Heterogeneous Ice Nucleation in the Tropical Tropopause Layer

Recent laboratory experiments have advanced our understanding of heterogeneous ice nucleation at low temperatures. We use these laboratory results, along with field measurements of aerosol composition, to constrain a parameterization of heterogeneous nucleation rate dependence on temperature and supersaturation. We apply this nucleation parameterization in detailed simulations of transport and cloud formation in the Tropical Tropopause Layer (TTL,similar or equal to 14-18 km) constrained by high-altitude aircraft measurements of TTL cirrus microphysical properties and supersaturation. The model results indicate that if the abundant TTL organic-containing aerosols were effective glassy ice nuclei (IN), as indicated by laboratory experiments with simple sugars/acids, then heterogeneous nucleation would prevent the occurrences of large ice concentrations and large ice supersaturations, both of which are clearly indicated by the in situ observations. If glassy organic aerosols are relatively poor IN, as indicated by laboratory experiments with more atmospherically relevant secondary organic aerosol-like composition, then the simulated ice concentrations and supersaturations are in reasonable agreement with the observations. Concentrations of effective mineral dust or ammonium sulfate IN larger than about 50/L can drive significant changes in cirrus microphysical properties and occurrence frequency. The cloud occurrence frequency can either increase or decrease, depending on the efficacy and abundance of IN added to the TTL. Relatively low TTL concentrations of mineral dust particles (<= 10 L-1 indicated by the limited available field measurements) have negligible impact on cirrus frequencies or microphysical properties. Given the assortment of uncertainties, meaningful estimates of TTL aerosol-cirrus indirect effects on radiative forcing are likely not possible at this time.