, respectively. The coating of primary particles (BC, OC, sea salt, and dust) by secondary species is simulated using four distinct tracers, which are defined as BC_SP, OC_SP, SS_SP, and DST_SP. Total SOA on secondary particles of different sizes is lumped into one tracer (SULF_LV). BC_LV, POC_LV, SS_LV, and DST_LV represent coated on BC, POC, sea salt, and dust particles.

2.2. Advanced Particle Microphysics (APM) Model

The advanced particle microphysics (APM) model is an advanced multitype, multicomponent, size-resolved microphysics code developed for various applications. It was originally developed to study the formation and evolution of aerosols and contrails in aircraft plumes (F. Yu, 1998) by resolving complex size-dependent interactions in a system consisting of ions, charged and uncharged aerosols, volatile and nonvolatile particles, and liquid and solid phases (F. Yu, 1998; F. Yu & Turco, 1997). The APM model adapted some structures and algorithms of a size-resolved aerosol microphysics model that Turco and colleagues originally developed for studying stratospheric aerosol (Hamill et al., 1982; Toon et al., 1988; Turco et al., 1979) and later expanded for air quality studies (Jacobson et al., 1994; Jacobson & Turco, 1995). The APM model contains many new features and algorithms as detailed by F. Yu (1998), including variable bin resolution and bin number, hybrid bin structure, kinetic nucleation modeling capabilities, explicit consideration of charge effects, modularization of various aerosol microphysics subroutines, and extensive usage of lookup tables for computing efficiency. In the past several decades, the APM model has been continuously improved and evolved into two separate but related versions: a comprehensive APM for detailed aerosol microphysics studies (including nucleation) (e.g., F. Yu et al., 2018, 2024) and a simplified APM for coupling with 3-D models for regional and global simulations of size-resolved particle microphysics (e.g., F. Yu & Luo, 2009). The simplified APM model used in this study incorporates key microphysical processes including nucleation, condensation and evaporation, coagulation, and thermodynamic equilibrium with local humidity (Figure 1b). The H2SO4–H2O–NH3 ternary ion-mediated nucleation (TIMN) scheme, developed by F. Yu et al. (2018, 2020), is employed in its latest iteration to study the nucleation process. The APM contains options to calculate nucleation rates based on other mechanisms such as organics-mediated nucleation (F. Yu et al., 2017) but these are not considered in the present study. The APM model is also flexible, facilitating straightforward specification of the number of bins, compositions, and aerosol types. Computational acceleration in APM is realized through techniques such as the extensive utilization of precomputed lookup tables and adaptive temporal discretization for disparate microphysical processes (Luo & Yu, 2011; F. Yu & Luo, 2009). More details of APM are given in the Section below.

2.3. APM Integration in CMAQ: Key Features

The APM model is integrated within the CMAQ CTM so that detailed aerosol microphysics are applied to update the aerosol state during each model time step. The algorithm flowchart of CMAQ-APM is shown in Figure 1c. The nucleation is determined by the TIMN mechanism (F. Yu et al., 2018, 2020). The model explicitly simulates the condensation of H2SO4 and low-volatility secondary organic gases onto various particle types (Figure 1b, see also F. Yu, 2011). It also incorporates equilibrium uptake or partitioning to determine the contributions of nitrate, ammonium, and semivolatile SOA to sulfate particle growth (Figure 1b). The model further considers the removal of secondary particles by other particle species, including sea salt, black carbon (BC), primary organic carbon (POC), and dust. The following section outlines the emission inputs and crucial aspects of the coupled model concerning size-resolved aerosol features.

2.3.1. Emissions

Emissions for a specific year are categorized into 14 source sectors, including airports, rail, commercial marine vessels, fugitive dust, agriculture, fires (including agricultural fires, wildfires, and prescribed burns), residential wood combustion, oil and gas, volatile chemical products, fuel combustion from industrial and other nonpoint sources, fuel combustion from industrial and other point sources, nonroad mobile sources, onroad mobile sources, and electric generating utilities (Foley et al., 2023). The input emissions include NOx, SO2, CO, PM2.5, primary organic carbon (POC), primary black carbon (BC), volatile organic compounds (VOCs), and NH3. Consideration is given to biogenic VOC emissions, calculated by the Model of Emissions of Gases and Aerosols from Nature (MEGAN) (Guenther et al., 2012), and sea spray aerosol emissions, but wind-blown dust emissions are disabled. In the present study, emitted primary fine sulfate is distributed into 40 bins within the APM module assuming two lognormal modes: a nucleation mode with a geometric mean diameter (dg) of 10 nm and a standard deviation (σ) of 1.6 and an accumulation mode with a dg of 70 nm and a σ of 2.0 (K. T. Whitby, 1978). The mass fractions allocated to these modes are 5% and 95%, respectively (K. T. Whitby, 1978). A coarse mode for primary coarse sulfate is included with dg and σ of 6 μm and 2.2 (Binkowski & Roselle, 2003).

Carbonaceous aerosol emissions are dominated by the combustion of fossil fuels, biofuels, and biomass burning. According to our prior research (Luo & Yu, 2011; F. Yu & Luo, 2009), BCOC resulting from fuel combustion and biomass burning is presumed to exhibit two lognormal distributions, characterized by a number median diameter of 60 and 150 nm, with a standard deviation of 1.8 for both distributions.

2.3.2. Representation of Major Particles and Precursors in the Coupled Model

The APM model allows for adjustable bin resolution to focus on specific size ranges. The bin structure is chosen to represent typical size ranges of various types of particles in the atmosphere and, for secondary particles, to have relatively high resolution for the size range important to their contribution to total number concentrations and CCN number abundance. In the coupled CMAQ-APM model, sulfate particles are represented using 40 sectional bins spanning dry diameters from 0.0012 to 12 μm. This includes 30 bins for the 0.0012–0.12 μm range, crucial for nucleated particle growth to CCN, and 10 bins for the 0.12–12 μm range. Sea salt is modeled with 20 bins, covering 0.012–12 μm, with 10 bins each for the 0.012–0.12 μm and 0.12–12 μm ranges. Dust is represented by 15 bins for 0.12–12 μm, whereas black carbon (BC) and primary organic carbon (POC) each use 15 bins for 0.0015–1 μm. In previous implementations of APM into GEOS-chem (F. Yu & Luo, 2009), WRF-Chem (Luo & Yu, 2011), and NAQPMS (X. Chen et al., 2014), BC and POC are treated with modal methods. In the later version of APM, including the one updated in WRF-Chem (F. Yu, Nadykto, et al., 2020; F. Yu, Luo, et al., 2020), the one implemented in IAP-AACM (X. Chen et al., 2021), and the one considered in this study for CMAQ, both BC and POC are represented by 15 sectional bins.

Ultrafine secondary particle scavenging by larger primary particles is important for assessing CCN contributions of secondary particles. Additionally, secondary species coating primary particles impact their optical and hygroscopic properties. The APM distinguishes secondary from primary particles and explicitly resolves the coating (a “semiexternally mixed” system, F. Yu & Luo, 2009). Four tracers—BC_SO4, OC_SO4, sea salt_SO4, and dust_SO4—are employed to track amount of sulfate coated on primary particles due to coagulation, condensation, and in-cloud aqueous chemistry. Each tracer represents the sulfate mass on a certain primary particle type. Nitrate and ammonium concentrations linked to sulfate are determined using the ISORROPIAII thermodynamic equilibrium model (Fountoukis & Nenes, 2007). Two pathways for the production of H2SO4, a crucial component for secondary particle formation and growth, are considered in this study. The first of these is the production (P_H2SO4) resulting from the reaction of SO2 with OH. The other pathway is attributed to aqueous-phase chemistry within clouds (APMSO4CLD) including reactions with hydrogen peroxide, ozone, O2 catalyzed by Mn2+ and Fe3+, methyl hydrogen peroxide, and peroxyacetic acid (Jacobson, 1997; Martin & Good, 1991).

