Modeling urban and regional aerosols—condensation and evaporation near acid neutrality

doi:10.1016/S1352-2310(98)00059-4

Atmospheric Environment

Volume 32, Issue 20, 25 September 1998, Pages 3527-3531

https://doi.org/10.1016/S1352-2310(98)00059-4 Get rights and content

Abstract

Mathematically predicting the size and composition distribution of atmospheric aerosols can help to elucidate the complex link between emissions and particulate air quality. Wexler et al. (1994) identified and analyzed the atmospheric aerosol processes that govern particulate mass concentrations and estimated the relative importance of each term using parameters typical in South Coast Air Basin (SoCAB). The result was a general dynamic equation including only the relevant terms. In this paper we describe practical difficulties integrating these equations under the acid-neutral conditions common in the SoCAB. By introducing an acid equilibrium assumption, that is, the aerosol hydrogen ion concentration can be assumed to be in equilibrium with the gas-phase acidity, the mass distribution can be predicted using modest computer resources.

Introduction

Size- and composition-resolved modeling of aerosols is important to our understanding of urban and regional air pollution, and the development and evaluation of emission control strategies that comply with PM2.5 standards proposed by the US EPA. The first generation of aerosol models assumed thermodynamic equilibrium between the gas and aerosol phases for the volatile compounds to predict the total particulate mass (e.g. Russell and Cass, 1986; Bassett et al., 1991) and the particle size and composition distribution (e.g. Hogo et al., 1985; Pilinis and Seinfeld, 1988). Wexler and Seinfeld (1990) showed that equilibrium may not always hold, especially when the aerosol loading and temperature are low. In addition, even when equilibrium holds, the size distribution of the secondary compounds cannot be predicted by thermodynamic considerations alone; gas-aerosol transport must also be considered. Therefore, it is desirable for an aerosol model to explicitly simulate the transport of volatile species between gas and particles. Wexler et al. (1994) identified and analyzed the physical and chemical processes that may influence the particulate size and composition, and estimated the relative importance of each term using parameters typical in South Coast Air Basin (SoCAB). The resulting general dynamic equation was hence simplified to include only the dominant terms and algorithms for solving this equation were proposed. Recently, a hybrid model has been used to simulate SoCAB PM. The hybrid model uses a gas-aerosol equilibrium assumption to predict the gas-phase concentrations and then uses the aerosol transport moment to partition condensate over the size distribution (Lurmann et al., 1997). Although the hybrid model is computationally efficient, it introduces uncertainty in the size distribution because all aerosol sizes are combined during the thermodynamic equilibrium phase of the calculation. This may substantially alter the predicted partial pressures of the semi-volatile compounds NH₃, HNO₃ and HCl. In this work we solve the governing equations without employing a gas-aerosol equilibrium assumption. We discuss here some of the problems encountered in our implementation of the model to neutral acidity conditions in the SoCAB and our solutions to these problems. Neutral acidity aerosols are common in urban areas of western North America and western Europe. In the papers that will follow, we will describe the application of the model to two episodes from Southern California Air Quality Study (SCAQS).

Section snippets

The imade equation

The spatial and temporal variation, chemical composition and size distribution of atmospheric aerosols are influenced by a host of physical and chemical processes including emission and deposition, condensation and evaporation, advection and turbulent diffusion, coagulation, nucleation, gravitational settling, and aerosol-phase chemical reactions. Under the assumption of internally mixed aerosol, i.e., the particles of the same size have identical chemical composition, these processes are

The host gas-phase model and the aerosol module

Because secondary aerosol species are formed, by definition, by gas-phase processes, we must solve Eq. (1)in conjunction with a similar equation that describes the dynamics of the gas-phase species; therefore we must solve Eq. (1)within a host gas-phase airshed model. In this work we use version IV of the Urban Airshed Model (UAM) (SAI, 1990aSAI, 1990bSAI, 1990cSAI, 1990dSAI, 1990e) with improvements by Lurmann et al. (1997) which include: (1) a flexible gas-phase chemical mechanism interface

Acid equilibrium

One prominent characteristic in the western United States is that aerosol acidity is often neutralized by ammonia, and the mole fraction of hydrogen ions is typically less than 3% of the total cations. This scarcity of hydrogen ions has special significance in our understanding of the physical nature of condensation and evaporation in near acid neutral circumstances.

