Ice Concentration of Tropical Tropopause Layer Cirrus Under the Existence of Bimodal Aerosol Distribution

Cirrus clouds in the Tropical Tropopause Layer (TTL) plays an important role in regulating the amount of water entering the stratosphere (Holton and Gettelman 2001). Intensive discussions about TTL cirrus have been made by focusing on the comparison between ice concentrations measured by aircraft observations and those simulated by cloud microphysical models (e.g., Jensen and Pfister 2004; Jensen et al. 2010, 2012; Spichtinger and Krämer 2013). However, the detailed cloud physical processes in the TTL are not fully understood. Mimura et al. (2016) found that the cooling rate of a few Kelvin per hour is required for the formation of thin (10¹ m) high ice concentration (10⁴ L^-1) TTL cirrus observed by Airborne Tropical Tropopause Experiment (ATTREX) 2011 if monodisperse aerosols are assumed. From a series of model simulations assuming isentropic ascent with constant cooling rate, following features are found for the maximum ice concentration (N_i).

1. N_i increases as the cooling rate of air parcel (–dT/dt) increases,
2. N_i decreases as initial water vapor mixing ratio (X₀) increases,
3. N_i decreases as accommodation coefficient (α_d) increases, and
4. N_i decreases as radius of aerosol particles (r) increases, especially when r is greater than ~0.5 μm.

                The fourth feature above is explained by the improved water uptake efficiency by aerosols with large nuclei (thus large ice particles) and suppression of new ice particle formation (hereafter referred to as “effect 1”). Another effect, that is, the increase of ice nucleation rate due to the increase of aerosol volume (hereafter referred to as “effect 2”), need to be considered. Due to this effect2, larger aerosols could nucleate under lower RH_i condition.

                The present paper discusses the behavior of N_i under the existence of bimodal aerosol distribution paying special attention to these two contrasting effects. The experimental condition is set to –dT/dt = 2.8 K h⁻¹, X₀= 6 ppmv, and α_d =0.3, that lead to N_i = 10⁴ L⁻¹ for monodisperse aerosol radius of 0.05 μm. In order to focus on the interaction between bimodal aerosols with different radius (r_a, r_b), the number concentrations of two aerosol types are set to the same (10⁵ L^-1). As the first step of our analysis, we fix r_a = 0.05 μm and sweep r_b from 0.01 to 10 μm, and discuss about the result by focusing on N_i and number concentrations of ice nucleated from two size of aerosols (N_ia, N_ib).

                It is readily seen that N_ia decreases as r_b increases because of the effect 2. On the other hand, N_ib takes maximum at r_b = 0.084 μm. When r_b is much smaller than r_a, aerosols with radius r_a nucleate at lower RH_i, suppressing the nucleation of aerosols with radius r_b. On the other hand, when r_b is larger than r_a, the proportion of N_ib in N_i becomes larger due to effect 2. However, when r_b is larger than 0.084 μm, the suppression of N_ib by effect 1 along with increase of r_b exceeds the promotion of N_ib by effect2. Another point of special interest is that N_i takes maximum at r_b = 0.035 μm exceeding that at r_a = r_b = 0.05 μm. This may be the result of effect2 of the aerosols r_b leading to the decrease in N_ia and increase in N_ib. The results from the case of different r_a will be also discussed at the time of presentation.