Experimental Investigation of the Governing Parameters of Atmospheric Ice Nucleation Using Field-Collected and Laboratory Generated Aerosol Particles and its Application in Cloud Resolving Models

Principal Investigator(s):
Daniel Knopf, The Research Foundation for SUNY Stony Brook University

How aerosol particles affect the radiative properties of clouds, the so-called, “indirect effect” of aerosols, is recognized as one of the largest sources of uncertainty in climate prediction. However, the traditional understanding of the indirect effect does not consider the role of atmospheric ice crystals. Atmospheric ice formation is an important process impacting the distribution of water vapor, precipitation, and ice cloud formation, thereby modulating cloud albedo and thus climate. Processes that lead to the formation of ice crystals from aerosol particles are still insufficiently understood and not well represented in cloud resolving, general circulation models, and Earth system models. The reasons for this are manifold, e.g. i) ice in the atmosphere can form by different nucleation pathways such as ice nucleates from a supercooled aqueous droplet that contains an ice nucleating particle (INP), a process termed immersion freezing, or ice forms from the supersaturated gas-phase on the surface of an INP, termed deposition ice nucleation; ii) only a very small number of the ambient aerosol particles act as INP. Currently, we still have an insufficient fundamental understanding of the relationship between aerosol types and ice nucleating particle (INP) numbers, which in turn has to be parameterized for application in cloud and climate models. In recent years substantially differing mathematically interpretations of laboratory freezing data have been suggested. They can result in different predictions of INP and ice crystal numbers when applied in models. These parameterizations are usually derived by employing experiments that use a single type of aerosol particles, e.g. made up from one component. However, in the atmosphere, the aerosol population is usually multicomponent in nature. How a diverse particle population will initiate ice formation has so far not been systematically investigated. Different aerosol particle types, acting as INPs, will compete for water vapor and ice nucleation. Prediction of the most successful INP is non-trivial, since it may depend on a variety of processes that have not been well understood so far, such as the relative abundance of different particle types, particle size, nucleation time, and multiple aerosol particle-ice activation cycles (i.e. particles that initiated ice previously may act differently the second time). To improve our predictive capability of ice nucleation in an ambient aerosol population, it is imperative to understand the impact of these fundamental parameters on ice formation. This research project will apply a combination of experimental ice nucleation studies. Laboratory experiments will be guided by an Earth system model that includes a newly developed dust module with different mineral species as well as other aerosol types. The model will be used to predict typical aerosol population characteristics for the location of three field campaign studies namely, Mixed-Phase Arctic Cloud Experiment (M-PACE), the Indirect and Semi-Direct Aerosol Campaign (ISDAC), and the Small Particles in Cirrus (SPARTICUS) campaign. In addition, ice nucleation will be studied from ambient particles collected at the Southern Great Plains (SGP) site, representing an authentic well characterized aerosol population. Particle samples will be designed in the laboratory to reflect field-measured and modeled ambient populations in terms of mineralogy, number, and size of particles. These particle samples will be scrutinized for their ability to initiate ice nucleation, thereby testing the predictive capability of current ice nucleation parameterizations. Successful descriptions will be evaluated by calculating global INP concentration fields off-line from global aerosol fields. These fields will be derived from simulations with the Earth system model for the different parameterizations and their uncertainty ranges as determined from the laboratory measurements. The results will be compared to the field measurements. The overarching goal of the project is to find the parameterization with the best skill to reproduce INP for physically realistic thermodynamic conditions and aerosol populations, and to get informed which ones are the parameters with the largest uncertainties, on which future research should be focused. This research aims to reduce model uncertainty that is founded in ice nucleation parameterizations describing the relationship between aerosol composition and INP numbers.