Sulfate production via aqueous chemistry within clouds is distributed to sea salt, BC, POC, and dust particles or sectional sulfate particles based on the number concentrations of the activated particles. A separate tracer for low-volatility secondary organic gases (LVSOGs) with volatility less than 10−2 μg/m3 is added in CMAQ-APM. The production and sinks of LVSOGs (P_LVSOGs) are considered to explicitly predict the concentrations of LVSOGs in the APM module. As discussed in a previous work (F. Yu, 2011), total SOA on secondary particles of different sizes is aggregated into one tracer (SULF_LV) and subsequently performs a surface area-dependent redistribution across these particles. The quantification of low-volatility SOA coating on distinct primary aerosol types—dust, sea salt, BC, and POC—is achieved through tracers DST_LV, SS_LV, BC_LV, and POC_LV. The CCN concentrations are contingent upon the size distribution and composition of atmospheric particles at a specific supersaturation ratio. The APM model uses Köhler theory to determine CCN by assessing particle hygroscopicity and critical activation diameters (F. Yu & Luo, 2009). The APM module calculates an effective hygroscopicity parameter (κ) for each aerosol category, including sulfate, sea salt, dust, black carbon, and organic carbon. The critical activation diameters are established by lookup tables that integrate the hygroscopicity parameter and temperature. A particle is counted as a CCN when its dry diameter meets or exceeds the critical activation diameter for a specific supersaturation level.

2.3.3. Essential Microphysical Processes Governing Particle Size Distributions

2.3.3.1. Nucleation

Nucleation is essential for simulating particle number concentration and CCN abundance. This study employs the TIMN mechanism chosen for its strong physical basis and validation against laboratory data (F. Yu et al., 2018). To improve computational speed, pregenerated lookup tables from a detailed kinetic TIMN model (F. Yu, Nadykto, et al., 2020; F. Yu, Luo, et al., 2020) are used. Precise TIMN in each grid cell can be obtained from the lookup tables using a straightforward multivariable interpolation algorithm. The potential impact of amines and other compounds on both neutral and ion-mediated nucleation is recognized with the magnitude of this influence varying according to atmospheric concentration levels. Further research is necessary to develop amine emission estimates and incorporate amine interactions into the model, allowing for the evaluation of other significant nucleation pathways on atmospheric particle numbers.

2.3.3.2. Condensation, Equilibrium, and Coagulation

Particle growth is simulated through the condensation and evaporation of vapors. This study explicitly calculates the condensation of H2SO4 and low-volatility secondary organic gases (LVSOGs). Nitrate, ammonium, and semivolatile secondary organic aerosol (SOA) uptake are determined using equilibrium partitioning with inorganic equilibrium calculated by ISORROPIA II (Fountoukis & Nenes, 2007). The partitioning of nitrate and ammonium to each sectional sulfate particle as well as the amount of bulk sulfate coating the primary aerosol components are parameterized as proportional to the corresponding sulfate mass. In this regard, the model evaluates the impact of nitrate and ammonium uptake on the growth of secondary particles. In addition, we link the total mass of AMT2-AMT5, ALVOO2, ASVOO1-ASVOO3, AISO1-AISO3, AAVB2-AAVB4, AORGC, AOLGA, AOLGB, AMTHYD, and AMTNO3 from original CMAQ, with detailed description of these species available at https://github.com/USEPA/CMAQ/blob/main/CCTM/src/MECHS/mechanism_information/cb6r3_ae7_aq/cb6r3_ae7_aq_species_table.md, into APM to address the SOA partition to secondary particle growth in a way similar to that described by F. Yu (2011). The self-coagulation of sectional sulfate, BC, POC, and sea salt, along with the removal of sulfate particles by primary particles via coagulation scavenging, is simulated using the mass-conserving semi-implicit numerical scheme reported by Jacobson et al. (1994). Coagulation is computationally intensive, making its optimization crucial for 3D sectional aerosol microphysics models. To reduce computational cost, two schemes are implemented: pregenerated lookup tables for coagulation kernel calculations and a variable time step for coagulation calculations (F. Yu, 1998; F. Yu & Luo, 2009).

2.3.3.3. Dry and Wet Deposition

In the present study, dry deposition in CMAQv5.4 is calculated using the M3Dry model with the resistance-in-series approach of Pleim and Ran (2011). The CMAQ model incorporates cloud processes, featuring an aqueous chemistry model for sulfur and secondary organic aerosol (Carlton et al., 2008) processing and a basic mechanism for scavenging. The wet deposition algorithms were derived from the study conducted by Chang et al. (1987). The APM-related tracers are computed using Henry’s Law equilibrium equation to determine final concentration and deposition amounts.

2.4. Computational Overhead of the CMAQ-APM Model

All computational tasks in this study were performed on uniform Linux workstations utilizing the 2.6 GHz Intel (R) Xeon(R) Gold Opteron Processor 6348. The model system is compiled using IntelMPI (https://www.intel.com/content/www/us/en/developer/tools/oneapi/toolkits.html) for parallel processing. All simulations are executed on four-node hardware with 28 Intel Broadwell physical cores per node. In the present study, the mechanism of cb6r5_ae7_aq is used with 224 tracers (129 gases, 84 aerosols, and 11 nonreactive tracers). Standard CMAQ takes ~60 min for 1 day simulation over a domain of 299 × 459 grid cells (12 km horizontal spacing, 35 vertical layers) covering the contiguous United States, parts of Mexico and Canada, and adjacent ocean regions. The coupled CMAQ-APM model, which includes complete sectional particle microphysics (nucleation, condensation, coagulation, deposition, and scavenging) and CB6 chemistry, required approximately 110 min for the same simulation. This model utilizes 339 tracers, adding 116 to account for size-resolved aerosol processes. Despite the significant increase in model complexity and the inclusion of detailed microphysics, the computational cost increase is relatively small, highlighting the efficiency of the coupled model. It needs to be pointed out that the performance metrics may fluctuate due to variables including the particular CPU architecture, the aggregate number of processors employed, compiler selection, and I/O performance.

3. Simulations and Assessment

3.1. Model Configuration and Study Periods

In the study, the M3Dry dry-deposition option, the CB6r3_AE7_AQ chemical mechanism, and aero7 have been selected. CMAQ-APM was set up with a 12 km × 12 km horizontal grid covering the contiguous United States (CONUS) and 35 layers spanning from the surface up to 50 hPa, which is the same as the EPA’s (Environmental Protection Agency) Air QUAlity TimE Series (EQUATES) project (https://www.epa.gov/cmaq/equates, last access: 2 December 2024). We apply meteorology and emission inputs from EQUATES (https://drive.google.com/drive/folders/1G2_LBLy7_n91Ur0ulsLZ9zwGs3luTzn2, last access: 2 December 2024) to drive the model. As described in Section 2.3.1, since the redistributions of sulfate, BC, and POC emissions are internally coded within the APM, there are no extra emission inputs requirements compared to those in original CMAQ. A summary of configurations for meteorological fields and emissions is presented by Appel et al. (2021) and Foley et al. (2023). This study utilized lateral boundary and initial conditions for its 12 km grid spacing CONUS domain sourced from a 108 km grid spacing Northern Hemispheric simulation from EQUATES (https://drive.google.com/drive/folders/1G2_LBLy7_n91Ur0ulsLZ9zwGs3luTzn2, last access: 2 December 2024). For APM tracers considered in initial and boundary conditions, we redistributed the sulfate, BC, and POC mass into each bin based on the different lognormal distributions described in Section 2.3.1.

We conduct the simulations for two periods: two summer months (28 May–31 July 2013) and two winter months (28 October–31 December 2013) with a spin-up time of 4 days. These specific periods were chosen based on the availability of observational data on particle size distributions, total particle number concentration, and CCN abundance, allowing for a comprehensive evaluation of the model’s seasonal performance. Due to the limitation of the available meteorological data during the simulation periods, we validated the simulated temperature (T) and relative humidity (RH) with the measurements at the Pinnacle State Park (PSP) site (Figures S1 and S2 in Supporting Information S1). Overall, CMAQ simulated diurnal variations in T and RH are in good agreement with measurements. In present study, we focus on characterizing the model’s predictive capabilities for particle size distribution, number concentration, and CCN abundance building upon the established accuracy of meteorological and pollutant mass concentration simulations (Beck et al., 2020; Hogrefe et al., 2021; Sarwar et al., 2023). Pearson’s correlation coefficient (r), Normalized Mean Bias (NMB), and Index of Agreement (IoA) have been widely used in model evaluation (Willmott, 1981; S. Yu et al., 2006). Focusing on the APM performance regarding the simulated particle size distributions and number concentrations, we chose the hourly averaged outputs to evaluate the model’s general behavior and response under various typical conditions. Hourly averages are generally more useful for direct comparison with regulatory air quality standards or observational data. Monthly/bimonthly averages effectively smooth out short-term noise caused by hourly fluctuations, thus better revealing the model’s performance in simulating seasonal trends in pollutant concentrations and long-term systematic biases.

Aerosol number concentrations and PSDs have been measured at some sites in the United States. To enable a comparison between simulations and observational data, a list of available surface observations during the simulation periods was compiled as detailed in Table 1. The geographical positions of seven sites are visualized in Figure 2.

