1. Introduction
Anthropogenic aerosols are one of the major contributors to human-induced climate change (e.g., Twomey 1974; Charlson et al. 1992). A substantial effort has been made to quantify the magnitude of the aerosol effect both through direct scattering and absorption of shortwave radiation as well as indirectly on cloud formation, cloud characteristics, and precipitation. Clouds themselves also play an important role in transporting, scavenging, and processing aerosols. Convective clouds have been recognized as an important mechanism for transferring chemical compounds from the surface to the free troposphere (e.g., Ridley et al. 2004; Lawrence et al. 2003; Pickering et al. 2001; Wang and Prinn 2000). Several observations (e.g., Clarke 1992, 1993; Nyeki et al. 1999; Ström et al. 1999; Clarke and Kapustin 2002; Petzold et al. 2002; Schröder et al. 2002; Twohy et al. 2002; Hermann et al. 2003; Minikin et al. 2003; Lee et al. 2004) have indicated high number concentrations of small aerosols in the vicinity of the anvils of convective clouds. One theory has been that the environment in this area is favorable for the nucleation of particles as both relative humidity and concentration of aerosol precursors are relatively high. Another explanation for the small aerosols found near the top of convective clouds could be direct transport from the surface to the free troposphere by strong vertical updrafts. However, a great number of aerosols are water soluble and/or effective as cloud condensation nuclei (CCN) and are hence likely to be scavenged by heavy precipitation. In addition, aerosols in the free troposphere can serve as ice nuclei and are thus likely to be scavenged through the formation of ice clouds. An important fact is that the lifetime of aerosols is much longer in the free troposphere than in the planetary boundary layer. The free tropospheric aerosols may thereby have a major influence on the earth’s radiative budget.
A few adiabatic parcel model studies have been performed to examine the origin of free tropospheric aerosols (deReus et al. 1998; Kulmala et al. 1998; Clement et al. 2002). Zhang et al. (1998) incorporated a two-moment aerosol model into a two-dimensional cloud and sulfate chemistry model to simulate the effects of clouds on aerosol redistribution and production in cumulonimbus clouds. They found that the nucleation rate after cloud dissipation in the upper troposphere increased by one order of magnitude compared to the nucleation rate before cloud formation. Using a 2D cloud-resolving model including an explicit aerosol module, Ekman et al. (2004) found that the direct transport of small-/medium-sized particles (5.84 nm ≤ d ≤ 31.0 nm) within a convective cloud is substantial. Up to 10% of the surface concentration may reach the free troposphere.
In the present study, a 3D version of the aerosol cloud-resolving model (CRM) utilized by Ekman et al. (2004) is employed. We evaluate the performance of the aerosol CRM by examining if the model is able to reproduce important features of observed variables within a convective cloud. We also further examine the magnitude of direct transport of aerosols within a convective cloud and estimate the number of new aerosols formed in the vicinity of the anvil. The model and the simulated case are first described in section 2. Thereafter we compare the 3D model with available observations in section 3. An estimate of the aerosol transport within the convective cloud is presented in section 4. In section 5 we present results of a box model calculation of new particle formation and aerosol growth performed at the top of the cloud and driven by the output of the 3D model data fields. Summary and conclusions are given in the last section.
2. Model and simulated case
a. Cloud-resolving model
The present version of the CRM is a 3D version of the model reported in Ekman et al. (2004). There are many benefits of using a 3D model instead of a 2D model. Using a 3D model, it is possible to represent horizontal variations in the initial tracer compound data. These variations could be important for simulating the correct average and maximum amount of tracer compounds that are transported from the surface to the cloud anvil. Deep convective clouds are also known themselves to have complicated spatial structures that are impossible to be fully captured using a 2D model. Spatial variations in, for example, cloud updraft and downdraft may be important when average statistics of transported aerosol concentrations from the surface to the cloud anvil are calculated.
The dynamics–physics module in the CRM consists of the nonhydrostatic momentum equations, the continuity equations for water vapor and airmass density, the thermodynamic equation, and the equation of state (Wang and Chang 1993a). Also included are prognostic equations for the mixing ratios as well as number concentrations of cloud droplets, raindrops, ice crystals, and graupel particles. The microphysical transformations are formulated based on a “two moment” scheme incorporating the size spectra of particles (Wang and Chang 1993a; Wang et al. 1995). A δ-four-stream radiation module based on Fu and Liou (1993) is incorporated in the model using predicted concentrations of gases (including H2O and O3) and hydrometeors to calculate radiative fluxes and heating rates.
In the CRM, the number of cloud condensation nuclei (CCN) and ice nuclei (IN) available for cloud droplet nucleation is predicted using the aerosol module by including the transport, mixing, and various physical and chemical conversions of aerosols in the model (cf. next section). The chemistry submodule predicts atmospheric concentrations of 25 gaseous and 16 aqueous (in both cloud droplets and raindrops) chemical compounds including important aerosol precursors, such as sulfate and nitrate, undergoing more than 100 reactions as well as transport and microphysical conversions. A module of heterogeneous chemistry on ice particles has been developed and is included in the present version of the model (Wang 2005). This module calculates surface uptake of several key chemical species including HNO3, SO2, H2O2, and CH3OOH by ice particles.
The CRM has been used in studies of dynamics, microphysics, chemistry, and aerosol transport in continental deep convection (e.g., Wang and Chang 1993a, b; Wang and Crutzen 1995; Ekman et al. 2004) and deep convection over the Pacific (Wang et al. 1995; Wang and Prinn 1998, 2000). Results of the chemical and dynamical parts of the model have also been compared with available observations including aircraft, radar, and satellite data. The spatial resolution of the model can be flexibly set; a horizontal grid interval of 2 km and a vertical grid interval of 400 m are used in this study.
b. Aerosol module
The evolution in time and space of aerosols consisting of sulfate, organic carbon, black carbon, and mixtures thereof is described using a multimodal aerosol model originally developed by Wilson et al. (2001) and modified by Ekman et al. (2004). A schematic picture of the module is shown in Fig. 1. Five different modes are used to represent the aerosol population. These five modes are 1) nucleation mode sulfate aerosols (here defined by 0 ≤ d ≤ 5.84 nm), 2) Aitken mode sulfate aerosols (here defined by 5.84 nm ≤ d ≤ 31.0 nm), 3) accumulation mode sulfate aerosols (here defined by d ≥ 31.0 nm), 4) mixed aerosols, 5) and black carbon (BC) aerosols. In the present version of the model, mixed mode aerosols are assumed to have basically the same properties as sulfate aerosols; that is, they have the same density as sulfate aerosols, are water soluble, and may serve as both CCN and IN. One difference though is that the mixed mode aerosols have a source from the BC mode, as BC aerosols “age” when H2SO4 condense on the aerosols.
