How can we make use of the observed variability of ice nucleating particle concentrations in models?

The aerosols that can act as ice nucleating particles (INPs) are of a large variety in both size and chemical composition. This leads to different temperatures at which INPs can nucleate ice. In recent years, the type and properties of INP species that initiate freezing at certain conditions have been investigated in more and more detail. If this growing knowledge is to be used in models, it implies that the models hold very detailed information about the aerosol species abundant in clouds. Especially for climate models, this is a difficult challenge.

In this study, we approach the problem of parameterizing heterogeneous ice nucleation from another angle: assuming well-diluted background aerosol, the probability of a specific INP concentration at the current temperature follows a log-normal distribution. We derived relative frequency distribution functions (RFDs) from measurements in marine environments (Welti et al., 2018). The number of INPs for the current temperature is being drawn from this RFD following its weighting. Thus, one randomly selected INP concentration is used from the range of all possible INP concentrations derived from observations at the given temperature. The advantage of our freezing parameterization is that it does not need any information about the chemical composition or size of the aerosols present in the cloud. It is valid for remote locations that are not close to a source of INPs, e.g. maritime or polar sites.

We implemented this new parameterization into the large-eddy simulation model MIMICA (Savre et al., 2014) and evaluated it for a mixed-phase Arctic cloud observed during the ASCOS expedition (Tjernström et al., 2014). For the Arctic there is large uncertainty about the types of INPs as well as their concentration, which is a challenge for modelling mixed phase clouds in this region. In our talk, we show that our new scheme does work well for the simulated case. We will present the performance of this new framework, as well as its sensitivity to RFD distribution variables and the model domain resolution. Additionally, we compare the new parameterization to “classic” heterogeneous nucleation schemes, such as a simple active sites parameterization.

Savre, J., Ekman, A. M. L., and Svensson, G. (2014), Technical note: Introduction to MIMICA, a large-eddy simulation solver for cloudy planetary boundary layers, J. Adv. Model. Earth Syst., 6, 630–649, doi:10.1002/ 2013MS000292.
Tjernström, M.; Leck, C., et al. (2014), The Arctic Summer Cloud Ocean Study (ASCOS): overview and experimental design, Atmos. Chem. Phys., 14, 2823–2869.
Welti, A., Müller, K., Fleming, Z. L., and Stratmann, F. (2018), Concentration and variability of ice nuclei in the subtropical maritime boundary layer, Atmos. Chem. Phys., 18, 5307–5320.