Table 1.

Details of Observation Sites Used in This Study, Including Names, Locations, Relevant Information, and References for Available Aerosol Number Concentration, Particle Size Distribution (PSD), and CCN Data

Sites Site location Site Charact. Measured variables Instruments Measured periods Source of data
Pinnacle State Park (PSP) (36.21°N, 81.69°W) Rural CN10
SO2 		TSI model 3783 CPC
Thermo model 43i		 5–30 NOVEMBER 2013 A
Temperature
Relative humidity		 meteorological sensors 1 NOVEMBER–31
DECEMBER 2013
1 JUNE–31 JULY 2013
Appalachian State University (APP) (42.10°N, 77.21°W) Rural CN10 TSI model 3010 CPC 1 NOVEMBER–31
DECEMBER 2013
1 JUNE–31 JULY 2013		 B
Southern Great Plains (SGP) (36.8°N, 97.5°W) Rural CN10 TSI model 3010 CPC 1 NOVEMBER–31
DECEMBER, 2013
1 JUNE–31 JULY 2013		 B
PSDs SMPS + APS 1–30 NOVEMBER 2013 C
CCN Cloud Condensation Nuclei (CCN) counter 1 NOVEMBER–19
DECEMBER 2013
1 JUNE–31 JULY 2013		 E
Centreville (CTR) (32.9°N, 87.25°W) Rural CN10 SMPS 20 JUNE–15 JULY 2013 D
SO2 MARGA 1 JUNE–15 JULY 2013
PSDs SMPS 1 JUNE–15 JULY 2013
CCN Cloud Condensation Nuclei (CCN) counter 20 JUNE–15 JULY 2013
Storm Peak Laboratory (SPL) (40.46°N, 106.74°W) Forest CN10 TSI model 3010 CPC 1 NOVEMBER–31
DECEMBER 2013
1 JUNE–31 JULY 2013		 B
PSDs SMPS 1 JUNE–31 JULY 2013 B
Bondville (40.05°N, 88.37°W) Rural CN10 TSI model 3760 CPC 1 NOVEMBER–31
DECEMBER 2013
1 JUNE–15 JULY 2013		 B
Trinidad Head (THD) (41.05°N, 124.15°W) Coast CN10 TSI model 3760 CPC 5–30 NOVEMBER 2013
1 JUNE–31 JULY 2013		 B
Open in a new tab

Figure 2.

Figure 2.

Open in a new tab

Horizontal spatial distributions of simulated surface layer precursors and meteorology (SO2, OH, P_H2SO4, TCS, H2SO4, NH3, T, RH, and LVSOG) important for new particle formation during the period of 1 June–31 July (summer) 2013.

3.2. Simulated Particle Formation, Number Concentration, and CCN Abundance in the US

Figures 2 and 4 show simulated ground horizontal spatial distributions of precursors (SO2, OH), the production rate of sulfuric acid (P_H2SO4), total condensation sink (TCS), H2SO4, NH3, LVSOG, temperature (T), and relative humidity (RH) controlling new particle formation during the period of 1 June–31 July (summer) 2013 and 1 November–31 December (winter) 2013, respectively. Nucleation rate (J), particle number with diameter above 3 nm (CN3), inorganic aerosols (SNA, which refers to the sum of sulfate, nitrate, and ammonium), SOA, CN10, and CCN concentration at a supersaturation of 0.4% (CCN0.4) are shown in Figures 3 and 5 for each simulation period. Typical summertime simulated concentrations of aerosol precursors in the surface across the CONUS are ~0.03–3 ppbv for SO2 and ~0.06–0.22 pptv for OH. The production rate of sulfuric acid via the reaction of SO2 and OH is 103–106 cm−3 s−1, whereas the condensation sink is ~0.001–0.011 # cm−3 s−1. Sulfuric acid, driven by the SO2 oxidation and condensation sink, ranges from 0.03 to 3 pptV. In addition to H2SO4, NH3 and temperature control the variations in nucleation rate. NH3 and temperature are ~0.1–70 ppbv and 8–36°C, respectively (Figure 2). Relative humidity ranges from 10% to exceed 90% increasing from west to east on CONUS shown in Figure 2. LVSOG is in the range of 0.3–7 pptV. In Figure 3, the calculated monthly mean nucleation rates generally range from 0.001 to 2 cm−3s−1 with their spatial distributions closely aligned with aerosol precursor patterns. Similarly, the horizontal distributions of CN3 are strongly related to the nucleation rate with number concentrations of CN3 ranging from 2,000 to 15,000 # cm−3. In the current study, we treat SNA and SOA via uptake and partitioning to contribute to particle growth. SNA and SOA are in the range of 0.3–7 and 0.3–4 μg m−3, respectively. Calculations based on the simulated particle number size distributions yielded CN10 values ranging from approximately 2,000 to 15,000 cm−3 and CCN0.4 values ranging from approximately 300 to 5,000 cm−3. Like those in the summertime, the wintertime simulated SO2, OH, P_H2SO4, TCS, H2SO4, and NH3 are ~0.03–5 ppbV, below 0.1 pptV, 103–8 × 104 # cm−3 s−1, ~0.001–0.01 # cm−3 s−1, ~0.03–0.8 ppbV, and ~0.1–30 ppbV, respectively; temperature and RH are in the range of −5–20°C and 30%–90%; LVSOG ranges from 0.1 to 5 pptV (Figure 4).

Figure 4.

Figure 4.

Open in a new tab

Same as Figure 2, but in the winter months (1 November–31 December 2013).

Figure 3.

Figure 3.

Open in a new tab

Same as Figure 2, but for nucleation rate (J), CN3, SNA, SOA, CN10, and CCN0.4 (1 June–31 July 2013).

Figure 5.

Figure 5.

Open in a new tab

Same as Figure 3, but in the winter months (1 November–31 December 2013).

Compared to nucleation rates in summer, the values in winter are higher even though H2SO4 and NH3 are lower. Laboratory and theoretical studies indicate that lower temperatures enhance nucleation (e.g., Tiszenkel et al., 2019; F. Yu et al., 2018). CN3, SNA, SOA, CN10, and CCN0.4 are in the range of 2,000–20,000 # cm−3 s−1, 0.3–8 μg m−3, less than 1.5 μg m−3, ~1,000–15,000, and ~300–5,000 cm−3 (Figure 5). Aerosol precursor spatial distributions, as modeled, are consistent with their respective source regions. Specifically, the SO2 distribution exhibits a spatial correlation with the National Emissions Inventory (https://drive.google.com/drive/folders/19Uz7dEvzDBLfZv5MKgQFCzwad_6i9Xgp, last access: 6 June 2025), delineating coal-fired power plant locations, particularly within the Ohio Valley. Similarly, NH3 hotspots coincide with regions with significant agricultural and animal feeding activities, notably in the Midwest US and the west coastline of California. Substantial seasonal and regional fluctuations exist for SO2, OH, H2SO4, NH3, nucleation rate, particle number concentrations (CN3 and CN10), and CCN0.4. The values of aerosol concentrations (CN3 and CN10) in the eastern US are higher in winter and strongly correlated with nucleation. CCN0.4 hotspots shift from northeast (summer) to southeast (winter) United States, which is highly relevant to the nucleation rate change.

3.3. Comparisons With Measurements

3.3.1. Particle Size Distributions

Particle number concentration is strongly influenced by size distributions (Jacquot & Sartelet, 2024) and nucleation. Based on available measurements, we present the comparisons of size distribution at SPL (summer) and SGP (winter) sites. Figures 6 and 7 present observed and simulated PSDs, CN10, and simulated SO2, NH3, H2SO4, and nucleation rates during the period of 1 June–31 July 2013 (summer) at the SPL site and 1–30 November 2013 (winter) at SGP site, respectively. In the summer case at the SPL site, the evolution of PSDs distinctly indicates the presence of intense nucleation and growth events on certain days, resulting in substantial elevations in CN10. The model can predict frequent new particle formation (NPF) events aligning with in situ observations. H2SO4 dominates the modeled nucleation process, hence influencing CN10 concentrations (Figure 6). In the winter case at the SGP site, less frequent and weaker NPF events are seen than those that occurred in summer at the SPL site partly due to relatively weak photochemistry and less biogenic emission (Figure 7). The model can identify the NPF events but it slightly overestimates the growth. Consistent with F. Yu, Luo, et al. (2020), our simulations show that particles, nucleated at a few nanometers, effectively grow to the 10–30 nm range via H2SO4 condensation and the equilibrium uptake of HNO3, NH3, and H2O on nucleation event day. Moreover, the temporal evolutions of PSDs and CN10 demonstrate a substantial increase in CN10 subsequent to nucleation events.