The size distribution within each aerosol mode is assumed to be lognormal and is described by three parameters: number, mass, and standard deviation. To reduce the computational burden, the standard deviations are prescribed (1.59 for all sulfate modes and 2.0 for the mixed and BC mode). To close these aerosol spectra, both number concentrations and mass mixing ratios of the five aerosol modes, that is, all together 10 variables, are incorporated in the cloud-resolving model as prognostic variables undergoing transport, mixing, dry deposition, and nucleation as well as impact scavenging besides aerosol microphysical processes (Ekman et al. 2004). The advection scheme used to calculate the transport of these aerosol variables is the same revised Bott scheme as developed in Wang and Chang (1993a).
In the default version of the model, due to the short integration time, emissions of both carbonaceous and sulfuric aerosols are set to zero. Therefore, carbonaceous aerosols have no additional sources other than the given initial loadings during the model integration. A continuous source in addition to the given initial loading of the whole sulfate aerosol population (three modes) is the nucleation of new aerosols from H2SO4 (supplied by SO2 oxidation calculated in the chemistry module of the model) and H2O (Vehkamäki et al. 2002). The condensation coefficient as well as the intra- and intermodal coagulation coefficients for each aerosol mode is determined from the theory of Fuchs (1964), using the geometric mean radius of each mode.
The activation of a drop at a certain supersaturation depends on the composition of the solute. The number of aerosols available to form cloud droplets is determined by calculating the critical radius corresponding to the critical saturation ratio for drop activation using the Köhler equation (cf. Ekman et al. 2004). For any aerosol size bin that has the critical radius R* within its boundaries (Rmin < R* < Rmax), the bin is split so that only particles with radius larger than R* are activated. The total number of aerosols activated can be obtained by integrating the distribution function from R* to Rmax. Nucleation of ice crystals can occur both by homogeneous-freezing nucleation of liquid particles and by heterogeneous nucleation caused by aerosol particles, but it is unclear which process is dominating in the atmosphere (e.g., Cziczo et al. 2004; Haag and Kärcher 2004; DeMott et al. 2003; Haag et al. 2003). Both of the above nucleation processes are included in the CRM. Only pure sulfate aerosols and mixed aerosols are considered to constitute CCN or IN. An additional path for nucleation of ice crystals, that is, direct freezing of aerosols to form ice crystals, has been proposed recently (e.g., Kärcher and Lohmann 2003). This process is not included in our model.
Another path for scavenging of aerosols is through collision with falling raindrops, graupel, or ice crystals, that is, the precipitation (impaction) scavenging. In the model, the collision efficiency E varies with size and is prescribed for the different aerosol bins (as in Ekman et al. 2004). The removal is efficient for small and large particles whereas the collision efficiency for particles in the size range from 0.1 to 1.0 μm is relatively small. Resuspension is not treated in the CRM; that is, the aerosols are assumed to be scavenged when they are in droplets or ice crystals.
c. Simulated case
The selected case to simulate is a cumulonimbus cloud with an extended anvil over northern Germany, observed during the Stratosphere–Troposphere Experiment by Aircraft Measurements (STREAM) on 29 July 1994 (Ström et al. 1999, hereafter S99). This case is the same as simulated in Ekman et al. (2004), but in the present study the 3D version of the model is utilized instead of the 2D version. At the observed location, several smaller groups of thunderstorms were formed along a cold front, and aircraft measurements of aerosols, cloud water, relative humidity, carbon monoxide, and ozone were conducted along a cross section through the center of the anvil of one of these storm clouds. The research aircraft entered the cumulonimbus cloud at approximately 1436 UTC at an altitude of ∼10 400 m. The aircraft traveled a distance of ∼260 km across a frontal zone before leaving the cloudy air at about 1503 UTC. In situ data from this level are presented in Fig. 5 in S99.
The meteorological part of the CRM simulation is initialized using analyzed 3D initial data fields of pressure, temperature, winds, and specific humidity obtained from the National Centers for Environmental Prediction (NCEP). Horizontally interpolated fields of NO2, O3, and SO2 obtained from surface observations conducted by the Cooperative Program for Monitoring and Evaluation of the Long-Range Transmission of Air Pollutants in Europe (EMEP) (Hjellbrekke and Hanssen 1998) are used to initialize the chemistry module. There are no observations of gaseous HNO3 or H2SO4 available from EMEP. For HNO3, the initial concentration is instead obtained by combining results from previous simulations using the CRM (Wang and Chang 1993a; Wang and Crutzen 1995) and measured particulate concentrations of NO−3 from EMEP. For H2SO4, the initial concentration is assumed to be equal to zero. Vertical profiles of all chemical compounds are prescribed as to decrease with height except for O3, which is based on previous work (for initial profiles of SO2, CO, and O3, cf. Ekman et al. 2004).
For black carbon and mixed mode aerosols, a horizontally constant surface concentration of 500 cm−3 is assumed initially, corresponding to a mass concentration of 100 ng m−3. The surface concentration of Aitken mode and accumulation mode aerosols is set to be 50 000 and 3000 cm−3, respectively. These aerosol concentrations are representative for what may be observed in urban continental air (R. Krejci 2004, personal communication; Seinfeld and Pandis 1998). All aerosol concentrations are initially prescribed to decrease with height as a function of air density (Fig. 2). This type of vertical dependence is in fairly good agreement with the observations by Schröder et al. (2002) and Petzold et al. (2002). The initial mass concentration for each mode is calculated by assuming spherical particles with a density of 1.7 g cm−3 and a radius of 6.29 nm for the Aitken mode and 48.5 nm for the accumulation, mixed, and black carbon modes. The nucleation mode aerosol concentration is assumed to be zero at the beginning of the simulation.
