CAPS: 1999 Publications

Three different methods are used to predict secondary organic aerosol (SOA) concentrations in the San Joaquin Valley of California during the winter of 1995-1996 [Integrated Monitoring Study, (IMS95)]. The first of these methods estimates SOA by using elemental carbon as a tracer of primary organic carbon. The second method relies on a Lagrangian trajectory model that simulates the formation, transport, and deposition of secondary organic aerosol. The model includes a recently developed gas-particle partitioning mechanism. Results from both methods are in good agreement with the chemical speciation of organic aerosol during IMS95 and suggest that most of the OC measured during IMS95 is of primary origin. Under suitable conditions (clear skies, low winds, low mixing heights) as much as 15-20 mu g C m(-3) of SOA can be produced, mainly due to oxidation of aromatics. The low mixing heights observed during the winter in the area allow accumulation of SOA precursors and the acceleration of SOA formation. Clouds and fog slow down the production of secondary compounds, reducing their concentrations by a factor of two or three from the above maximum levels. In addition, it appears that there is significant diurnal variation of SOA concentration. A strong dependence of SOA concentrations on temperature is observed, along with the existence of an optimal temperature for SOA formation. (C) 1999 Elsevier Science Ltd. All rights reserved. C1 Carnegie Mellon Univ, Dept Chem Engn, Pittsburgh, PA 15213 USA. Carnegie Mellon Univ, Dept Engn, Pittsburgh, PA 15213 USA. Carnegie Mellon Univ, Dept Publ Policy, Pittsburgh, PA 15213 USA. Carnegie Mellon Univ, Dept Civil Environm Engn, Pittsburgh, PA 15213 USA. Sonoma Technol Inc, Santa Rosa, CA 95403 USA.

The condensation of organic vapors on a nearly monodisperse externally mixed aerosol population was measured in order to test the hypothesis that organic species may preferentially condense on specific substrates. The organic species tested were glutaric acid, a typical secondary organic species, and DOP, a model organic for primary species. A series of organic and inorganic substrate pairs were investigated: NaCl, (NH4)(2)SO4, glutaric acid, and adipic acid. The growth of these aerosol populations was measured using a Tandem Differential Mobility Analyzer and the results were analyzed using transition regime mass transfer theory. The results show that no detectable preferential condensation occurs for either species on any of the substrates studied. The implications of these results to SOA formation models are discussed. C1 Carnegie Mellon Univ, Dept Chem Engn, Pittsburgh, PA 15213 USA.

Heterogeneous sulfate production in the remote marine boundary layer is studied by combining measurements collected during the first Southern Hemisphere marine Aerosol Characterization Experiment (ACE 1)with two numerical models: one simulating heterogeneous sulfate production in sea-salt aerosol water (SSAW) and the other simulating cloud processing. The models calculate oxidation of SO2 by O-3 and H2O2 via aqueous-phase reactions both in SSAW and in cloud droplets. Model calculations indicate that as much as 50-75% of the observed non-sea-salt (NSS) sulfate in particles in the ambient diameter size range of 0.9-16 mu m during the ACE 1 measurement period is due to production in SSAW, with cloud processing producing the remainder. The initial alkalinity of the wind-generated sea-salt aerosols can strongly influence sulfate production. Changing the effective alkalinity of SSAW from a value typical of seawater to a value based on the measurements taken at Cape Grim results in a better match to the observed NSS sulfate size distribution with up to 75% of the predicted supermicrometer NSS sulfate being due to heterogeneous conversion in SSAW. The average rates of removal of sulfate due to dry and wet deposition for the 0.4-16 mu m size range are 38 and 14 ng m(-3) d(-1) respectively. C1 Carnegie Mellon Univ, Dept Chem Engn, Pittsburgh, PA 15213 USA. Univ Colorado, Dept Environm & Geog, Denver, CO 80217 USA. Carnegie Mellon Univ, Dept Chem Engn, Pittsburgh, PA 15213 USA. Carnegie Mellon Univ, Dept Engn & Publ Policy, Pittsburgh, PA 15213 USA.