Let us consider the time scales over which the aerosol inorganic concentrations change significantly. For hydrogen ion this time

Coupled condensation/evaporation

Wexler and Seinfeld (1990) obtained expressions for the flux of acid and ammonia onto a particle for both the coupled and uncoupled situations. Using their flux expressions in the definition of H_i we obtain $H_{i} = πD_{p} D ̄ C ̄ (1+β) m 1− 1−4 C_{NH_{3, ∞}} C_{HX,∞} −K_{NH_{4}x} C ̄^{2} (1+β)^{2} (coupled)$ whenever the acid and base condensation is coupled, where $D ̄ = D_{NH_{3}} D_{HX}$ is the average diffusivity, $C ̄ =(D_{NH_{3}} C_{NH_{3}},∞+D_{HX} C_{HX,∞})/ D ̄$ is the diffusion weighted average of the concentrations far from the particle, β=2λ/(αD_p) where λ is the mean

Conclusions

An important idea in the aerosol model developed by Wexler et al. (1994) is the desirability of explicitly modeling the transport of volatile species between gas and particles as opposed to assuming equilibrium between these phases. In this paper we described some difficulties encountered in our attempt to implement the model under the special conditions in the SoCAB and introduced the acid equilibrium assumption to solve these problems. That is, for the near acid-neutral aerosols, the aerosol

Acknowledgements

Support for this work was provided by EPA grant CR 823634-01, California Air Resources Board grant 92-311, and the IBM Environmental Research Program.

References (24)

W.P.L. Carter
A detailed mechanism for the gas-phase atmospheric reactions of organic compounds
Atmospheric Environment
(1990)
W.P.L. Carter et al.
Evaluation of a detailed gas phase atmospheric reaction mechanism using environmental chamber data
Atmospheric Environment
(1991)
F.W. Lurmann et al.
Modeling urban and regional aerosolsII. Application to California’s South Coast Air Basin
Atmospheric Environment
(1997)
S.N. Pandis et al.
Secondary organic aerosol formation and transport
Atmospheric Environment
(1992)
C. Pilinis et al.
Development and evaluation of an Eulerian photochemical gas-aerosol model
Atmospheric Environment
(1988)
S. Potukuchi et al.
Identifying solid-aqueous phase transitions in atmospheric aerosols—II. Acidic solutions
Atmospheric Environment
(1995)
S. Potukuchi et al.
Predicting vapor pressures using neural networks
Atmospheric Environment
(1997)
A.G. Russell et al.
Verification of a mathematical model for aerosol nitrate and nitric acid formation and its use for control measure evaluation
Atmospheric Environment
(1986)
A.S. Wexler et al.
The Distribution of Ammonium Salts among a Size and Composition Dispersed Aerosol
Atmospheric Environment
(1990)
A.S. Wexler et al.
Modeling urban and regional aerosols—I. Model Development
Atmospheric Environment
(1994)

Bassett, M. E., Cassmassi, J. C., Durkee, K. R. and Hogo, H. (1991) Final air quality management plan, 1991 revision....

A. Bott

A positive definite advection scheme obtained by nonlinear renormalization of the advective fluxes

Monthly Weather Review

(1989)

Cited by (37)