Figure 6.

Figure 6.

Open in a new tab

Observed and simulated PSDs, CN10, and simulated SO2, NH3, H2SO4, and nucleation rates during the period of 1 June–31 July 2013 (summer) at the SPL site.

Figure 7.

Figure 7.

Open in a new tab

Observed and simulated PSDs, CN10, and simulated SO2, NH3, H2SO4, and nucleation rates during the period of 1–30 November 2013 (winter) at the SGP site. Selected 3 days period from 19 to 21 November 2013 with nucleation events and nonevents is marked with a black dashed rectangle.

To analyze nucleation processes and their contribution to regional CN10, we selected a 3-day period (19–21 November 2013, Figure 7) encompassing both nucleation and non-nucleation events. The day 19 November has a strong nucleation event at the SGP site; 20 November has a weaker nucleation event at the SGP site, and 21 November shows nucleation nonevent at the SGP site. Figures 8 and 9 show, over United States, the simulated horizontal distribution of SO2, TCS, H2SO4, NH3, J, and CN10 in the surface. The considerable daily variations in SO2 and TCS across the US are readily apparent in Figure 8. The influence of SO2 oxidation and the condensation sink on H2SO4, which in turn affects the nucleation rate, is clearly observable. This finding aligns with nucleation occurrences documented at the SGP site during the 3-day period (Figure 6). Figure 9 demonstrates that nucleation occurs regionally with a spatial pattern similar to H2SO4 distribution. CN10 exhibits a distribution analogous to the nucleation rate across the CONUS from November 19 to 20. On November 21, CN10 continues to exhibit elevated values with no nucleation events occurring over the Midwestern CONUS.

Figure 8.

Figure 8.

Open in a new tab

For each day of the 3 days period from (left to right) 19 to 21 November 2013, (top to bottom) modeled horizontal spatial distribution of SO2, TCS (Total Condensation Sink), and H2SO4 over the US with the measurement sites THD, SPL, SGP, CTR, PSP, APP, and Bondville marked on the maps.

Figure 9.

Figure 9.

Open in a new tab

Same as Figure 8, but for NH3, J, and CN10.

3.3.2. Condensation Nuclei Larger Than 10 nm (CN10) and CCN Abundance

Figure 10 illustrates the comparison of measured and simulated CN10 using coupled CMAQ-APM and original CMAQ model during the simulation periods at other five sites (THD, SPL, SGP, APP, and Bondville). The original CMAQ exhibits substantial limits in estimating particle number concentrations with the normalized mean bias (NMBo) varying from −80.93% to 417.33% and the highest Pearson correlation coefficients (r) of 0.5 (Table 2). Structural biases include the lack of atmospherically relevant nucleation mechanisms and the simple (and computationally fast) three-mode aerosol scheme. Additionally, there is a challenge of the accurate representation of nucleated particle growth before their inclusion into the Aitken mode. In the original CMAQ, when the mass production rate derived from the particle number production rate surpasses the H2SO4 vapor production rate, it is limited to the H2SO4 vapor production rate. Consequently, the particle number production rate is adjusted, presuming that the particles have a diameter of 2 nm. The production rate of newly generated particles at 2 nm is utilized to determine alteration in the number concentrations of Aitken mode using the methodology of Binkowski and Roselle (2003). This approach yields an upper-limit estimate of nucleated particle number and mass concentrations in the Aitken mode, as CMAQ overlooks the loss of part of the particles with their growth from 2 nm to 10–100 nm (Y. Zhang et al., 2010).

Figure 10.

Figure 10.

Open in a new tab

The comparisons of measured and simulated CN10 during the simulation periods at other five sites (THD, SPL, SGP, APP, and Bondville). NMBa and NMBo represent the calculated normalized mean bias based on CMAQ-APM and original CMAQ model, respectively.

Table 2.

Pearson Correlation Coefficients (r) of CMAQ-APM Number Concentration Above 10 nm (CN10) at Five Sites (Bondville, APP, SGP, THD, and SPL) During Summer and Winter Simulations

Sites Summer case
 Winter case
CMAQ-APM CMAQ CMAQ-APM CMAQ
Bondville 0.30 0.22 0.39 −0.10
APP 0.57 0.50 0.66 0.09
SGP 0.02 −0.03 0.40 −0.08
THD 0.15 −0.02 0.12 0.18
SPL 0.53 0.15 0.59 0.13
Open in a new tab

The performance using CMAQ-APM exhibits spatiotemporal heterogeneity evidenced by the mean NMBa value of 41.24% (−6.42%–192.32%). The performance of CN10 at the APP and SGP locations, regardless of summer or winter, is much superior to that of other sites. The time variations in CN10 at APP and SPL sites are in good agreement with observations (Table 2) with a Pearson correlation coefficient (r) of 0.66 and 0.59 compared to 0.09 and 0.13 using original CMAQ. CMAQ-APM substantially overestimates CN10 during winter at the Bondville site. The primary origin of tiny particles is the nucleation process. The TIMN nucleation scheme posits that H2SO4 and NH3 are essential nucleation precursors, but factors such as temperature, relative humidity, ionization rate, and the surface area of pre-existing particles also affect nucleation rates (F. Yu et al., 2018, 2020). In the winter boundary layer, galactic cosmic rays and soil-derived radioactive substances are the primary sources of ions (F. Yu, Nadykto, et al., 2020; F. Yu, Luo, et al., 2020). Uncertainties in meteorology (e.g., wind, temperature, relative humidity, radiation), surface emissions, SO2 oxidation, NH3 concentrations, transport, and other processes may all lead to the under or overestimation of CN10. Overall, the model simulations match well with observed CN10 and can capture most peak values, indicating the performance of the coupled model is reasonable and acceptable. Table 1 indicates that only partial measurements are accessible at the two sites (CTR and PSP) during the simulated periods. The data at the CTR location are available from 1 June to 15 July 2013, whereas the data at the PSP site are available in November. Furthermore, since observed SO2 data are accessible at both sites during the simulation periods, we want to study the relationship between SO2 and CN10. Consequently, we have addressed the two sites individually below (Figure 11).

Figure 11.

Figure 11.

Open in a new tab

Comparison of observed and simulated SO2 and CN10 concentrations at PSP (5–30 November 2013) and CTR (1 June–15 July 2013) sites.

The comparison between observed and simulated SO2 and aerosol number concentrations (CN10) observations at PSP site from 5 to 20 November 2013 and CTR site from 1 June to 15 July 2013 is presented in Figure 11. It shows NMB and Index of Agreement (IoA) of SO2 are 79.39% and 0.78 at PSP site, 43% and 0.55 at CTR site, respectively, which show reasonable agreement with observations. As shown in Figure 10 (right panel), the NMB and IoA of CN10 are 82.65% and 0.55 at the PSP site and 24.93% and 0.46 at the CTR site. The conversion of gaseous SO2 to sulfuric acid, via oxidation, is a crucial factor in the nucleation process. Consequently, if SO2 levels are overestimated, it is probable that the formation of newly nucleated atmospheric particles will be overestimated, potentially influencing other aerosol microphysical processes, such as condensation and coagulation, which subsequently affect particle number.

Particle size distribution, chemical composition, and water vapor supersaturation dictate the number of aerosols that become cloud droplets. Using simulated particle size distributions, we calculated CCN concentration at a given supersaturation. The time series of observed and simulated CCN concentration at a supersaturation of 0.4% (CCN0.4) are presented in Figure 12. Overall, CMAQ-APM simulates the diurnal variations in CCN0.4 in good agreement with measurements with a NMB (IoA) of 36.06% (0.55) during summertime (1 June–31 July, Figures 12a) and 31.93% (0.78) during wintertime (10 November–10 December, Figure 12b) at the SGP site, respectively. Similarly, Figure 12c shows that simulated CCN0.4 concentrations and temporal variability (0 to ~3,500 #/cm3) at the CTR site are overall consistent with observations with NMB and IoA of −11.44% and 0.66 during the period of 20 June to 15 July, respectively. The results once more underscore the performance and utility of the CMAQ-APM model.

Figure 12.

Figure 12.

Open in a new tab

Time series comparison of observed and simulated CCN0.4 concentrations at the SGP and CTR sites.

