TROPOSPHERIC AEROSOL FORMATION: PROCESSES, OBSERVATIONS AND SIMULATIONS

摘要

Observations interpreted using new modeling techniques suggest that under typical atmospheric conditions, the formation of new particles is likely to be enhanced by the presence of natural ionization. This ion-mediated nucleation (IMN) process may be important in environments where nanometer-sized particles are otherwise marginally stable, and homogeneous nucleation is highly improbable. Owing to the presence of charge, IMN is capable of explaining measurements of enhanced particle growth rates (by a factor as large as 10, linked to species such as H_2SO_4, H_2O and NH_3) up to measurable sizes ～3 nm. The discussion presented here demonstrates that IMN theory is consistent with a range of field data describing the behavior of ultrafine aerosols. This mechanism also explains the detection of new particles under conditions for which homogeneous nucleation is seemingly precluded. It follows that an ion-based theory is likely to improve our ability to model particle formation under atmospheric conditions. The number of nucleated particles that contributes to the background population of cloud condensation nuclei (CCN) depends on the ionization rate, initial ion abundance, and precursor vapor concentrations. The influence of the ionization rate on CCN production may have a bearing on the correlation noted between variations in global cloud coverage (up to 3-4%) and galactic cosmic ray fluxes (up to 20% over the 11-year solar cycle). Air ionization rates are also affected by surface radioactivity over land, and excess free charge created by lighting activity and corona discharge within clouds. Even bursting bubbles and droplet impacts on surfaces are known to generate electrically charged micro-particles. These additional mechanisms might also initiate IMN, leading to local increases in the abundances of ultrafine aerosols and CCN. The aerosol nucleation rates associated with IMN are normally limited by the background ionization rate, which varies roughly from ～1-30 ion-pairs/cm~3s in the troposphere. This property of IMN is consistent with the average maximum particle production rates of ≤0.5-10/cm~3s estimated from a number of in situ observations. However, at higher precursor vapor supersaturations, direct activation of pre-existing ions could produce bursts of aerosols numbering more than 100,000/cm~3. Thus, the studies cited above point to three distinct phases of aerosol nucleation, corresponding to an increasing state of supersaturation of air: Phase 1. Low precursor supersaturation: leads to "slow" nucleation at average rates ≤0.1-10~(-2)/cm~3s, connected with the formation of stable neutral clusters following the recombination of large ambient ions; generates ultrafine aerosol densities of up to 100's-1000's/cm~3. Phase 2. Intermediate precursor supersaturation: causes selective ion activation and growth in an environment of charged and neutral molecular clusters; generates ultrafine particle concentrations of 1000's-10,000' s/cm~3. Phase 3. High precursor supersaturation: triggers homogeneous nucleation, either binary or ternary, possibly in regions of high humidity; creates ultrafine particle abundances reaching 100,000's/cm~3. Measurements of the air mobility spectrum seem to add considerable information toward an understanding of aerosol formation and growth at sizes below a few nanometers. Hence, to characterize nucleation mechanisms more precisely, such data should be included in experimental designs. Lagrangian aerosol sampling techniques would also be favored, since this approach can yield data on microphysical evolution without the complicating effects of a changing air mass. Further laboratory studies should be undertaken to quantify the thermodynamic data that define ion properties under tropospheric conditions, at ion sizes and compositions relevant to aerosol nucleation. The sparseness of such data imposes a limitation on our ability to quantify ion-based nucleation mechanisms. The authors acknowledge support f