Simulating the phase partitioning of NH<inf>3</inf>, HNO<inf>3</inf>, and HCl with size-resolved particles over northern Colorado in winter
2016, Atmospheric Environment
Citation Excerpt :
However, representing the processes that influence inorganic particle constituents in air quality models is computationally challenging, particularly for large-scale (e.g., continental U.S.) applications. Approaches have been developed in the last few decades to efficiently simulate inorganic aerosol thermodynamics (e.g., Jacobson, 1999; Kim et al., 1993; Nenes et al., 1998; Wexler and Seinfeld, 1991; Zaveri et al., 2005a, 2005b) and dynamics (e.g., Binkowski and Shankar, 1995; Jacobson, 2005; Pilinis et al., 2000; Sun and Wexler, 1998; Zaveri et al., 2008) in air quality models. Yet, despite the ability of these approaches to represent key aerosol processes, challenges remain in simulating inorganic particle size-composition distributions (e.g, Nolte et al., 2015).
Numerical modeling of inorganic aerosol processes is useful in air quality management, but comprehensive evaluation of modeled aerosol processes is rarely possible due to the lack of comprehensive datasets. During the Nitrogen, Aerosol Composition, and Halogens on a Tall Tower (NACHTT) campaign in February and March 2011, the phase partitioning of soluble trace gases with size-resolved particles and related meteorological conditions were measured continuously at a site in Colorado about 33 km north of Denver. These size-resolved measurements of particulate ${SO}_{4}^{2 -}$ , ${NH}_{4}^{+}$ , ${NO}_{3}^{-}$ , Cl⁻, Na⁺, Ca²⁺, Mg²⁺, and K⁺ and inorganic gases are used here to assess the ability of a continental U.S. modeling platform to simulate the gas-particle partitioning of NH₃, HNO₃, and HCl at this location. Modeling is based on the Community Multiscale Air Quality (CMAQ) model with 12 km horizontal resolution. Baseline, sensitivity, and source apportionment simulations are conducted to fully characterize the model predictions. Considering the limitations in representing regional terrain features in the national 12 km modeling, predictions agree reasonably well with measured concentrations and gas-particle partitioning at this location. However, the median sum of ${NH}_{4}^{+}$ and NH₃ is underpredicted in the baseline simulation by a factor of four suggesting a need for improved bottom-up NH₃ emissions inventories in this area. The median sum of Cl⁻ and HCl is underpredicted by a factor of 2.8, while levels of Ca²⁺, Mg²⁺, and K⁺ are overpredicted. Improvements in windblown and fugitive dust emissions may improve and/or better constrain these predictions. Mass size distributions for inorganic particle constituents are generally simulated well, although the modeled fine particle mode is shifted to slightly larger diameters relative to measurements. Source apportionment modeling estimates of source sector and boundary contributions to air quality at the site are provided.
Simulating the fine and coarse inorganic particulate matter concentrations in a polluted megacity
2010, Atmospheric Environment
Citation Excerpt :
The major disadvantage of this approach is that infinite solutions to the problem may exist if the particles are solid (Wexler and Seinfeld, 1990). In the “dynamic” approach, active mass transfer is considered for each aerosol size “section” or “mode” (Meng and Seinfeld, 1996; Meng et al., 1998; Jacobson et al., 1996; Jacobson, 1997a,b; Sun and Wexler, 1998a,b; Pilinis et al., 2000; Zhang and Wexler, 2008; Zaveri et al., 2008). Although this approach is the most accurate, it is much more computationally intensive than the bulk equilibrium approach.
A three dimensional chemical transport model (PMCAMx) is applied to the Mexico City Metropolitan Area (MCMA) in order to simulate the chemical composition and mass of the major PM₁ (fine) and PM_1–10 (coarse) inorganic components and determine the effect of mineral dust on their formation. The aerosol thermodynamic model ISORROPIA-II is used to explicitly simulate the effect of Ca, Mg, and K from dust on semi-volatile partitioning and water uptake. The hybrid approach is applied to simulate the inorganic components, assuming that the smallest particles are in thermodynamic equilibrium, while describing the mass transfer to and from the larger ones. The official MCMA 2004 emissions inventory with improved dust and NaCl emissions is used. The comparison between the model predictions and measurements during a week of April of 2003 at Centro Nacional de Investigacion y Capacitacion Ambiental (CENICA) “Supersite” shows that the model reproduces reasonably well the fine mode composition and its diurnal variation. Sulfate predicted levels are relatively uniform in the area (approximately 3 μg m⁻³), while ammonium nitrate peaks in Mexico City (approximately 7 μg m⁻³) and its concentration rapidly decreases due to dilution and evaporation away from the urban area. In areas of high dust concentrations, the associated alkalinity is predicted to increase the concentration of nitrate, chloride and ammonium in the coarse mode by up to 2 μg m⁻³ (a factor of 10), 0.4 μg m⁻³, and 0.6 μg m⁻³ (75%), respectively. The predicted ammonium nitrate levels inside Mexico City for this period are sensitive to the physical state (solid versus liquid) of the particles during periods with RH less than 50%.
Designing ambient particulate matter monitoring program for source apportionment study by receptor modeling
2009, Atmospheric Environment
Citation Excerpt :
In addition, the emission inventory cannot directly account for secondary particles that are formed in the atmosphere. Dispersion modeling, in principle, can estimate the impact of source emissions to ambient air pollution levels and advanced chemical transport models have the ability to predict secondary aerosol formation in the atmosphere (Mysliwiec and Kleeman, 2002; Sun and Wexler, 1998). However, advanced 3D dispersion models dealing with PM formation require intensive input data, e.g. detailed gridded emission inventory, and other resources that may not be readily available everywhere.