3. Comparison with observations
Several complications arise when comparing the simulated and observed properties of aerosols and gases within the convective cloud. The observed and modeled clouds are not identical. They can only be considered to represent different “samples” along the line of convection. Hence, the time scale of the cloud development and the location of the anvil may slightly differ. In addition, the measurements are conducted along one cross section within the cloud, whereas the model output may be collected in all cloudy grid points at a certain level. Nevertheless, the general characteristics of the modeled and observed cloud should be similar. A comparison should give us an indication of potential shortcomings of the chemistry and aerosol modules.
After 3 hours of simulation, the cloud has reached its decaying state, which is also the time reported for the observations. General features of the cloud during the simulation are summarized in Table 1. The modeled cloud anvil covers an area of approximately 100 × 100 km2 (Fig. 3), which is somewhat smaller than what was observed by the aircraft measurements (the aircraft was estimated to fly 150 km through cloudy air). In addition, the observed cloud appeared to have two major bodies, whereas the modeled cloud only consists of one body (cf. Fig. 4 in the present study and Fig. 5 in S99).
a. Chemical compounds
Carbon monoxide is in general a good indicator of vertical transport of air from a polluted boundary layer to the free troposphere, as CO is emitted mainly from combustion processes and it is not affected by chemical reactions on a time scale of a few hours. High concentrations of O3 in the free troposphere may, on the other hand, both be an indicator of vertical transport of polluted boundary layer air (photochemical smog) as well as downward transport of ozone-rich stratospheric air. The modeled minimum, average, and maximum concentrations of CO and O3 along the cross section at 10.4-km altitude agree fairly well with the observed corresponding values (approximately 20% difference, cf. Fig. 5). There is a tendency to underestimate the CO concentration and overestimate the O3 concentration, which could be a sign of too little transport of air from the boundary layer. Before the convective event on 29 July, a buildup of high concentrations of O3, CO, SO2, and other pollutants in the boundary layer occurred due to several weeks of clear weather and weak winds. According to S99, a strong correlation between O3 and CO was noted as high as 10.4 km within the cloud. The correlation between CO and O3 within the modeled cloud is high, just as in S99 (0.65, cf. Fig. 4). Carbon monoxide is also well correlated with SO2 and Aitken mode aerosols, which indicates that all these variables originate from the polluted boundary layer.
b. Meteorological variables
The modeled minimum, average, and maximum temperature are in good agreement with observations (less than 3% difference, cf. Fig. 6a). For the ice water content and ice crystal number, the variation in their values occurs over several orders of magnitude (both in model and in observations); hence it is more appropriate to compare median and percentile values instead of average values (Figs. 6b,c). Note that for this comparison, all cloudy grid points at the 10.4-km level (instead of only along the cross section) are compared with observations in order to increase the number of sample points. The modeled ice water content and ice crystal number are, in general, in good agreement (within a factor of 2) with the observations.
c. Aerosols
A large number of small- (diameter >7 nm) and medium-sized (d > 18 nm) aerosols were observed at 10.4-km altitude by S99. The model predicts small- and medium-size aerosol concentrations up to 1.6 × 104 cm−3 and 8.3 × 103 cm−3, respectively at 10 km, which is 52% and 68% lower, respectively, than the observed values of 3.3 × 104 cm−3 and 2.6 × 104 cm−3 (Fig. 7). The model simulates the distribution of aerosol concentrations at 10.4 km fairly well (Fig. 7), but the variability is lower. There are several possible explanations for the discrepancy: the observations are performed along only one cross section whereas the model output is taken in all cloudy grid points; difference in location of the observed cross section and modeled surface within the cloud; difference in time point chosen for the comparison; difference in how the selection of a cloudy grid point is made; and the assumption of a lognormal distribution for the modeled aerosols.
S99 also performed measurements at the absolute top of the cloud, at 12.2-km altitude. High concentrations of aerosols, up to 20 000 cm−3, could be observed at this level. S99 suggested that these particles had been transported from the boundary layer up to the cloud anvil, but did not exclude the possibility that particle production might have occurred at the top of the cloud where evaporating crystals humidify the cold air. The model simulations also display high concentrations of Aitken mode aerosols at 12-km altitude (up to 14 × 103 cm−3, not shown). The Aitken mode aerosol concentration at this altitude is highly correlated with CO and SO2 concentrations (correlation coefficient equal to 0.98 and 0.78, respectively), again indicating transport of aerosols from the polluted boundary layer to the top of the cloud.
4. Aerosol transport within the convective cloud
Figure 8 shows the distribution of nucleation, Aitken, and BC mode aerosols for isosurfaces equal to 0.1, 6000, and 50 cm−3, respectively, after 3 hours of simulation. The vertical transport of nucleation, Aitken, and BC particles is also illustrated by the vertical cross sections shown in Figs. 9a,b,d (the location in the y direction of the cross section is the same as shown in Fig. 3, y = 120 km). Accumulation mode and mixed mode aerosols are efficient as CCN and are hence almost completely scavenged by precipitation, mainly through nucleation scavenging (cf. Fig. 9c and Ekman et al. 2004).
During the simulation, SO2 is oxidized to H2SO4 and hence a possibility for new particle formation is provided. As displayed by Figs. 8a and 9a, a small amount of nucleation mode aerosols are actually formed within the cloud. This type of formation was not observed in the 2D modeling study by Zhang et al. (1998). Nucleation mode aerosols are subjected to impaction scavenging by precipitation and may also be coagulated to form larger aerosols. However, some of the particles reach the top of the convective cloud. After 3 hours of simulation, the maximum concentration is 3.4 cm−3 in the cloud anvil. The transport of aerosols can also be studied by calculating the aerosol budget at different levels of the cloud as well as below the cloud. There is a small supply of nucleation mode aerosols both below and within the cloud during the simulation (Fig. 10a). Not investigated in this study, due to a lack of input data, is a potential nucleation mechanism via organic vapors. This type of nucleation process could generate hydrophobic particles that would have a better chance to survive the cloud processing.