A comparison is conducted between 3 atmospheric equilibrium models: GFEMN, ISORROPIA, SCAPE2, and SEQUILIB. While ISORROPIA, SCAPE2, and SEQUILIB simplify the problem at hand in an effort to reduce computational rigor, GFEMN does not employ many of the simplifying assumptions used in previous models, thus allowing it to accurately predict multistage aerosol behavior and deliquescence depression. We examine model performance for representative atmospheric environments over an extended composition, temperature, and RH domain and against observations in Southern California. The predictions of GFEMN, ISORROPIA, SCAPE2, and SEQUILIB are in general agreement, but the latter 3 do not adequately reproduce multistage deliquescence behavior for multicomponent systems. The most notable differences in model predictions occur for H+ and aerosol water concentrations; discrepancies in predictions of aerosol nitrate and total dry inorganic Phl concentrations are not as significant. The models predict different deliquescence relative humidities for multicomponent systems, but for ammonia poor environments, these discrepancies do not introduce differences in total dry inorganic PM predictions. Against measurements taken during the Southern California Air Quality Study (SCAQS), all models qualitatively reproduce but generally underpredict aerosol nitrate concentrations. Finally, based on its overall agreement with GFEMN and its computational efficiency, ISORROPIA appears to be the model of choice for use in large-scale aerosol transport models. In places where crustal material comprises a significant portion of total PM, SCAPE2 is an alternative. C1 Carnegie Mellon Univ, Dept Chem Engn, Pittsburgh, PA 15213 USA. Carnegie Mellon Univ, Dept Engn & Publ Policy, Pittsburgh, PA 15213 USA.

The atmosphere overlying the ocean is very sensitive-physically chemically and climatically-to air pollution. Given that clouds over the ocean are of great climatic significance, and that sulphate aerosols seem to be an important control on marine cloud formation(1), anthropogenic inputs of sulphate to the marine atmosphere could exert an important influence on climate. Recently, sulphur emissions from fossil fuel burning by international shipping have been geographically characterized(2), indicating that ship sulphur emissions nearly equal the natural sulphur nux from ocean to atmosphere in many areas(3). Here we use a global chemical transport model to show that these ship emissions can be a dominant contributor to atmospheric sulphur dioxide concentrations over much of the world's oceans and in several coastal regions. The ship emissions also contribute significantly to atmospheric non-seasalt sulphate concentrations over Northern Hemisphere ocean regions and parts of the Southern Pacific Ocean, and indirect radiative forcing due to ship-emitted particulate matter (sulphate plus organic material) is estimated to contribute a substantial fraction to the anthropogenic perturbation of the Earth's radiation budget. The quantification of emissions from international shipping forces a re-evaluation of our present understanding of sulphur cycling and radiative forcing over the ocean. C1 Carnegie Mellon Univ, Dept Chem Engn, Pittsburgh, PA 15213 USA. Carnegie Mellon Univ, Dept Engn & Publ Policy, Pittsburgh, PA 15213 USA. Carnegie Mellon Univ, Dept Social & Decis Sci, Pittsburgh, PA 15213 USA. Duke Univ, Nicholas Sch Environm, Durham, NC 27708 USA.