4. Summary and Discussion

In this study, an advanced particle microphysics (APM) model, known for its computational efficiency and previously validated in GEOS-Chem and WRF-Chem, was integrated into CMAQ to model particle size distribution, total number, and CCN abundance across the United States. The implementation of CMAQ-APM results in a computational overhead of approximately 110 min, reflecting an 83% augmentation relative to the 60-min execution time of CMAQv5.4 utilizing the CB6r3_AE7_AQ chemical mechanism for equivalent daily simulations on the same computational platform. This modest rise in computational costs, incurred by incorporating full size-resolved microphysics, is considered acceptable given the benefits for detailed aerosol modeling. Nonetheless, it may become necessary to improve runtime efficiency for specific applications where long-term or high-resolution simulations are required. Future considerations for efficient solutions include enhanced parallelization and adjusting model configurations (such as minimizing the number of aerosol bins).

Model simulations of aerosol number concentrations, particle size distributions, and CCN abundance are validated by the measurements at seven sites over the United States during the summer (June 1–July 31) and winter (November 1–December 31) months of 2013. The model results are in overall agreement with observations. The values of nucleation rate, CN10 and CCN0.4, are typically from ~0.001 to ~2 cm−3 s−1, ~2,000–15,000, and ~300–5,000 cm−3. Substantial seasonal and regional fluctuations exist for SO2, H2SO4, NH3, nucleation rate, CN10, and CCN0.4. Aerosol number concentrations (CN3 and CN10) and CCN have considerable seasonal and regional variability strongly correlating with changes in the nucleation rates. CCN0.4 hotspots shift from the northeast in summer to the southeast in winter.

The results obtained for particle formation and number concentration over the US are reasonable but uncertainties persist, highlighting the need for additional research to improve model performance and reduce these uncertainties. Additional nucleation schemes should be considered that involve both biogenic and anthropogenic pollutants such as organics, amines, and HNO3, which may play considerable roles in NPF in some regions (Shrivastava et al., 2024; Wang et al., 2022; Yao et al., 2018; Zhao et al., 2024). In addition, there are still uncertainties in the emissions and chemistry leading to the formation of both POC and SOA (Murphy et al., 2017, 2023; Pye et al., 2023, 2024), and CMAQ-APM size distribution predictions will be sensitive to these species through both NPF and condensational growth processes. The incorporation of APM into CMAQ substantially improves the capability of CMAQ in representing size-resolved particle microphysics and simulating the number concentration of particles of various size regimes, including ultrafine particles and CCN, which are important for public health and climate/weather impact studies. The increasing attention on the impacts of aerosols on human health and climate requires that particle microphysical processes be understood, and we envision CMAQ-APM will become a powerful tool for further development.

Supplementary Material

Supplement1

Supporting Information:

Supporting Information may be found in the online version of this article.

Key Points:

  • Integrated sectional aerosol model enhances the ability to track size-dependent aerosol properties

  • Coupled model increases computational demand by 83% but enables high-fidelity aerosol number simulations

  • Model predictions align well with observations on particle number concentrations and size distributions

Acknowledgments

This work is supported by the New York State Energy Research and Development Authority (NYSERDA; contract no. 137487) and Department of Energy (DOE, Grant DE-SC0022886). Some of the data used to compare with model simulations were obtained from: (a) the Atmospheric Radiation Measurement (ARM) user facility, a U.S. Department of Energy (DOE) Office of Science user facility managed by the Biological and Environmental Research Program, (b) EBAS database operated by NILU—Norwegian Institute for Air Research (https://ebas-data.nilu.no/Default.aspx), and (c) the NOAA Earth System Research Laboratories (ESRL) (https://csl.noaa.gov/groups/csl7/measurements/2013senex/). We thank James Sherman for the APP data, and NOAA/Global Monitoring Laboratory for THD and BND data used in this study. We acknowledge Ian B. McCubbin, Ty Aikens, Dan Gilchrist, Chris Rapp, Megan Ostile, and Maria Garcia for technical support for the collection of data at SPL. The authors thank Drs. Havala Pye and Chris Nolte for comments/suggestions that improve the manuscript. The views expressed in this article are those of the authors and do not necessarily reflect the views or polices of the U.S. EPA.

Footnotes

Conflict of Interest

The authors declare no conflicts of interest relevant to this study.

Data Availability Statement

Observed aerosol number concentrations, particle size distributions, and CCN are available via F. Yu, Luo, et al. (2020), Marinescu et al. (2019), and Perkins et al. (2022). Part of data of particle size distributions and number concentrations were accessed at https://ebas-data.nilu.no/Default.aspx and https://csl.noaa.gov/groups/csl7/measurements/2013senex/. Figures were plotted using MATLAB R2023b and NCAR Command Language (NCL) version 6.3.0. MATLAB is available under its license at https://www.mathworks.com/products/matlab.html and NCL can be accessed under its license at https://www.ncl.ucar.edu/.