Aitken mode aerosols are to some extent washed out by precipitation, but a substantial amount of this type of particles is transported toward the top of the convective cloud (Figs. 8b, 9b and 10b). These aerosols are not large enough to be effective CCN and not small enough to be efficiently scavenged through impaction scavenging (cf. Ekman et al. 2004). As the formation of new particles within the cloud is small, most Aitken mode particles observed at the top of the cloud are likely to originate from transport. This theory is also supported by Fig. 10b. After a 3-h simulation, the average concentration within the cloud anvil is approximately 10% of the surface concentration. This corresponds to an increase of the aerosol concentration in the upper parts of the cloud by 36% (from 2.4 × 103 to 3.3 × 103 cm−3). Zhang et al. (1998) noted a similar increase (∼1000 cm−3) of the small aerosol number concentration in their 2D simulations of a convective cloud. In the boundary layer, the average Aitken mode aerosol concentration is equal to 36% of the initial surface concentration. The removal of Aitken mode aerosols below the cloud occurs both through impaction scavenging by precipitating raindrops and through transfer to the accumulation mode aerosol size bin.
Accumulation mode and mixed mode aerosols are efficiently transported from the boundary layer to the cloud base and subsequently nucleation scavenged within the cloud. The average concentration of accumulation mode aerosols in the boundary layer at the end of the simulation is approximately 30% of the initial average surface concentration, and the average mixed mode concentration is about 50% (Figs. 10c,d). Within the anvil level, the average concentration of mixed mode aerosols is about 40% of the initial value in the upper troposphere, whereas the accumulation mode aerosol concentration is close to zero.
As the cloud develops, BC aerosols in the boundary layer are transported to the middle and high parts of the cloud by the convective updraft (Figs. 8b, 9d and 10e). Although BC aerosols are hydrophobic, they may still be subjected to impact scavenging by falling precipitation. At the end of the simulation, the average BC aerosol concentration in the boundary layer is 57% of the initial value. In the cloud anvil, the average concentration is 9% of the initial boundary layer concentration, an increase by 28% (from 24 to 31 cm−3) compared to the initial anvil concentration. The maximum concentration of BC aerosols at the cloud top is 105 cm−3, corresponding to a mass concentration of 20 ng m−3.
5. Aerosol processing in the free troposphere
After 3 hours of simulation using the CRM, the maximum SO2 and H2SO4 concentration at 10.4 km is 62 pptm (approximately 30% of the initial surface concentration) and 0.02 pptm, respectively. The average SO2 and H2SO4 concentrations within and below the cloud are displayed in Table 2. There is an overall increase of sulfur compounds in the free troposphere due to the convective activity. According to Wang and Prinn (2000), there are three factors contributing to the incomplete dissolving of SO2 into cloud water and thus the transport of SO2 to the cloud anvil: the much lower solubility of SO2 relative to H2O2, the conversion of water to ice phase particles that terminates the aqueous reactions, and the relatively short lifetime and limited coverage of the liquid phase portions of the cloud.
The peak in the Aitken mode aerosol concentration at the top of the convective cloud could be attributed both to direct transport from the lower troposphere as well as to newly formed aerosols in the upper troposphere. The latter process, if proved to be an important factor, could enable enhanced new aerosol formation to last longer than the convective core. Due to the limited model domain the current 3D simulation cannot fully cover this time period. To study the importance of particle nucleation in the free troposphere after the cloud’s dissipation, a box model has been designed. Using the box model, we also examine the possibility of the newly formed nucleation mode aerosols and transported Aitken mode aerosols to grow to suitable CCN or IN size (i.e., approximately the accumulation mode size bin of the model).
The box model includes aerosol microphysical and chemical processes such as H2SO4–H2O nucleation, coagulation, condensation of H2SO4 and oxidation of SO2 by OH to form H2SO4. Sulfur dioxide is assumed to be depleted as the oxidation to form H2SO4 occurs. Output from the CRM at t = 3 h and z = 10.4 km are used for initialization of the box model (Table 3; Fig. 11). Because the radiation model is not included in the box model, we need to make some assumptions about the OH concentration. Two model simulations, one with OH being depleted as SO2 is oxidized and one with constant OH concentration are thus introduced. These two simulations provide the lower and upper bound for SO2 oxidation. A box model simulation using a diurnally varying OH concentration has also been conducted for comparison with the constant OH concentration simulation. This simulation yielded similar results to the constant OH concentration simulation.
Figure 11a displays the results from the first run with OH being depleted. The 60-h time evolution of all sulfate aerosol modes as well as the H2SO4 concentration in the grid point of the box model with highest initial H2SO4 concentration is shown. Note that for a simulation time of this length the results should be interpreted with some caution, as there is no meteorological or radiative processes included in the box model (e.g., diffusion, cloud formation, etc.). The formation of nucleation mode aerosols is substantial; up to 1.0 × 104 cm−3 new particles are formed. In contrast to the CRM simulation, these aerosols cannot be impaction scavenged by precipitation. The number aerosols simulated is in agreement both with the model study by Zhang et al. (1998) and by several observations (Lee et al. 2004; Twohy et al. 2002; Clarke et al. 1999), which indicate new particle concentrations downwind of convective anvils in the range from 1.0 × 104 cm−3 to 5 × 104 cm−3.
Aitken mode aerosols do grow; mostly due to coagulation (as a result, the particle number concentration is reduced whereas the mass increases). However, in our simulation none of these aerosols can reach the prescribed accumulation size limit of the model (31 nm). The wind speed at 10.4-km altitude is on average 9 m s−1 in the CRM. This is in good agreement with radiosonde data reported from the same area (more information available online at http://www.weather.uwyo.edu/upperair). A 60-h simulation would correspond to a transport distance of 2000 km assuming a constant wind speed.
The main reason that the formation of new particles in the first box model run cease to exist with time is that the OH concentration rapidly approaches zero. Hence, there is no more production of H2SO4. When a constant OH concentration (= 6.1 × 106 molecules cm−3) is assumed, as in the second box model simulation (Fig. 11b), H2SO4 is continuously formed and there is thus a continuous formation of nucleation mode aerosols. As a result (and different from the box model simulation with OH being depleted), not only the mass within the Aitken mode increases, but also the number concentration as aerosols from the nucleation mode size bin grow and are transferred to the Aitken mode.