The mass transfer rate of pure ammonium nitrate between the aerosol and gas phases was quantified experimentally by the use of the tandem differential mobility analyzer/scanning mobility particle sizer (TDMA/SMPS) technique. Ammonium nitrate particles 80-220 nm in diameter evaporated in purified air in a laminar flow reactor under temperatures of 20-27 degrees C and relative humidities in the vicinity of 10%. The evaporation rates were calculated by comparing the initial and final size distributions. A theoretical expression of the evaporation rate incorporating the Kelvin effect and the effect of relative humidity on the equilibrium constant is developed. The measurements were consistent with the theoretical predictions but there was evidence of a small kinetic resistance to the mass transfer rate. The discrepancy can be explained by a mass accommodation coefficient ranging from 0.8 to 0.5 as temperature increases from 20-27 degrees C. The corresponding timescale of evaporation for submicron NH,NO, particles in the atmosphere is of the order of a few seconds to 20 min. (C) 1999 Elsevier Science Ltd. All rights reserved. C1 Carnegie Mellon Univ, Dept Chem Engn & Engn, Pittsburgh, PA 15213 USA. Carnegie Mellon Univ, Dept Publ Policy, Pittsburgh, PA 15213 USA.

A computationally efficient and rigorous thermodynamic model (ISORROPIA) that predicts the physical stale and composition of inorganic atmospheric aerosol is presented. The advantages of this particular model render it suitable for incorporation into urban and regional air quality models. The model is embodied into the UAM-AERO air quality model, and the performance is compared with two other thermodynamic modules currently in use, SEQUILIB 1.5 and SEQUILIB 2.1. The new model yields predictions that agree with experimental measurements and the results of the other models, but at the same time proves to be much faster and computationally efficient. Using ISORROPIA accelerates the thermodynamic calculations by more than a factor of six, while the overall speed-up of UAM-AERO is at least twofold. This speedup is possible by the optimal solution of the thermodynamic equations, and the usage of precalculated tables, whenever possible. (C) 1999 Elsevier Science Ltd. All rights reserved. C1 Univ Aegean, Dept Environm Sci, GR-81100 Mytilene, Greece. Univ Miami, Rosenstiel Sch Marine & Atmospher Sci, Div Marine & Atmospher Chem, Miami, FL 33149 USA. Carnegie Mellon Univ, Dept Chem Engn, Pittsburgh, PA 15213 USA. Carnegie Mellon Univ, Dept Engn & Publ Policy, Pittsburgh, PA 15213 USA.

We present geographically resolved global inventories of nitrogen and sulfur emissions from international maritime transport for use in global atmospheric models. Current inventories of globally resolved sources of natural and anthropogenic emissions do not include the significant contribution of SO2 or NOx from oceangoing ships [Benkovitz et al., 1996]. We estimate the global inventory of ship emissions,using current emission test data for ships [Carlton et al., 1995] and a fuel-based approach similar to that used for automobile inventories [Singer and Harley, 1996]. This study estimates the 1993 global annual NOx and SO2 emissions from ships to be 3.08 teragrams (Tg, or 10(12) g) as N and 4.24 Tg S, respectively. Nitrogen emissions from ships are shown to account for more than 14% of all nitrogen emissions from fossil fuel combustion, and sulfur emissions exceed 5% of sulfur emitted by all fuel combustion sources including coal. Ship sulfur emissions correspond to about 20% of biogenic dimethylsulfide (DMS) emissions. In regions of the Northern Hemisphere, annual sulfur emissions from ships can be of the same order of magnitude as estimates of the annual flux of DMS [Chin et al., 1996] Monthly inventories of ship sulfur and nitrogen emissions presented in this paper are geographically characterized on a 2 degrees x 2 degrees resolution. Temporal and spatial characteristics of the inventory are presented, Uncertainty in inventory estimates is assessed: the fifth and ninety-fifth percentile values for global nitrogen emissions are 2.66 Tg N and 4.00 Tg N, respectively; the fifth and ninety-fifth percentile values for sulfur emissions are 3.29 Tg S and 5.61 Tg S, respectively. We suggest that these inventories, available via the Ship Emissions Assessment (SEA) web site, be used in models along with the Global Emissions Inventory Activity (GEIA) inventories for land-based anthropogenic emissions and modeled with ocean-biogenic inventories for DMS. C1 Carnegie Mellon Univ, Dept Engn & Publ Policy, Pittsburgh, PA 15213 USA. Carnegie Mellon Univ, Dept Chem Engn, Pittsburgh, PA 15213 USA. Carnegie Mellon Univ, Dept Social & Decis Sci, Pittsburgh, PA 15213 USA.