References

  1. An J, Lu Y, Huang DD, Ding X, Hu Q, Yan R, et al. (2024). Sectoral size-resolved particle number emissions with speciation: Emission profile-based quantification and a case study in the Yangtze River Delta Region, China. Journal of Geophysical Research: Atmospheres, 129(18), e2024JD041234. 10.1029/2024jd041234 [DOI] [Google Scholar]
  2. Antony C, & Vladimir K (2010). Hemispheric aerosol vertical profiles: Anthropogenic impacts on optical depth and cloud nuclei. Science, 329(5998), 1488–1492. 10.1126/science.1188838 [DOI] [PubMed] [Google Scholar]
  3. Appel KW, Bash JO, Fahey KM, Foley KM, Gilliam RC, Hogrefe C, et al. (2021). The community multiscale air quality (CMAQ) model versions 5.3 and 5.3.1: System updates and evaluation. Geoscientific Model Development, 14(5), 2867–2897. 10.5194/gmd-14-2867-2021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Beck J, Brown J, Dudhia J, Gill D, Hertneky T, Klemp J, et al. (2020). An evaluation of a hybrid, terrain-following vertical coordinate in the WRF-Based RAP and HRRR models. Weather and Forecasting, 35(3), 1081–1096. 10.1175/WAF-D-19-0146.1 [DOI] [Google Scholar]
  5. Binkowski FS, & Roselle SJ (2003). Models-3 community multiscale air quality (CMAQ) model aerosol component, 1, Model description. Journal of Geophysical Research, 108(D6), 4183. 10.1029/2001JD001409 [DOI] [Google Scholar]
  6. Binkowski FS, & Shankar U (1995). The regional particulate matter model: 1. Model description and preliminary results. Journal of Geophysical Research, 100(D12), 26191–26209. 10.1029/95jd02093 [DOI] [Google Scholar]
  7. Blichner SM, Sporre MK, & Berntsen TK (2021). Reduced effective radiative forcing from cloud–aerosol interactions (ERFaci) with improved treatment of early aerosol growth in an Earth system model. Atmospheric Chemistry and Physics, 21(23), 17243–17265. 10.5194/acp-21-17243-2021 [DOI] [Google Scholar]
  8. Carlton AG, Turpin BJ, Altieri KE, Seitzinger SP, Mathur R, Roselle SJ, & Weber RJ (2008). CMAQ model performance enhanced when in-cloud secondary organic aerosol is included: Comparisons of organic carbon predictions with measurements. Environmental Science & Technology, 42(23), 8798–8802. 10.1021/es801192n [DOI] [PubMed] [Google Scholar]
  9. Chang JS, Brost RA, Isaksen ISA, Madronich S, Middleton P, Stockwell WR, & Walcek CJ (1987). A three-dimensional Eulerian acid deposition model: Physical concepts and formation. Journal of Geophysical Research, 92(D12), 14681–14700. 10.1029/jd092id12p14681 [DOI] [Google Scholar]
  10. Chen L, Gao Y, Ma M, Wang L, Wang Q, Guan S, et al. (2023). Striking impacts of biomass burning on PM2.5 concentrations in Northeast China through the emission inventory improvement. Environmental Pollution, 318, 120835. 10.1016/j.envpol.2022.120835 [DOI] [PubMed] [Google Scholar]
  11. Chen X, Wang Z, Li J, & Yu F (2014). Development of a regional chemical transport model with size-resolved aerosol microphysics and its application on aerosol number concentration simulation over China. Sola, 10, 83–87. 10.2151/sola.2014-017 [DOI] [Google Scholar]
  12. Chen X, Yu F, Yang W, Sun Y, Chen H, Du W, et al. (2021). Global–regional nested simulation of particle number concentration by combing microphysical processes with an evolving organic aerosol module. Atmospheric Chemistry and Physics, 21(12), 9343–9366. 10.5194/acp-21-9343-2021 [DOI] [Google Scholar]
  13. Debry E, & Sportisse B (2007). Solving aerosol coagulation with size-binning methods. Applied Numerical Mathematics, 57(9), 1008–1020. 10.1016/j.apnum.2006.09.007 [DOI] [Google Scholar]
  14. Dusek U, Frank GP, Hildebrandt L, Curtius J, Schneider J, Walter S, et al. (2006). Size matters more than chemistry for cloud-nucleating ability of aerosol particles. Science, 312(5778), 1375–1378. 10.1126/science.1125261 [DOI] [PubMed] [Google Scholar]
  15. Elleman RA, & Covert DS (2009). Aerosol size distribution modeling with the community multiscale air quality modeling system in the Pacific Northwest: 1. Model comparison to observations. Journal of Geophysical Research, 114(D11), D11206. 10.1029/2008jd010791 [DOI] [Google Scholar]
  16. Foley KM, Pouliot GA, Eyth A, Aldridge MF, Allen C, Appel KW, et al. (2023). 2002–2017 anthropogenic emissions data for air quality modeling over the United States. Data in Brief, 47, 109022. 10.1016/j.dib.2023.109022 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Fountoukis C, & Nenes A (2007). ISORROPIA II: A computationally efficient thermodynamic equilibrium model for K+Ca2+–Mg2+–NH4+–Na+–SO42−–NO3−–Cl−–H2O aerosols. Atmospheric Chemistry and Physics, 7(17), 4639–4659. 10.5194/acp-74639-2007 [DOI] [Google Scholar]
  18. Gelbard F, Tambour Y, & Seinfeld JH (1980). Sectional representations for simulating aerosol dynamics. Journal of Colloid and Interface Science, 76(2), 541–556. 10.1016/0021-9797(80)90394-X [DOI] [Google Scholar]
  19. Guenther AB, Jiang X, Heald CL, Sakulyanontvittaya T, Duhl T, Emmons LK, & Wang X (2012). The model of emissions of gases and aerosols from nature version 2.1 (MEGAN2. 1): An extended and updated framework for modeling biogenic emissions. Geoscientific Model Development, 5(6), 1471–1492. 10.5194/gmd-5-1471-2012 [DOI] [Google Scholar]
  20. Hamill P, Turco RP, Kiang CS, Toon OB, & Whitten RC (1982). An analysis of various nucleation mechanisms for sulfate particles in the stratosphere. Journal of Aerosol Science, 13(6), 561–585. 10.1016/0021-8502(82)90021-0 [DOI] [Google Scholar]
  21. Han Y, Zhu T, Guan T, Zhu Y, Liu J, Ji Y, et al. (2016). Association between size-segregated particles in ambient air and acute respiratory inflammation. Science of the Total Environment, 565, 412–419. 10.1016/j.scitotenv.2016.04.196 [DOI] [PubMed] [Google Scholar]
  22. HEI. (2013). HEI review panel on ultrafine particles. In Understanding the health effects of ambient ultrafine particles, HEI perspectives (Vol. 3). Health Effects Institute. Boston, MA: Retrieved from https://www.healtheffects.org/system/files/Perspectives3.pdf [Google Scholar]
  23. Hogrefe C, Gilliam R, Mathur R, Henderson B, Sarwar G, Appel KW, et al. (2021). CMAQv5.3.2 ozone simulations over the Northern Hemisphere: Model performance and sensitivity to model configuration. In 20th Annual CMAS Conference, Nov 2, 2020, (virtual). Retrieved from https://www.cmascenter.org/conference/2021/slides/hogrefe-cmaq5-3-2-ozone-2021.pdf [Google Scholar]
  24. Jacobson MZ (1997). Development and application of a new air pollution modeling system II. Aerosol module structure and design. Atmospheric Environment, 31(2), 131–144. 10.1016/1352-2310(96)00202-6 [DOI] [Google Scholar]
  25. Jacobson MZ, Turco R, Jensen E, & Toon O (1994). Modeling coagulation among particles of different composition and size. Atmospheric Environment, 28(7), 1327–1338. 10.1016/1352-2310(94)90280-1 [DOI] [Google Scholar]
  26. Jacobson MZ, & Turco RP (1995). Simulating condensational growth, evaporation and coagulation of aerosols using a combined moving and stationary size grid. Aerosol Science & Technology, 22(1), 73–92. 10.1080/02786829408959729 [DOI] [Google Scholar]
  27. Jacquot O, & Sartelet K (2024). Numerical investigations on the modelling of ultrafine particles in SSH-aerosol-v1. 3a: Size resolution and redistribution. Geoscientific Model Development Discussions, 2024, 1–22. [Google Scholar]
  28. Karamchandani P, Vijayaraghavan K, Chen SY, Rochelle BB, & Knipping EM (2010). Development and application of a parallelized version of the advanced modeling system for transport, emissions, reactions and deposition of atmospheric matter (AMSTERDAM): 1. Model performance evaluation and impacts of plume–in–grid treatment. Atmospheric Pollution Research, 1(4), 260–270. 10.5094/apr.2010.033 [DOI] [Google Scholar]
  29. Knibbs LD, Cole-Hunter T, & Morawska L (2011). A review of commuter exposure toultrafine particles and its health effects. Atmospheric Environment, 45(16), 2611–2622. 10.1016/j.atmosenv.2011.02.065 [DOI] [Google Scholar]
  30. Kwon H-S, Ryu MH, & Carlsten C (2020). Ultrafine particles: Unique physicochemical properties relevant to health and disease. Experimental & Molecular Medicine, 52(3), 318–328. 10.1038/s12276-020-0405-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Lee JA, Alessandrini S, Kim JH, Meech S, Kumar R, Djalalova IV, & Wilczak JM (2024). Comparison of CAMS and CMAQ analyses of surface-level PM2.5 and O3 over the conterminous United States (CONUS). Atmospheric Environment, 338, 120833. 10.1016/j.atmosenv.2024.120833 [DOI] [Google Scholar]
