What controls cloud droplet number concentration of trade wind cumuli?

Aerosol and cloud microphysical properties of trade wind cumuli were determined during the CARRIBA (Clouds, Aerosol, Radiation, and tuRbulence in the trade wInd regime over BArbados, Siebert et al., 2012) experiment in November 2010 and April 2011. Measurements were performed with the helicopter-borne measurement platform ACTOS (Airborne Cloud Turbulence Observation System, Siebert et al., 2006) with high temporal and spatial resolution.

ACTOS is equipped with sensors to measure basic meteorological parameters (3d wind velocity, temperature, humidity, etc.), turbulence parameters as well as aerosol and cloud microphysical properties.

Aerosol particle number size distributions were measured in the size range of 6 nm < Dp < 2.5 μm (combination of Scanning Mobility Particle Sizer and Optical Particle Counter). Additionally total particle number concentration (N) and cloud condensation nucleus concentration at different supersaturations (N_CCN) were measured with a Condensational Particle Counter (CPC) and a miniaturized Cloud Condensation Nucleus counter (mCCNc).

Cloud droplet characteristics and liquid water content were determined using a Phase Doppler Interferometer (PDI) and the Particle Volume Monitor (PVM). Detailed information on the instrumentation can be found in Siebert et al., 2012 and Ditas et al., 2012.

This study deals with the statistical analysis of aerosol-cloud-interactions of almost 700 individual clouds measured during 10 research flights under comparable meteorological conditions. The analysis provides information on the number concentration of activated particles (N_act), activation diameter (D_{p, act}) and critical supersaturation (S_crit). The calculation of N_act is based on the comparison of total particle number concentration outside of clouds and the interstitial particle number concentration inside clouds. This method is validated against cloud droplet number concentration data (N_d). Activation diameter and critical supersaturation are derived using the measured particle number size distribution and Köhler theory. Furthermore, major factors controlling the cloud droplet number concentration are identified with the help of the helicopter-borne observations and a comprehensive sensitivity study using a cloud microphysical parcel model.

The observed clouds were sorted for evolution stages. Focusing on the 30% most active clouds, the influence of vertical wind velocity (w) inside the cloud and the CCN concentration below clouds were examined.

Fig. 1: Boxplot of N_act versus N_CCN at 0.26% supersaturation. Median values are illustrated with thick horizontal lines, the box (error bars) show 50% (75%) of the data.

Figure 1 shows a boxplot of N_act versus N_CCN at 0.26% supersaturation. In spite of the scatter, median N_act are strongly influenced by the CCN concentration, even stronger than by w (not shown here).

Motivated by the observations we performed a sensitivity study using the spectral cloud microphysical parcel model by Simmel & Wurzler, 2006. Within this study the sensitivity of the cloud droplet number concentration towards changes in the particle number size distribution, total particle number concentration, particle hygroscopicity (κ) and cloud base updraft is investigated within the range of our observations during the CARRIBA campaign.

Fig. 2: Contour plot of modeled cloud droplet number concentration N_d as a function of updraft velocity w and total particle number concentration N. γ denotes the ratio between Aitken and accumulation mode particle number concentration.

Figure 2 shows model results of calculated N_d for different updraft velocities and total particle number concentration. The initial particle number size distribution is represented by two lognormal modes with equal number concentration, equal standard deviation (σ = 1.45), and mean diameters at D_p1 = 35 nm and D_p2 =140 nm. The orientation of the contour lines in Fig. 2 is more or less parallel to the ordinate, indicating that changes in Nw. Hence, in the plotted example the cloud droplet number concentration is limited by the amount of available aerosols.

Finally, we calculated the sensitivity of N_d towards all varied parameters (N, κ, w, shape of particle number size distribution γ). Our results show that the cloud droplet number concentration of trade wind cumuli is very sensitive to changes in the total particle number concentration and the particle number size distribution.

We thank the Caribbean Institute for Meteorology and Hydrology, Horizon Helicopters, enviscope GmbH, and Barbados Concorde Experience. This project was funded by DFG-grant SI 1534/3-1 and the TROPOS.