Particulate matter (PM) receptor modeling requires specific intensive input data that is always a challenge to produce cost effectively. A well-designed monitoring program is important to collect such PM ambient data in urban areas with diverse and densely distributed sources. This paper presents a general framework for designing such a monitoring program while emphasizing appropriate quality assurance and quality control elements that are particularly applicable where limited resources are available. Topics for discussion include selection of monitoring sites, sampling and analytical techniques, and the uncertainty estimation for ambient concentration input data. The design framework is illustrated by a case study of a monitoring program for PM source apportionment in the Bangkok Metropolitan Region in which 24-h fine and coarse PM samples were collected using two collocated dichotomous samplers. Comparison between black carbon measurements by Smoke Stain reflectometry and Thermal Optical Transmittance method is highlighted.
Modeling urban and regional aerosols-Development of the UCD Aerosol Module and implementation in CMAQ model
2008, Atmospheric Environment
This paper presents a mechanistic, fully dynamic, internally mixed, sectional aerosol module, the UCD Aerosol Module, which evolves from the Aerosol Inorganic Model (AIM). The UCD Aerosol Module employs three gas-to-particle mass transport schemes, replacement, coupled and uncoupled, and simplified aerosol thermodynamics to predict gas–particle partitioning, aerosol phase state and water content efficiently and reasonably. Aerosol dynamics is integrated using an Asynchronous Time-Stepping (ATS) integration method, where different sized particles integrate with different time scales. CMAQ-UCD is an implementation of the UCD Aerosol Module in the CMAQ modeling system. With all of these features, CMAQ-UCD is designed to be a computationally efficient and scientifically sound air quality model that predicts particulate mass concentrations and size distributions over urban or regional scales for regulatory and scientific purposes.
Modeling urban and regional aerosols-Application of the CMAQ-UCD Aerosol Model to Tampa, a coastal urban site
2008, Atmospheric Environment
The University of California at Davis (UCD) aerosol module, an internally mixed, sectional aerosol model with dynamic mass transfer between the gas and particle phases, has been coupled to the Community Multiscale Air Quality (CMAQ) model. This paper describes the application of the CMAQ-UCD model to simulate air quality in Tampa, a large city with a population of 2M on the west coast of Florida, USA. Modeled aerosol size and composition distributions are evaluated against size-segregated ambient measurements of ${SO}_{4}^{2 -}$ , ${NH}_{4}^{+}$ , ${NO}_{3}^{-}$ , ${Na}^{+}$ , and ${Cl}^{-}$ collected at three Tampa-area sites during May 2002, and against semi-continuous HNO₃ and total aerosol ${SO}_{4}^{2 -}$ , ${NH}_{4}^{+}$ , ${NO}_{3}^{-}$ , and ${Cl}^{-}$ measurements collected at a single site. Sea-salt emissions over the open ocean and the surf zone are parameterized as a function of modeled wind speed and relative humidity. Modeled total aerosol sulfate and ammonium concentrations and size distributions agree with measurements, with an overall normalized mean bias (NMB) of 2% and −23% and normalized mean error (NME) of 46% and 38%, respectively, and correctly identifying the size bin in which the peak concentration is observed. Sea-salt size distributions are also simulated well, with the distribution dominated by the coarse mode and total aerosol sodium and chloride NMB of −2% and 17% and NME of 32% and 38%. Though the model correctly identifies that nitrate is predominantly in the coarse $(D_{p} > 2.5 μ m)$ size sections, aerosol nitrate concentrations are underpredicted by a factor of two. The availability of highly time-resolved measurements provides a unique opportunity to evaluate the model's partitioning of total nitrate and the simulation of chloride depletion as a function of particle size.
Role of ammonia chemistry and coarse mode aerosols in global climatological inorganic aerosol distributions
2007, Atmospheric Environment
We use an inorganic aerosol thermodynamic equilibrium model in a three-dimensional chemical transport model to understand the roles of ammonia chemistry and natural aerosols on the global distribution of aerosols. The thermodynamic equilibrium model partitions gas-phase precursors among modeled aerosol species self-consistently with ambient relative humidity and natural and anthropogenic aerosol emissions during the 1990s.
Model simulations show that accounting for aerosol inorganic thermodynamic equilibrium, ammonia chemistry and dust and sea-salt aerosols improve agreement with observed SO₄, NO₃, and NH₄ aerosols especially at North American sites. This study shows that the presence of sea salt, dust aerosol and ammonia chemistry significantly increases sulfate over polluted continental regions. In all regions and seasons, representation of ammonia chemistry is required to obtain reasonable agreement between modeled and observed sulfate and nitrate concentrations. Observed and modeled correlations of sulfate and nitrate with ammonium confirm that the sulfate and nitrate are strongly coupled with ammonium. SO₄ concentrations over East China peak in winter, while North American SO₄ peaks in summer. Seasonal variations of NO₃ and SO₄ are the same in East China. In North America, the seasonal variation is much stronger for NO₃ than SO₄ and peaks in winter.
Natural sea salt and dust aerosol significantly alter the regional distributions of other aerosols in three main ways. First, they increase sulfate formation by 10–70% in polluted areas. Second, they increase modeled nitrate over oceans and reduce nitrate over Northern hemisphere continents. Third, they reduce ammonium formation over oceans and increase ammonium over Northern Hemisphere continents. Comparisons of SO₄, NO₃ and NH₄ deposition between pre-industrial, present, and year 2100 scenarios show that the present NO₃ and NH₄ deposition are twice pre-industrial deposition and present SO₄ deposition is almost five times pre-industrial deposition.

View all citing articles on Scopus

View full text