The maximum concentration of nucleation mode particles is 2.7 × 104 cm−3. When the aerosols within the Aitken mode reach a critical size (at ∼0.5 h), the condensation coefficient increases substantially. This increase results in that the H2SO4 concentration drops and that the formation of new particles becomes much slower. Aerosols within both the nucleation and Aitken mode grow through coagulation and condensation, but the growth rate of the Aitken mode aerosols is low. After 24 hours of simulation, or an approximate transport distance of 800 km, no aerosols have reached the accumulation mode size bin of the model. After 40 hours of simulation the formation of new particles is approximately zero. Aerosols within the nucleation mode are continuously transferred to the Aitken mode and after 46 hours of simulation (or 1500 km), there is some formation of accumulation mode aerosols. Note again, that this formation occurs assuming the conditions within the grid box remain the same during the whole simulation period.
The amount of available H2SO4 for nucleation is not only dependent on the oxidation rate of SO2, but also on the initial (and thus continuously simulated) Aitken mode aerosol concentration. If the initial Aitken mode number concentration is doubled, less H2SO4 is available for nucleation, and the maximum concentration of nucleation mode aerosols is reduced to 7.0 × 103 cm−3. At the same time, the growth of the Aitken mode aerosol particles becomes faster and accumulation mode aerosols are formed already after 24 hours of simulation. In an equivalent manner, if the initial Aitken mode aerosol concentration is halved, more nucleation mode aerosols are formed (maximum concentration ∼3.2 × 104 cm−3), the growth of the Aitken mode aerosols is slower and accumulation mode aerosols are formed later during the simulation.
6. Summary and conclusions
Several observational studies have indicated high concentrations of small aerosols in the vicinity of anvils of convective clouds. In the present study, we examine the possible origin of these aerosols using a 3D cloud-resolving model including an explicit aerosol module. We have first evaluated the performance of the model by comparing the model results with aircraft observations. Five different types of aerosols are considered: nucleation mode sulfate aerosols (here defined by 0 ≤ d ≤ 5.84 nm), Aitken mode sulfate aerosols (here defined by 5.84 nm ≤ d ≤ 31.0 nm), accumulation mode sulfate aerosols (here defined by d ≥ 31.0 nm), mixed aerosols, and black carbon aerosols.
The simulated average values and variability of meteorological and chemical variables are in general in good agreement with measurements. Polluted air is transported from the boundary layer during the convective event, and relatively high concentrations of chemical compounds such as CO and SO2 can be found at the top of the cloud. A plume of O3-rich air is transported from the surface to the cloud anvil, which is also noted in the observations.
Nucleation mode, accumulation mode, and mixed mode aerosols are efficiently scavenged by the heavy precipitation within the convective cloud through impaction scavenging (nucleation mode) and nucleation scavenging (accumulation and mixed mode) and only small numbers of these types of aerosols can reach the cloud anvil. However, and in agreement with the results obtained by Ekman et al. (2004) and Zhang et al. (1998), a substantial part of the Aitken mode and black carbon aerosol populations are transported to the upper troposphere. As the cloud reaches its decaying state, ∼10% of the initial surface concentrations of Aitken mode and black carbon aerosols are present at the top of the convective cloud. The average number of small aerosols simulated at 10.4-km altitude is of the same order of magnitude as in the observations. The results strongly suggest the critical role of vertical convective transport in the redistributions of sulfate aerosols in Aitken size range and for noncoated black carbon aerosols.
During the 3-h CRM simulation, new aerosols are formed within the convective cloud through binary H2O–H2SO4 nucleation, but the formation is rather small. At the end of the simulation a maximum concentration of approximately 3–4 cm−3 is noted in the cloud anvil. A different pathway to form more new particles within the cloud could be via organic vapors. Due to a lack of input data, this type of nucleation process has not been considered in the present study.
Using a box model including all the major microphysical and chemical processes of aerosols, additional simulations have been carried out to study the formation of new aerosols at the top of the cloud after the cloud has evaporated and the possibility of these aerosols to grow to suitable CCN or IN size. The concentration of SO2 in the anvil is relatively high and a substantial amount of new particles are formed (up to 1–3 × 104 cm−3). This number is in agreement both with observations downwind of convective clouds (Lee et al. 2004; Twohy et al. 2002; Clarke 1993) and with previous 2D simulations by Zhang et al. (1998). Nucleation and Aitken mode aerosols grow through coagulation and condensation, but the growth rate is low. Provided that there is enough OH available to oxidize the SO2, and that the ambient conditions within the box model remain unchanged during the simulation, some of the aerosols (concentrations ∼400 cm−3) can reach the accumulation mode size bin of the box model after 46 hours of simulation.
Altogether, the CRM and box model simulations suggest that the impact of convective transport on upper-tropospheric aerosol concentrations can last several days and extend over areas as wide as 1000 km2. During the active stage of convection direct transport through convective activity is responsible for injecting large number concentrations of small aerosols to the upper troposphere. In addition, after the cloud has dissipated, increased aerosol precursors in the upper troposphere can lead to formation of small aerosols in concentrations very close to that of the directly transported ones.
The fate of black carbon aerosols in the free troposphere is still quite unclear and so is the climate impact of these aerosols. There is a possibility that these aerosols could become “aged” with time and thus convert into large hygroscopic aerosols, interesting from a liquid cloud formation point of view. In addition, aged black carbon aerosols may be efficient as IN through heterogeneous freezing below water saturation. These issues need to be further investigated and a parameterization of heterogeneous freezing of aerosols (e.g., Khvorostyanov and Curry 2004) is planned to be included in the CRM.
Acknowledgments
The first author would like to thank the Knut and Alice Wallenberg foundation, Sweden, postdoctoral fellowship program on sustainability and the environment for research funding. This work was also partially supported by the NOAA Climate and Global Change Program Grant GC97-474, by NSF Grant ATM-0329759, by the Ford–MIT Alliance, and by the industrial consortium of the MIT Joint Program on the Science and Policy of Global Change.
REFERENCES
Charlson, R. J., S. E. Schwartz, J. M. Hales, R. D. Cess, J. A. Coakely Jr., J. E. Hansen, and D. J. Hofmann, 1992: Climate forcing by anthropogenic aerosols. Science, 255 , 423–430.
Clarke, A. D., 1992: Atmospheric nuclei in the remote free troposphere. J. Atmos. Chem., 14 , 479–488.
