The microphysical properties of corona-producing ice clouds observed in a cloud chamber experiment

The recent studies of natural corona-producing ice clouds have shown that, in contrary to earlier believes, the colorful corona displays in cirrus clouds are produced by untypically small and unusually shaped ice crystals with diameter of 14-25 μm (e.g. Sassen 1991; Sassen et al. 1998; Shaw and Neiman 2003; Shaw and Pust 2011). Up to now, the knowledge of cloud microphysical conditions related to the nucleation of these cirrus clouds, is limited to photographic, lidar and meteorological studies. Homogeneous freezing of sulphuric acid droplets of stratospheric origin is suggested to be the main ice nucleation pathway of these ice crystals (Sassen et al. 1998), but no direct measurements of the nucleation pathway or chemistry of the ice crystals have been made. In our study we explain the microphysics of these type cirrus clouds and present the first measurements from corona-producing clouds produced in a cloud chamber.

The experimental results presented here were conducted at the AIDA (Aerosol Interaction and Dynamics in the Atmosphere) cloud chamber (i.e. Möhler et al. (2003)). The forward scattering light from the nucleating ice cloud was studied with a laser scattering and depolarization instrument (SIMONE, Schnaiter et al. (2012)) and simultaneously the size distribution and shape of the ice crystals were recorded with an optical particle counters (PPD, Welas). The conditions in the cloud chamber were logged constantly. Our goal was to create a corona-producing cloud at the same conditions described with the natural corona display cases, i.e. create a narrow growing ice crystal size distribution. We used as an ice nuclei hematite, that we know causes fast ice nucleation, and sulphuric acid droplets, like suggested in previous study. As a starting temperature for the expansion, we used -40 °C.

In three of the conducted experiment runs, we were able to observe oscillation pattern in the forward scattering measured at the angle of 1.8°. We related the maxima and minima of the observed oscillation pattern to a size simultaneously measured with PPD or with Welas. The maxima were correlated to a size of 19 μm and the minima to a size of 28 μm according to the PPD. The measured sizes were 6 μm smaller than the size responsible of creating 2nd order maxima and 2nd order minima calculated with the Fraunhofer diffraction theory for the same wavelength and angular radius. The difference can be explained with inaccuracy in the PPD sizing and with the uniformity in the ice crystal shape.

To further confirm the observation and eliminate the possibility of measurement artifact, we used the measured size distribution and Mie-calculations to simulate the observed oscillation in the forward scattering spectra. The observed oscillation pattern was reproducible using a lognormal size distribution with mean diameter and sigma retrieved from the PPD measurements.

With our cloud chamber experiments we have successfully managed to nucleate a corona-producing ice crystal cloud. To the complement of earlier measurements of natural corona-producing cirrus clouds, we showed that these types of cirrus clouds are composed of small ice crystals created in cold temperatures trough rapid ice nucleation. We were also able to show that homogeneous nucleation on sulphuric acid droplets can cause these types of cirrus clouds, as speculated in earlier study. This new type of cirrus cloud differs significantly from the cirrus clouds used in climate models. In order to improve the our understanding of the effect of Earth's cirrus cloud cover to the radiative forcing, the prevalence as well as the radiative properties of this cirrus cloud type need to be studied more.