To evaluate the impact of aerosols on climate we must consider the aerosol dynamics of the remote marine atmosphere. Marine aerosols are subject to losses due to precipitation, dry deposition, and coagulation; yet, observed remote marine aerosol concentrations and size distributions are relatively constant. This maintenance of the aerosol distribution requires a particle source. This work focuses on the potential of H2SO4 nucleation within the marine boundary layer (MBL) to supply these particles. Spatial and temporal variability in meteorology and species concentrations are considered in a mathematical model to evaluate the effect of natural deviations from average MBL conditions on the highly nonlinear aerosol system. A dynamic, vertically dimensioned, size-resolved aerosol model is used with parameterized heterogeneous chemical processes. The results suggest that MBL nucleation may be an important source of new particle number in the remote MEL. However, though our model shows that typical remote MBL aerosol distributions can on average be maintained by MBL nucleation and sea-salt emissions, large oscillations in particle number concentration occur. Because such oscillations are only occasionally reported in measurements, MBL nucleation may not be the dominant source of new particles in the remote MEL. The nucleation events, which cause these oscillations, are predicted to occur at the top of the MBL after rain and/or entrainment of clean free tropospheric air. Predictions are particularly sensitive to the H2SO4 accommodation coefficient, nucleation tuner, and washout efficiency. Reduction of the accommodation coefficient is shown to increase the predicted accumulation mode concentration because the nucleation rate is enhanced. Entrainment of clean free tropospheric air is shown to increase the frequency of nucleation events within the MBL and may help to explain the observed correlation between subsidence and MBL small particles. A small constant addition of particles from the free troposphere, ocean, etc. suppresses H2SO4 nucleation and can lead to a reduction in total predicted aerosol number. This is because nucleation events require low total aerosol surface area and the constant addition of particles reduces the severity of aerosol surface area minimums. Larger external aerosol sources can maintain the observed remote MBL aerosol distribution. However, the temporal and spatial variability of such sources could have a large impact on aerosol concentrations and requires further investigation. C1 Carnegie Mellon Univ, Dept Chem Engn, Pittsburgh, PA 15213 USA. Duke Univ, Nicholas Sch Environm, Durham, NC 27708 USA. Carnegie Mellon Univ, Dept Engn & Publ Policy, Pittsburgh, PA 15213 USA.

Many existing models calculate the composition of the atmospheric aerosol system by solving a set of algebraic equations based on reversible reactions derived from thermodynamic equilibrium. Some models rely on an a priori knowledge of the presence of components in certain relative humidity regimes, and often fail to accurately predict deliquescence point depression and multistage aerosol growth. The present approach, relying on adjusted thermodynamic parameters of solid salts and a state of the art activity coefficient model, directly minimizes the Gibbs free energy (according to thermodynamic equilibrium principles) given temperature, relative humidity and the total (gas plus aerosol) ammonia, nitric acid, sulfate, sodium, and hydrochloric acid concentrations. A direct minimization, while requiring nb additional assumptions in its algorithm, allows the elimination of many of the assumptions used in previous models such as divided relative humidity (rh) and composition domains where only certain reactions are assumed to occur and constant DRH values despite varying temperature and composition. Moreover, the current approach predicts aerosol deliquescence and efflorescence behavior explaining the existence of supersaturated aerosol solutions. A comparison is conducted between our approach and available experimental results under several conditions. The current model agrees with experimental results for single salt systems although it shows sensitivity to thermodynamic parameters used in the minimization algorithm. A set of Delta G(f)(0) for solid salts is estimated that is consistent with available laboratory measurements and significantly improves model performance. I;or multicomponent systems, the current approach with adjusted Delta G(f)(0) accurately reproduces observed multistage growth patterns and deliquescence point depression over a broad temperature range. Finally, the direct Gibbs free energy minimization accurately reproduces aerosol efflorescence behavior. (C) 1999 Elsevier Science Ltd. All rights reserved. C1 Carnegie Mellon Univ, Dept Chem Engn, Pittsburgh, PA 15213 USA. Carnegie Mellon Univ, Dept Engn & Publ Policy, Pittsburgh, PA 15213 USA.