Clarke, A. D., 1993: Atmospheric nuclei in the Pacific midtroposphere: Their nature, concentration, and evolution. J. Geophys. Res., 98 , 20633–20647.
Clarke, A. D., and V. N. Kapustin, 2002: A Pacific aerosol survey. Part I: A decade of data on particle production, transport, evolution, and mixing in the troposphere. J. Atmos. Sci., 59 , 363–382.
Clarke, A. D., and Coauthors, 1999: Nucleation in the equatorial free troposphere: Favorable environments during PEM-Tropics. J. Geophys. Res., 104 , 5735–5744.
Clement, C. F., I. J. Ford, C. H. Twohy, A. Weinheimer, and T. Campos, 2002: Particle production in the outflow of a midlatitude storm. J. Geophys. Res., 107 .4559, doi:10.1029/2001JD001352.
Cziczo, D. J., D. M. Murphy, P. K. Hudson, and D. S. Thomson, 2004: Single particle measurements of the chemical composition of cirrus ice residue during CRYSTAL-FACE. J. Geophys. Res., 109 .4201, doi:10.1029/2003JD004032.
DeMott, P. J., D. J. Cziczo, A. J. Prenni, D. M. Murphy, S. M. Kreidenweis, D. S. Thomson, R. Borys, and D. C. Rogers, 2003: Measurements of the concentration and composition of nuclei for cirrus formation. Proc. Natl. Acad. Sci. U.S.A., 100 , 14655–14660.
DeReus, M., J. Ström, M. Kulmala, L. Prijola, J. Lelieveld, C. Schiller, and M. Zöger, 1998: Airborne aerosol measurements in the tropopause region and the dependence of new particle formation on preexisting particle number concentration. J. Geophys. Res., 103 , 31255–31263.
Ekman, A., C. Wang, J. Wilson, and J. Ström, 2004: Explicit simulations of aerosol physics in a cloud-resolving model: A sensitivity study based on an observed convective cloud. Atmos. Chem. Phys., 4 , 773–791.
Fu, Q., and K. N. Liou, 1993: Parameterization of the radiative properties of cirrus clouds. J. Atmos. Sci., 50 , 2008–2025.
Fuchs, N. A., 1964: The Mechanics of Aerosols. Pergamon Press, 408 pp.
Haag, W., and B. Kärcher, 2004: The impact of aerosols and gravity waves on cirrus clouds at midlatitudes. J. Geophys. Res., 109 .D12202, doi:10.1029/2004JD004579.
Haag, W., B. Kärcher, J. Ström, A. Minikin, U. Lohmann, J. Ovarlez, and A. Stohl, 2003: Freezing thresholds and cirrus cloud formation mechanisms inferred from in situ measurements of relative humidity. Atmos. Chem. Phys., 3 , 1791–1806.
Hermann, M., J. Heintzenberg, A. Wiedensohler, A. Zahn, G. Heinrich, and C. A. M. Brenninkmeijer, 2003: Meridional distributions of aerosol particle number concentrations in the upper troposphere and lower stratosphere obtained by Civil Aircraft for Regular Investigation of the Atmosphere Based on an Instrument Container (CARIBIC) flights. J. Geophys. Res., 108 .4114, doi:10.1029/2001JD001077.
Hjellbrekke, A. G., and J. E. Hanssen, 1998: Data report 1996. Part 2. Monthly and seasonal summaries. NILU/CCC Rep. 2/98, The Norwegian Institute for Air Research, Lillestrom, Norway, 229 pp.
Kärcher, B., and U. Lohmann, 2003: A parameterization of cirrus cloud formation: Heterogeneous freezing. J. Geophys. Res., 108 .4402, doi:10.1029/2002JD003220.
Khvorostyanov, V. I., and J. A. Curry, 2004: The theory of ice nucleation by heterogeneous freezing of deliquescent mixed CCN. Part I: Critical radius, energy, and nucleation rate. J. Atmos. Sci., 61 , 2676–2691.
Kulmala, M., A. Laaksonen, and L. Pirjola, 1998: Parameterizations for sulphuric acid/water nucleation rates. J. Geophys. Res., 103 , 8301–8307.
Lawrence, M. G., and Coauthors, 2003: Global chemical weather forecasts for field campaign planning: Predictions and observations of large-scale features during MINOS, CONTRACE and INDOEX. Atmos. Chem. Phys., 2 , 267–289.
Lee, S-H., and Coauthors, 2004: New particle production observed in the tropical/subtropical cirrus clouds. J. Geophys. Res., 109 .D20209, doi:10.1029/2004JD005033.
Minikin, A., A. Petzold, J. Ström, R. Krejci, M. Seifert, P. van Velthoven, H. Schlager, and U. Schumann, 2003: Aircraft observations of the upper tropospheric fine particle aerosol in the Northern and Southern Hemispheres at midlatitudes. Geophys. Res. Lett., 30 .1503, doi:10.1029/2002GL016458.
Nyeki, S., M. Kalberer, M. Lugauer, E. Weingartner, A. Petzold, F. Schroder, I. Colbeck, and U. Baltensperger, 1999: Condensation nuclei (CN) and ultrafine CN in the free troposphere to 12 km: A case study over the Jungfraujoch high-alpine research station. Geophys. Res. Lett., 26 , 2195–2198.
Petzold, A., and Coauthors, 2002: Vertical variability of aerosol properties observed at a continental site during the Lindenberg Aerosol Characterization Experiment (LACE 98). J. Geophys. Res., 107 .8128, doi:10.1029/2001JD001043.
Pickering, K. E., and Coauthors, 2001: Trace gas transport and scavenging in PEM-Tropics B South Pacific convergence zone convection. J. Geophys. Res., 106 , 32591–32607.
Ridley, B., and Coauthors, 2004: Convective transport of reactive constituents to the tropical and rigid-latitude tropopause region: I. Observations. Atmos. Environ., 38 , 1259–1274.
Schröder, F., B. Kärcher, M. Fiebig, and A. Petzold, 2002: Aerosol states in the free troposphere at northern latitudes. J. Geophys. Res., 107 .8126, doi:10.1029/2000JD000194.
Seinfeld, J. H., and S. N. Pandis, 1998: Atmospheric Chemistry and Physics: From Air Pollution to Climate Change. John Wiley & Sons, 1326 pp.