  32. Lin S, Ryan I, Paul S, Deng X, Zhnag W, Luo G, et al. (2022). Particle surface area, ultrafine particle number concentration, and cardiovascular hospitalizations. Environmental Pollution, 310, 119795. 10.1016/j.envpol.2022.119795 [DOI] [PubMed] [Google Scholar]
  33. Luecken DJ, Yarwood G, & Hutzell WT (2019). Multipollutant modeling of ozone, reactive nitrogen and HAPs across the continental US with CMAQ-CB6. Atmospheric Environment, 201, 62–72. 10.1016/j.atmosenv.2018.11.060 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Luo G, & Yu F (2011). Simulation of particle formation and number concentration over the Eastern United States with the WRF-Chem+ APM model. Atmospheric Chemistry and Physics, 11(22), 11521–11533. 10.5194/acp-11-11521-2011 [DOI] [Google Scholar]
  35. Mao J, Zhang Y, Yu F, Bie S, Yu Q, Ma W, et al. (2025). Contribution of ship emissions to aerosol number concentrations: Parameterization of plume-scale nonlinear microphysics and application. Journal of Geophysical Research: Atmospheres, 130(4), e2024JD042867. 10.1029/2024jd042867 [DOI] [Google Scholar]
  36. Marinescu PJ, Levin EJT, Collins D, Kreidenweis SM, & van den Heever SC (2019). Quantifying aerosol size distributions and their temporal variability in the Southern Great Plains, USA. Atmospheric Chemistry and Physics, 19(18), 11985–12006. 10.5194/acp-19-11985-2019 [DOI] [Google Scholar]
  37. Martin RL, & Good TW (1991). Catalyzed oxidation of sulfur dioxide in solution: The iron-manganese synercism. Atmospheric Environment, 25A, 2395–2399. [Google Scholar]
  38. Moreno-Ríos AL, Tejeda-Benítez LP, & Bustillo-Lecompte CF (2022). Sources, characteristics, toxicity, and control of ultrafine particles: An overview. Geoscience Frontiers, 13(1), 101147. 10.1016/j.gsf.2021.101147 [DOI] [Google Scholar]
  39. Murphy BN, Nolte CG, Sidi F, Bash JO, Appel KW, Jang C, et al. (2021). The detailed emissions scaling, isolation, and diagnostic (DESID) module in the community multiscale air quality (CMAQ) modeling system version 5.3.2. Geoscientific Model Development, 14(6), 3407–3420. 10.5194/gmd-14-3407-2021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Murphy BN, Sonntag D, Seltzer KM, Pye HOT, Allen C, Murray E, et al. (2023). Reactive organic carbon air emissions from mobile sources in the United States. Atmospheric Chemistry and Physics, 23(20), 13469–13483. 10.5194/acp-23-13469-2023 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Murphy BN, Woody MC, Jimenez JL, Carlton AMG, Hayes PL, Liu S, et al. (2017). Semivolatile POA and parameterized total combustion SOA in CMAQv5.2: Impacts on source strength and partitioning. Atmospheric Chemistry and Physics, 17(18), 11107–11133. 10.5194/acp-1711107-2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Nair AA, Lin S, Luo G, Ryan I, Qi Q, Deng X, & Yu F (2023). Environmental exposure disparities in ultrafine particles and PM2.5 by urbanicity and socio-demographics in New York state, 2013–2020. Environmental Research, 239, 117246. 10.1016/j.envres.2023.117246 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Nolte CG, Appel KW, Kelly JT, Bhave PV, Fahey KM, Collett JL Jr., et al. (2015). Evaluation of the community multiscale air quality (CMAQ) model v5.0 against size-resolved measurements of inorganic particle composition across sites in North America. Geoscientific Model Development, 8(9), 2877–2892. 10.5194/gmd-8-2877-2015 [DOI] [Google Scholar]
  44. Nolte CG, Bhave P, Arnold J, Dennis R, Zhang K, & Wexler A (2008). Modeling urban and regional aerosols—Application of the CMAQ-UCD Aerosol Model to Tampa, a coastal urban site. Atmospheric Environment, 42(13), 3179–3191. 10.1016/j.atmosenv.2007.12.059 [DOI] [Google Scholar]
  45. Oliveira JRDC, Base LH, de Abreu LC, Ferreira Filho C, Ferreira C, & Morawska L (2019). Ultrafine particles and children’s health: Literature review. Paediatric Respiratory Reviews, 32, 73–81. 10.1016/j.prrv.2019.06.003 [DOI] [PubMed] [Google Scholar]
  46. Perkins RJ, Marinescu PJ, Levin EJT, Collins DR, & Kreidenweis SM (2022). Long- and short-term temporal variability in cloud condensation nuclei spectra over a wide supersaturation range in the Southern Great Plains site. Atmospheric Chemistry and Physics, 22(9), 6197–6215. 10.5194/acp-22-6197-2022 [DOI] [Google Scholar]
  47. Pierce JR, & Adams PJ (2009). Uncertainty in global CCN concentrations from uncertain aerosol nucleation and primary emission rates. Atmospheric Chemistry and Physics, 9(4), 1339–1356. 10.5194/acp-9-1339-2009 [DOI] [Google Scholar]
  48. Place BK, Hutzell WT, Appel KW, Farrell S, Valin L, Murphy BN, et al. (2023). Sensitivity of northeastern US surface ozone predictions to the representation of atmospheric chemistry in the community regional atmospheric chemistry multiphase mechanism (CRACMMv1.0). Atmospheric Chemistry and Physics, 23(16), 9173–9190. 10.5194/acp-23-9173-2023 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Pleim J, & Ran L (2011). Surface flux modeling for air quality applications. Atmosphere, 2(3), 271–302. 10.3390/atmos2030271 [DOI] [Google Scholar]
  50. Pun BK, Seigneur C, Vijayaraghavan K, Wu S, Chen S, Knipping EM, & Kumar N (2006). Modeling regional haze in the BRAVO study using CMAQ-MADRID: 1. Model evaluation. Journal of Geophysical Research, 111, D6. 10.1029/2004jd005608 [DOI] [Google Scholar]
  51. Pye HOT, Luecken DJ, Xu L, Boyd CM, Ng NL, Baker K, et al. (2015). Modeling the current and future roles of particulate organic nitrates in the southeastern United States. Environmental Science & Technology, 49(24), 14195–14203. 10.1021/acs.est.5b03738 [DOI] [PubMed] [Google Scholar]
  52. Pye HOT, Murphy BN, Xu L, Ng NL, Carlton AG, Guo H, et al. (2017). On the implications of aerosol liquid water and phase separation for organic aerosol mass. Atmospheric Chemistry and Physics, 17(1), 343–369. 10.5194/acp-17-343-2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Pye HOT, Place BK, Murphy BN, Seltzer KM, D’Ambro EL, Allen C, et al. (2023). Linking gas, particulate, and toxic endpoints to air emissions in the community regional atmospheric chemistry multiphase mechanism (CRACMM). Atmospheric Chemistry and Physics, 23(9), 5043–5099. 10.5194/acp-23-5043-2023 [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Pye HOT, Xu L, Henderson BH, Pagonis D, Campuzano-Jost P, Guo H, et al. (2024). Evolution of reactive organic compounds and their potential health risk in wildfire smoke. Environmental Science & Technology, 58(44), 19785–19796. 10.1021/acs.est.4c06187 [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Qi Q, Yu F, Nair AA, Lau SSS, Luo G, Mithu I, et al. (2024). Hidden danger: The long-term effect of ultrafine particles on mortality and its sociodemographic disparities in New York State. Journal of Hazardous Materials, 471, 134317. 10.1016/j.jhazmat.2024.134317 [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Qin M, Murphy BN, Isaacs KK, McDonald BC, Lu Q, McKeen SA, et al. (2021). Criteria pollutant impacts of volatile chemical products informed by near-field modeling. Nature Sustainability, 4(2), 129–137. 10.1038/s41893-020-00614-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Sartelet KN, Hayami H, Albriet B, & Sportisse B (2006). Development and preliminary validation of a modal aerosol model for tropospheric chemistry: MAM, aerosol. Science and Technology, 40(2), 118–127. 10.1080/02786820500485948 [DOI] [Google Scholar]
  58. Sarwar G, Godowitch J, Henderson B, Fahey K, Pouliot G, Hutzell B, et al. (2013). A comparison of atmospheric composition using the carbon bond and regional atmospheric chemistry mechanisms. Atmospheric Chemistry and Physics, 13(19), 9675–9712. 10.5194/acp-13-9695-2013 [DOI] [Google Scholar]
  59. Sarwar G, Hogrefe C, Henderson BH, Foley K, Mathur R, Murphy B, & Ahmed S (2023). Characterizing variations in ambient PM2.5 concentrations at the US Embassy in Dhaka, Bangladesh using observations and the CMAQ modeling system. Atmospheric Environment, 296, 119587. 10.1016/j.atmosenv.2023.119587 [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Semeniuk K, & Dastoor A (2018). Current state of aerosol nucleation parameterizations for air-quality and climate modeling. Atmospheric Environment, 179, 77–106. 10.1016/j.atmosenv.2018.01.039 [DOI] [Google Scholar]
  61. Shrivastava M, Zhang J, Zaveri RA, Zhao B, Pierce JR, O’Donnell SE, et al. (2024). Anthropogenic extremely low volatility organics (ELVOCs) govern the growth of molecular clusters over the Southern Great Plains during the springtime. Journal of Geophysical Research: Atmospheres, 129(21), e2024JD041212. 10.1029/2024JD041212 [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Skamarock WC, & Klemp JB (2008). A time-split nonhydrostatic atmospheric model for weather research and forecasting applications. Journal of Computational Physics, 227(7), 3465–3485. 10.1016/j.jcp.2007.01.037 [DOI] [Google Scholar]