Soil permeability to air can increase substantially with measurement length scale. We tested the hypothesis that the scale effect could resolve large model underpredictions of radon and soil-gas entry into two experimental basement structures located in natural sandy-loam soil at a field site in Pen Lomond, CA. Previously, the model input for permeability at the site had been assessed based on 0.5-m scale measurements. After determining the soil-structure interaction scale (system scale) to be similar to 3 m, the model input was changed to reflect 3-m scale permeability measurements. This adjustment reduced unacceptably large model underpredictions, of a factor of 3 to 5, to a range near that of acceptable experimental error, 20 to 40%, The permeability scale effect may explain large and persistent model underestimates of radon entry into real houses, The results argue strongly for determining permeability at a length scale consistent with that of the system under study. C1 San Jose State Univ, Dept Environm Studies, San Jose, CA 95192 USA. Carnegie Mellon Univ, Dept Engn & Publ Policy, Pittsburgh, PA 15213 USA. Carnegie Mellon Univ, Dept Mech Engn, Pittsburgh, PA 15213 USA. Lawrence Berkeley Natl Lab, Indoor Environm Dept, Environm & Energy Technol Div, Berkeley, CA USA. Univ Calif Berkeley, Dept Civil & Environm Engn, Berkeley, CA 94720 USA.

We describe a novel modeling technique, based on Duhamel's theorem, to study the effects of time-varying winds on radon transport in soil near buildings. The technique, implemented in the model RapidSTART, reduces computational times for transient, three-dimensional, wind-induced soil-gas and radon transport by three to four orders of magnitude compared with conventional finite-difference models. To test model performance, we compared its predictions to analytical solutions of one-dimensional soil-column flow, finite-difference simulations of how around a full-scale house, and measurements of transient soil-gas and radon entry into an experimental basement structure. These comparisons demonstrate that RapidSTART accurately simulates time-dependent radon transport through soil and its entry into buildings. As demonstrated in a previous study, steady winds can significantly affect radon entry. In this paper, we extend the findings of that study by applying RapidSTART to explore the impacts of fluctuating wind speed and direction on radon entry into a prototypical house. In soils with moderate to high permeability, wind fluctuations have a small to moderate effect on the soil-gas radon concentration field and entry rate into the building. Fluctuating wind direction dominates the impact on radon entry rates, while fluctuating wind speed has little effect. For example, in a soil with a permeability of 10(-10) m(2), diurnal oscillations in wind direction can increase the predicted radon entry rate by up to 30% compared to steady-state predictions. (C) 1999 Elsevier Science Ltd. All rights reserved. C1 Univ Calif Berkeley, Dept Environm Sci Policy & Management, Ecosyst Sci Div, Berkeley, CA 94720 USA. Sandia Natl Labs, Combust Res Facil, Livermore, CA 94551 USA. Lawrence Berkeley Natl Lab, Indoor Environm Program, Berkeley, CA 94720 USA. Univ Calif Berkeley, Dept Civil & Environm Engn, Berkeley, CA 94720 USA.