Ström, J., H. Fischer, J. Lelieveld, and F. Schröder, 1999: In situ measurements of microphysical properties and trace gases in two cumulonimbus anvils over western Europe. J. Geophys. Res., 104 , 12221–12226.
Twohy, C. H., and Coauthors, 2002: Deep convection as a source of new particles in the midlatitude upper troposphere. J. Geophys. Res., 107 .4560, doi:10.1029/2001JD000323.
Twomey, S., 1974: Pollution and the planetary albedo. Atmos. Environ., 8 , 1251–1256.
Vehkamäki, H., M. Kulmala, I. Napari, K. E. J. Lehtinen, C. Timmreck, M. Noppel, and A. Laaksonen, 2002: An improved parameterization for sulfuric acid–water nucleation rates for tropospheric and stratospheric conditions. J. Geophys. Res., 107 .4622, doi:10.1029/2002JD002184.
Wang, C., 2005: A modeling study of the response of tropical deep convection to the increase of CCN concentration: 1. Dynamics and microphysics. J. Geophys. Res., 110 .D21211, doi:10.1029/2004JD005720.
Wang, C., and J. S. Chang, 1993a: A three-dimensional numerical model of cloud dynamics, microphysics, and chemistry. 1: Concepts and formulation. J. Geophys. Res., 98 , 14827–14844.
Wang, C., and J. S. Chang, 1993b: A three-dimensional numerical model of cloud dynamics, microphysics, and chemistry. 4: Cloud chemistry and precipitation chemistry. J. Geophys. Res., 98 , 16799–16808.
Wang, C., and P. J. Crutzen, 1995: Impact of a simulated severe local storm on the redistribution of sulfur dioxide. J. Geophys. Res., 100 , 11357–11367.
Wang, C., and R. G. Prinn, 1998: Impact of the horizontal wind profile on the convective transport of chemical species. J. Geophys. Res., 103 , 22063–22071.
Wang, C., and R. G. Prinn, 2000: On the roles of deep convective clouds in tropospheric chemistry. J. Geophys. Res., 105 , 22269–22297.
Wang, C., P. J. Crutzen, V. Ramanathan, and S. F. Williams, 1995: The role of a deep convective storm over the tropical Pacific Ocean in the redistribution of atmospheric chemical species. J. Geophys. Res., 100 , 11509–11516.
Wilson, J., C. Cuvelier, and F. Raes, 2001: A modeling study of global mixed aerosol fields. J. Geophys. Res., 106 , 34081–34108.
Zhang, Y. P., S. Kreidenweis, and G. R. Taylor, 1998: The effects of clouds on aerosol and chemical species production and distribution. Part III: Aerosol model description and sensitivity analysis. J. Atmos. Sci., 55 , 921–939.
Schematic picture of processes included in the aerosol model (following work by Wilson et al. 2001).
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Schematic picture of processes included in the aerosol model (following work by Wilson et al. 2001).
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Schematic picture of processes included in the aerosol model (following work by Wilson et al. 2001).
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Vertical profiles of initial aerosol concentrations.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Vertical profiles of initial aerosol concentrations.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Vertical profiles of initial aerosol concentrations.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Isosurface for the mass mixing ratio of total condensed water = 0.01 g kg−1 after a 3-h simulation. The full line, aligned with the direction of the spreading of the anvil and close to the center of the convective core, indicates the location of the cross sections displayed in Figs. 4 and 9.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Isosurface for the mass mixing ratio of total condensed water = 0.01 g kg−1 after a 3-h simulation. The full line, aligned with the direction of the spreading of the anvil and close to the center of the convective core, indicates the location of the cross sections displayed in Figs. 4 and 9.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Isosurface for the mass mixing ratio of total condensed water = 0.01 g kg−1 after a 3-h simulation. The full line, aligned with the direction of the spreading of the anvil and close to the center of the convective core, indicates the location of the cross sections displayed in Figs. 4 and 9.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Simulated variables at t = 3 h along a cross section within convective cloud (y = 120 km, z = 10.4 km) at 10.4-km altitude: (top) ice hydrometeor number (light gray line) and ice water content (black line), (middle) carbon monoxide mixing ratio (light gray line) and O3 mixing ratio (black line), and (bottom) Aitken mode aerosol concentration (light gray line) and temperature (black line) at standard pressure and temperature (STP).
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Simulated variables at t = 3 h along a cross section within convective cloud (y = 120 km, z = 10.4 km) at 10.4-km altitude: (top) ice hydrometeor number (light gray line) and ice water content (black line), (middle) carbon monoxide mixing ratio (light gray line) and O3 mixing ratio (black line), and (bottom) Aitken mode aerosol concentration (light gray line) and temperature (black line) at standard pressure and temperature (STP).
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Simulated variables at t = 3 h along a cross section within convective cloud (y = 120 km, z = 10.4 km) at 10.4-km altitude: (top) ice hydrometeor number (light gray line) and ice water content (black line), (middle) carbon monoxide mixing ratio (light gray line) and O3 mixing ratio (black line), and (bottom) Aitken mode aerosol concentration (light gray line) and temperature (black line) at standard pressure and temperature (STP).
