Heterogeneous Ice Nucleation in Wave Clouds: Dependencies on Temperature, Coarse Aerosol and Nuclei Processing Time

An ice nucleation parameterization used in cloud models is the scheme developed by DeMott and coworkers (DeMott et al., 2010; D10). The basis for D10 is nuclei measurements made with a continuous flow diffusion chamber (CFDC). Because the CFDC captures only two nucleation mechanisms (freezing and deposition), and because the duration of nuclei processing within the CFDC is a constant, and smaller than processing times within clouds, we conducted an airborne study of ice nucleation in middle-tropospheric wave clouds. The bases of our investigation are level-flight in-situ measurements, upward-pointing airborne lidar measurements, and the sinusoid wave structure of the investigated clouds. We derived four parameters for each of the 86 streamlines we analyzed. 1) Coarse aerosol concentration (D>0.5 um), 2) minimum in-cloud streamline temperature, 3) in-cloud nuclei processing time, and 4) ice crystal concentration measured in the downwind cloud tail. The second and third of these were derived using a parcel model initialized with in-situ thermodynamic and kinematic measurements. For our data set the minimum temperature is between -34 to -17 C and the processing time is between 20 to 420 s. We fitted our 86 values of coarse aerosol, minimum temperature and crystal concentration to the form developed by D10 and obtained good agreement with D10. In addition, we found that 78% of our fit predictions are within a factor of two of the observed crystal concentrations. Thus, the fitted-versus-observed variance for our data set is smaller than that in D10. We also attempted to incorporate nuclei processing time into our fit and concluded that time-dependent effects are unresolvable. Our result validates the D10 parameterization used in cloud models.