  63. Spracklen D, Carslaw K, Kulmala M, Kerminen V-M, Si-hto S-L, Riipinen I, et al. (2008). Contribution of particle formation to global cloud condensation nuclei concentrations. Geophysical Research Letters, 35(6), L06808. 10.1029/2007GL033038 [DOI] [Google Scholar]
  64. Tilmes S, Mills MJ, Zhu Y, Bardeen CG, Vitt F, Yu P, et al. (2023). Description and performance of a sectional aerosol microphysical model in the Community Earth System Model (CESM2). Geoscientific Model Development, 16(21), 6087–6125. 10.5194/gmd-16-6087-2023 [DOI] [Google Scholar]
  65. Tiszenkel L, Stangl C, Krasnomowitz J, Ouyang Q, Yu H, Apsokardu MJ, et al. (2019). Temperature effects on sulfuric acid aerosol nucleation and growth: Initial results from the TANGENT study. Atmospheric Chemistry and Physics, 19(13), 8915–8929. 10.5194/acp-19-8915-2019 [DOI] [Google Scholar]
  66. Toon OB, Turco RP, Westphal D, Malone R, & Liu MS (1988). A multidimensional model for aerosols: Description of computational analogs. Journal of the Atmospheric Sciences, 45(15), 2123–2143. 10.1175/1520-0469(1988)045<2123:ammfad>2.0.co;2 [DOI] [Google Scholar]
  67. Turco RP, Hamill P, Toon OB, Whitten RC, & Kiang CS (1979). A one-dimensional model describing aerosol formation and evolution in the stratosphere, Part I, Physical processes and mathematical analogs. Journal of the Atmospheric Sciences, 36(4), 699–717. 10.1175/1520-0469(1979)036<0699:AODMDA>2.0.CO;2 [DOI] [Google Scholar]
  68. U.S. Environmental Protection Agency (USEPA). (2017). CMAQv5.2 operational guidance document (pp. 1–224). USEPA. [Google Scholar]
  69. Vehkamäki H, Kulmala M, Napari I, Lehtinen KEJ, Timmreck C, Noppel M, & Laaksonen A (2002). An improved parameterization for sulfuric acid-water nucleation rates for tropospheric and stratospheric conditions. Journal of Geophysical Research, 107(D22), 4622. 10.1029/2002JD002184 [DOI] [Google Scholar]
  70. Vignati E, Wilson J, & Stier P (2004). M7: An efficient size-resolved aerosol microphysics module for large-scale aerosol transport models. Journal of Geophysical Research, 109(D22), D22202. 10.1029/2003JD004485 [DOI] [Google Scholar]
  71. Wang M, Xiao M, Bertozzi B, Marie G, Rörup B, Schulze B, et al. (2022). Synergistic HNO3–H2SO4–NH3 upper tropospheric particle formation. Nature, 605(7910), 483–489. 10.1038/s41586-022-04605-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  72. Whitby ER, & McMurry PH (1997). Modal aerosol dynamics modeling. Aerospace Science and Technology, 27(6), 673–688. 10.1080/02786829708965504 [DOI] [Google Scholar]
  73. Whitby KT (1978). The physical characteristics of sulfur aerosols. Atmospheric Environment, 12(1–3), 135–159. 10.1016/0004-6981(78)90196-8 [DOI] [Google Scholar]
  74. Williamson CJ, Kupc A, Axisa D, Bilsback KR, Bui T, Campuzano-Jost P, et al. (2019). A large source of cloud condensation nuclei from new particle formation in the tropics. Nature, 574(7778), 399–403. 10.1038/s41586-019-1638-9 [DOI] [PubMed] [Google Scholar]
  75. Willmott CJ (1981). On the validation of models. Physical Geography, 2(2), 184–194. 10.1080/02723646.1981.10642213 [DOI] [Google Scholar]
  76. Xie Y, Paulot F, Carter W, Nolte C, Luecken D, Hutzell B, et al. (2013). Understanding the impact of recent advances in isoprene photooxidation on simulations of regional air quality. Atmospheric Chemistry and Physics, 13(16), 8439–8455. 10.5194/acp-13-8439-2013 [DOI] [Google Scholar]
  77. Yao L, Garmash O, Bianchi F, Zheng J, Yan C, Kontkanen J, et al. (2018). Atmospheric new particle formation from sulfuric acid and amines in a Chinese megacity. Science, 361(6399), 278–281. 10.1126/science.aao4839 [DOI] [PubMed] [Google Scholar]
  78. Yu F (1998). A study of the formation and evolution of aerosols and contrails in aircraft wakes: Development, validation and application of an advanced particle microphysics (APM) model (Doctoral Dissertation). UCLA. [Google Scholar]
  79. Yu F (2007). Improved quasi-unary nucleation model for binary H2SO4-H2O homogeneous nucleation. The Journal of Chemical Physics, 127(5), 054301. 10.1063/1.2752171 [DOI] [PubMed] [Google Scholar]
  80. Yu F (2011). A secondary organic aerosol formation model considering successive oxidation aging and kinetic condensation of organic compounds: Global scale implications. Atmospheric Chemistry and Physics, 11(3), 1083–1099. 10.5194/acp-11-1083-2011 [DOI] [Google Scholar]
  81. Yu F, Kärcher B, & Anderson BE (2024). Revisiting contrail ice formation: Impact of primary soot particle sizes and contribution of volatile particles. Environmental Science & Technology, 58(40), 17650–17660. 10.1021/acs.est.4c04340 [DOI] [PMC free article] [PubMed] [Google Scholar]
  82. Yu F, & Luo G (2009). Simulation of particle size distribution with a global aerosol model: Contribution of nucleation to aerosol and CCN number concentrations. Atmospheric Chemistry and Physics, 9(20), 7691–7710. 10.5194/acp-9-7691-2009 [DOI] [Google Scholar]
  83. Yu F, Luo G, Nadykto AB, & Herb J (2017). Impact of temperature dependence on the possible contribution of organics to new particle formation in the atmosphere. Atmospheric Chemistry and Physics, 17(8), 4997–5005. 10.5194/acp-17-4997-2017 [DOI] [Google Scholar]
  84. Yu F, Luo G, Nair AA, Schwab JJ, Sherman JP, & Zhang Y (2020). Wintertime new particle formation and its contribution to cloud condensation nuclei in the Northeastern United States. Atmospheric Chemistry and Physics, 20(4), 2591–2601. 10.5194/acp-20-2591-2020 [DOI] [Google Scholar]
  85. Yu F, Luo G, Pryor SC, Pillai PR, Lee SH, Ortega J, et al. (2015). Spring and summer contrast in new particle formation over nine forest areas in North America. Atmospheric Chemistry and Physics, 15(24), 13993–14003. 10.5194/acp-15-13993-2015 [DOI] [Google Scholar]
  86. Yu F, Nadykto AB, Herb J, Luo G, Nazarenko KM, & Uvarova LA (2018). H2SO4–H2O–NH3 ternary ion-mediated nucleation (TIMN): Kinetic-based model and comparison with CLOUD measurements. Atmospheric Chemistry and Physics, 18(23), 17451–17474. 10.5194/acp-18-17451-2018 [DOI] [Google Scholar]
  87. Yu F, Nadykto AB, Luo G, & Herb J (2020). H2SO4–H2O binary and H2SO4–H2O–NH3 ternary homogeneous and ion-mediated nucleation: Lookup tables version 1.0 for 3-D modeling application. Geoscientific Model Development, 13(6), 2663–2670. 10.5194/gmd-13-2663-2020 [DOI] [Google Scholar]
  88. Yu F, & Turco RP (1997). The role of ions in the formation and evolution of particles in aircraft plumes. Geophysical Research Letters, 24(15), 1927–1930. 10.1029/97GL01822 [DOI] [Google Scholar]
  89. Yu S, Eder B, Dennis R, Chu SH, & Schwartz SE (2006). New unbiased symmetric metrics for evaluation of air quality models. Atmospheric Science Letters, 7(1), 26–34. 10.1002/asl.125 [DOI] [Google Scholar]
  90. Zhang KM, & Wexler AS (2008). Modeling urban and regional aerosols—Development of the UCD Aerosol Module and implementation in CMAQ model. Atmospheric Environment, 42(13), 3166–3178. 10.1016/j.atmosenv.2007.12.052 [DOI] [Google Scholar]
  91. Zhang Y, Liu P, Liu XH, Jacobson MZ, McMurry PH, Yu F, et al. (2010). A comparative study of nucleation parameterizations: 2. Three-dimensional model application and evaluation. Journal of Geophysical Research, 115(D20). 10.1029/2010jd014151 [DOI] [Google Scholar]
  92. Zhao B, Donahue NM, Zhang K, Mao L, Shrivastava M, Ma PL, et al. (2024). Global variability in atmospheric new particle formation mechanisms. Nature, 631(8019), 98–105. 10.1038/s41586-024-07547-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  93. Zheng Z, West M, Zhao L, Ma P-L, Liu X, & Riemer N (2021). Quantifying the structural uncertainty of the aerosol mixing state representation in a modal model. Atmospheric Chemistry and Physics, 21(23), 17727–17741. 10.5194/acp-21-17727-2021 [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement1
NIHMS2107041-supplement-Supplement1.docx (1MB, docx)

Data Availability Statement

Observed aerosol number concentrations, particle size distributions, and CCN are available via F. Yu, Luo, et al. (2020), Marinescu et al. (2019), and Perkins et al. (2022). Part of data of particle size distributions and number concentrations were accessed at https://ebas-data.nilu.no/Default.aspx and https://csl.noaa.gov/groups/csl7/measurements/2013senex/. Figures were plotted using MATLAB R2023b and NCAR Command Language (NCL) version 6.3.0. MATLAB is available under its license at https://www.mathworks.com/products/matlab.html and NCL can be accessed under its license at https://www.ncl.ucar.edu/.

RESOURCES