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Modeled and observed minimum, average, and maximum (a) CO and (b) O3 concentrations along the cross section of convective clouds at 10.4-km altitude. Sample numbers are indicated within brackets.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Modeled and observed minimum, average, and maximum (a) CO and (b) O3 concentrations along the cross section of convective clouds at 10.4-km altitude. Sample numbers are indicated within brackets.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Modeled and observed minimum, average, and maximum (a) CO and (b) O3 concentrations along the cross section of convective clouds at 10.4-km altitude. Sample numbers are indicated within brackets.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Modeled and observed (a) temperature along the cross section of convective clouds, (b) ice water content in all cloudy grid points, and (c) ice hydrometeor concentration in all cloudy grid points at 10.4-km altitude. STP values shown are in (a) minimum, mean, and maximum; (b) median, 75% tile, 90% tile, and maximum; (c) 10% tile, 25% tile, median, 75% tile, 90% tile, and maximum. Sample numbers are indicated within brackets.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Modeled and observed (a) temperature along the cross section of convective clouds, (b) ice water content in all cloudy grid points, and (c) ice hydrometeor concentration in all cloudy grid points at 10.4-km altitude. STP values shown are in (a) minimum, mean, and maximum; (b) median, 75% tile, 90% tile, and maximum; (c) 10% tile, 25% tile, median, 75% tile, 90% tile, and maximum. Sample numbers are indicated within brackets.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Modeled and observed (a) temperature along the cross section of convective clouds, (b) ice water content in all cloudy grid points, and (c) ice hydrometeor concentration in all cloudy grid points at 10.4-km altitude. STP values shown are in (a) minimum, mean, and maximum; (b) median, 75% tile, 90% tile, and maximum; (c) 10% tile, 25% tile, median, 75% tile, 90% tile, and maximum. Sample numbers are indicated within brackets.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Modeled and observed number of aerosols (a) (d > 7 nm) and (b) (d > 18nm) at 10.4-km altitude in all cloudy grid points. STP values shown are minimum, 10% tile, 25% tile, median, 75% tile, 90% tile, and maximum. Sample numbers are indicated within brackets.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Modeled and observed number of aerosols (a) (d > 7 nm) and (b) (d > 18nm) at 10.4-km altitude in all cloudy grid points. STP values shown are minimum, 10% tile, 25% tile, median, 75% tile, 90% tile, and maximum. Sample numbers are indicated within brackets.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Modeled and observed number of aerosols (a) (d > 7 nm) and (b) (d > 18nm) at 10.4-km altitude in all cloudy grid points. STP values shown are minimum, 10% tile, 25% tile, median, 75% tile, 90% tile, and maximum. Sample numbers are indicated within brackets.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
(a) Modeled isosurfaces for nucleation mode aerosol concentration equal to 0.1 cm−3 (dark gray) and Aitken mode aerosol concentration equal to 6000 cm−3 (light gray) after 3-h simulation. (b) Modeled isosurface for a black carbon aerosol concentration equal to 50 cm−3.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
(a) Modeled isosurfaces for nucleation mode aerosol concentration equal to 0.1 cm−3 (dark gray) and Aitken mode aerosol concentration equal to 6000 cm−3 (light gray) after 3-h simulation. (b) Modeled isosurface for a black carbon aerosol concentration equal to 50 cm−3.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
(a) Modeled isosurfaces for nucleation mode aerosol concentration equal to 0.1 cm−3 (dark gray) and Aitken mode aerosol concentration equal to 6000 cm−3 (light gray) after 3-h simulation. (b) Modeled isosurface for a black carbon aerosol concentration equal to 50 cm−3.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
(a) Modeled cross section through cloud (along line displayed in Fig. 1) of nucleation mode aerosols. Note that concentrations have been recalculated to STP. (b) As in (a) but for Aitken mode aerosols. (c) As in (a) but for accumulation mode aerosols. (d) As in (a) but for black carbon aerosols.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
(a) Modeled cross section through cloud (along line displayed in Fig. 1) of nucleation mode aerosols. Note that concentrations have been recalculated to STP. (b) As in (a) but for Aitken mode aerosols. (c) As in (a) but for accumulation mode aerosols. (d) As in (a) but for black carbon aerosols.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
(a) Modeled cross section through cloud (along line displayed in Fig. 1) of nucleation mode aerosols. Note that concentrations have been recalculated to STP. (b) As in (a) but for Aitken mode aerosols. (c) As in (a) but for accumulation mode aerosols. (d) As in (a) but for black carbon aerosols.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
(a) Time evolution of average nucleation mode aerosol concentration at different levels below and within the convective cloud (full lines). Below the cloud, all grid points within the square (x, y) = (100–140 km, 100–140 km) are considered. Within the cloud, only grid points with CWC > 0.01 g kg−1 are considered. The initial average concentration for the grid box below the cloud is also shown for comparison (dashed lines). (b) As in (a) but for Aitken mode aerosols. (c) As in (a) but for accumulation mode aerosols. (d) As in (a) but for BC mode aerosols. (e) As in (a) but for mixed mode aerosols.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
(a) Time evolution of average nucleation mode aerosol concentration at different levels below and within the convective cloud (full lines). Below the cloud, all grid points within the square (x, y) = (100–140 km, 100–140 km) are considered. Within the cloud, only grid points with CWC > 0.01 g kg−1 are considered. The initial average concentration for the grid box below the cloud is also shown for comparison (dashed lines). (b) As in (a) but for Aitken mode aerosols. (c) As in (a) but for accumulation mode aerosols. (d) As in (a) but for BC mode aerosols. (e) As in (a) but for mixed mode aerosols.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
(a) Time evolution of average nucleation mode aerosol concentration at different levels below and within the convective cloud (full lines). Below the cloud, all grid points within the square (x, y) = (100–140 km, 100–140 km) are considered. Within the cloud, only grid points with CWC > 0.01 g kg−1 are considered. The initial average concentration for the grid box below the cloud is also shown for comparison (dashed lines). (b) As in (a) but for Aitken mode aerosols. (c) As in (a) but for accumulation mode aerosols. (d) As in (a) but for BC mode aerosols. (e) As in (a) but for mixed mode aerosols.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Time development of aerosol number concentration and H2SO4 concentration at grid point x = 114 km, y = 180 km; 10.4 km calculated using the box model (a) with and (b) without OH depletion. The discontinuities seen in the figures are a result of the “remapping” procedure of aerosols between the different size bins.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Time development of aerosol number concentration and H2SO4 concentration at grid point x = 114 km, y = 180 km; 10.4 km calculated using the box model (a) with and (b) without OH depletion. The discontinuities seen in the figures are a result of the “remapping” procedure of aerosols between the different size bins.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
Time development of aerosol number concentration and H2SO4 concentration at grid point x = 114 km, y = 180 km; 10.4 km calculated using the box model (a) with and (b) without OH depletion. The discontinuities seen in the figures are a result of the “remapping” procedure of aerosols between the different size bins.
Citation: Journal of the Atmospheric Sciences 63, 2; 10.1175/JAS3645.1
General features of the modeled convective cloud.
Average initial (init) and after 4 hours of simulation (end) concentrations of trace gases below the cloud (0–3.2 km), and in low (3.2–5.2 km), middle (5.2–8 km), and high parts (8–14 km) of the convective cloud.
Initial values applied for the box model simulations.
