Inferred Differences in Ice Crystal Nucleation Rates between Continental and Maritime Deep Convective Clouds

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Wednesday, 7 January 2015
David L. Mitchell, DRI, Reno, NV; and M. A. Avery and A. Garnier

Based on in situ measurements of the ice particle size distribution (PSD) from two aircraft field campaigns (SPARTICUS & TC4) and MODIS satellite retrievals of the temperature dependence of the 12/11 μm effective absorption optical depth ratio or βeff, ice crystal nucleation rates appear to be anomalously high near the tops of continental thunderstorms relative to maritime thunderstorms. We present in situ and remotely sensed evidence for the following working hypothesis: Heterogeneous nucleation dominates during deep continental convection until ice nuclei (IN) in the updraft cannot prevent the supersaturation from increasing. As it increases, homogeneous nucleation eventually occurs near cloud top (T < -60C), with much faster ice crystal production rates. This is not the case in maritime anvil cirrus, where updrafts associated with deep convection are slower, promoting heterogeneous nucleation. We hypothesize that differences in updraft velocities and their effect on supersaturation might create a difference in the ice nucleation modes near cloud top.

The ice crystal nucleation rate, having units of g-1 s-1, is more related to the ratio of ice particle number concentration/ice water content (or N/IWC, with units of g-1) than to number concentration N. Hence the N/IWC ratio can be used as an index for ice nucleation rates in ice clouds since the N/IWC ratio should be most affected by the ice nucleation rate. Since it is well known that homogeneous nucleation generally produces ice crystals at a much faster rate than heterogeneous nucleation, the N/IWC ratio may provide a means of assessing whether homo- or heterogeneous nucleation is likely to dominate ice production.

Calculating βeff from the from the TC4 and SPARTICUS in situ PSD data, a surprisingly tight relationship was discovered between βeff and N/IWC. This enables accurate retrievals of N/IWC over the higher range of N/IWC ratios where βeff > 1.07 or N/IWC > 210^7 g-1 (based on the 12 and 11 μm channels). Preliminary results suggest that homogeneous nucleation dominates ice production for βeff > 1.15, or N/IWC > 10^8 g-1, although this needs to be confirmed through cloud modeling. It was also confirmed that satellite retrievals of βeff during TC4 were consistent with βeff calculated from TC4 PSD data. These retrievals verified that deep convection during TC4 over water did not produce the much higher N/IWC ratios observed during SPARTICUS in continental anvil cirrus.

These results and the above hypothesis imply that maritime anvil cirrus are formed through heterogeneous nucleation at all height levels and thus their microphysical and radiative properties may be affected by the concentrations of mineral dust and other IN. On the other hand, since homogeneous ice nucleation depends less on aerosol concentrations, and since a cloud's albedo is most sensitive to its microphysical properties near cloud top, the albedo of continental anvil cirrus may be relatively insensitive to changes in aerosol and mineral dust concentration based on the anvils sampled. While the albedo of maritime and continental anvil cirrus may be comparable, higher IWCs enhance albedo in the maritime case while higher ice production rates with smaller ice crystals enhance albedo in the continental case.

The imaging infrared radiometer (IIR) aboard CALIPSO has channels at 8.7, 10.5 and 12.0 μm and provides a data record of βeff dating back to 2006, as well as vertical profiles of IWC, extinction, depolarization and 1064/532 nm backscatter ratio from the CALIOP lidar. We will compare the MODIS-derived βeff - N/IWC relationship with that derived using the IIR data. We will also investigate the relationship between N/IWC, βeff and the vertically-resolved lidar parameters to better characterize the modes of nucleation and to determine whether a change in nucleation mode produces a measurable change in the vertical distribution of cloud ice. While many aspects of this work are still being explored, it appears possible that these relationships may be used to determine when and where homo- and heterogeneous nucleation dominate ice production in cirrus clouds as a function of season and latitude.