Particle number concentrations and size distributions in the stratosphere: implications of nucleation mechanisms and particle microphysics

Abstract

While formation and growth of particles in the troposphere have been extensively studied in the past two decades, very limited efforts have been devoted to understanding these in the stratosphere. Here we use both Cosmics Leaving OUtdoor Droplets (CLOUD) laboratory measurements taken under very low temperatures (205–223 K) and Atmospheric Tomography Mission (ATom) in situ observations of particle number size distributions (PNSDs) down to 3 nm to constrain nucleation mechanisms and to evaluate model-simulated particle size distributions in the lowermost stratosphere (LMS). We show that the binary homogenous nucleation (BHN) scheme used in most of the existing stratospheric aerosol injection (a proposed method of solar radiation modification) modeling studies overpredicts the nucleation rates by 3–4 orders of magnitude (when compared to CLOUD data) and particle number concentrations in the background LMS by a factor ∼ 2–4 (when compared to ATom data). Based on a recently developed kinetic nucleation model, which gives rates of both ion-mediated nucleation (IMN) and BHN at low temperatures in good agreement with CLOUD measurements, both BHN and IMN occur in the stratosphere. However, IMN rates are generally more than 1 order of magnitude higher than BHN rates and thus dominate nucleation in the background stratosphere. In the Southern Hemisphere (SH) LMS with minimum influence of anthropogenic emissions, our analysis shows that ATom-measured PNSDs generally have four apparent modes. The model captures reasonably well the two modes (Aitken mode and the first accumulation mode) with the highest number concentrations and size-dependent standard deviations. However, the model misses an apparent second accumulation mode peaking around 300–400 nm, which is in the size range important for aerosol direct radiative forcing. The bimodal structure of accumulation mode particles has also been observed in the stratosphere well above tropopause and in the volcano-perturbed stratosphere. We suggest that this bimodal structure may be caused by the effect of charges on coagulation and growth, which is not yet considered in any existing models and may be important in the stratosphere due to high ionization rates and the long lifetime of aerosols. Considering the importance of accurate PNSDs for projecting a realistic radiation forcing response to stratospheric aerosol injection (SAI), it is essential to understand and incorporate such potentially important processes in SAI model simulations and to carry out further research to find out what other processes the present models might have missed.