Global sulfate aerosol composition is simulated online in the Goddard Institute for Space Studies general circulation model II' (GISS GCM II-prime). Four sulfur species, hydrogen peroxide, gas phase ammonia, and particulate ammonium are the prognostic tracer species, the emissions, transport, and deposition of which are explicitly simulated. Nitric acid fields are prescribed based on a global chemical transport model. An online thermodynamic equilibrium calculation determines the partitioning of ammonia and nitrate between gas and aerosol phases, and the quantity of aerosol water based on the temperature, relative humidity, and sulfate concentration in each GCM grid cell. The total global burden of sulfate, nitrate, ammonium, and aerosol water is 7.5 Tg and is most sensitive to changes in sulfur emissions. Tropospheric lifetimes for ammonium and ammonia are 4.2 and 0.9 days, respectively; the tropospheric ammonium burden is 0.30 Tg N, compared with 0.14 Tg N for ammonia. Simulated ammonium concentrations are generally within a factor of 2 of observations. Subgrid variability in measured concentrations hinders comparison of observations to predictions. Ammonium nitrate aerosol plays an important role in determining total aerosol mass in polluted continental areas. In the upper troposphere and near the poles, cold temperatures allow unneutralized nitric acid to condense into the aerosol phase. Acidic aerosol species tend to be neutralized by ammonia to a greater degree over continents than over oceans. The aerosol is most basic and gas phase ammonia concentrations are highest over India. Water uptake per mole of sulfate aerosol varies by two orders of magnitude because of changes in relative humidity and aerosol composition. Spatial variations in aerosol composition and water uptake have implications for direct and indirect aerosol radiative forcing. C1 CALTECH, Dept Chem Engn, Pasadena, CA 91125 USA. NASA, Goddard Inst Space Studies, New York, NY 10025 USA.

Previous estimates of dry deposition to water surfaces were generally based on deposition to flat, solid surfaces. This paper examines the effects of waves on dry deposition rates by numerically simulating particle trajectories over wave surfaces. Airflows over two-dimensional sine waves with height-to-length ratios 2a/lambda, = 0.1, 0.07, and 0.03 were calculated with a commercial computational fluid dynamics model. Results from the airflow simulations (velocity, kinetic energy, energy dissipation rate, and shear stress) provided inputs for a stochastic particle trajectory model. Particles were released from a height of 300 non-dimensional wall units at different locations along the wave. For those between 1 and 20 mu m, deposition was found to be greatest for particles released to the upslope portion of the wave,followed by the trough, crest and downslope. Overall deposition rates were enhanced due to the presence of waves. Increases ranged from 5% (d(p) = 80 mu m) to 100% (d(p) = 1 mu m) for waves with 2a/lambda = 0.07 and 0.1 and were approximately 50% greater (d(p) = 1 - 80 mu m) for 2a/lambda = 0.03. Deposition rates were enhanced due to increases in impaction and turbulent transport, both of which increase with increasing wave slope. However, an increased slope also produced regions of low or reversed flow in the trough and downslope, which decreased deposition rates. Due to these competing effects with respect to wave slope, deposition rates did not increase monotonically with wave slope. (C) 1999 Elsevier Science Ltd. All rights reserved. C1 Carnegie Mellon Univ, Dept Civil & Environm Engn, Pittsburgh, PA 15213 USA.

Wind tunnel measurements of particle dry deposition to wavy and flat surfaces were made to estimate the enhancement of deposition rates due to waves on water surfaces. Measurements were made of 4.0 and 6.7 mu m uranine particles at wind speeds of 5 and 10 m s(-1) to sinusoidal waves with height to length ratios 2a/lambda = 0.1 and 0.03 and to flat surfaces. Results showed that deposition was greatest to the upslope portion of the wave, accounting for 40-45% of the total mass, followed by the trough (30%), downslope (15%), and crest (10-15%). These results generally agreed within experimental variability with modeling predictions (Zufall et al., 1999). Deposition was enhanced at the upslope due to the effects of particle interception and impaction on the wave. Total deposition to the wave surfaces was greater than deposition to the flat surface for a large majority of the cases. The average increase in deposition to both wave surfaces for both particle sizes and wind speeds over deposition to the flat surface was 80%. (C) 1999 Elsevier Science Ltd. All rights reserved. C1 Carnegie Mellon Univ, Dept Civil & Environm Engn, Pittsburgh, PA 15213 USA.