Examining the relationship between deep convection and cirrus formation in the Tropical Tropopause Layer

Abstract:

Cirrus clouds in the tropical tropopause layer (TTL) can result from either persistence of ice crystals detrained from deep convection or in situ nucleation of ice crystals within the TTL. Recent modeling studies have suggested that radiatively-driven evolution of anvil cirrus can explain much of the cirrus observed in the tropical upper troposphere. Here, we examine the relationship between convection and TTL cirrus using a detailed model of TTL transport and cloud microphysics, with input from deep convection diagnosed from satellite observations. The simulations are evaluated by comparison with CALIOP measurements of TTL cirrus. We focus on two questions related to convective influence with the model simulations: (1) How does the age since convective influence affect the occurrence of TTL cirrus? and (2) How are the TTL cirrus occurrence frequency and regional distribution affected by convection above different altitudes in the TTL? As shown by previous studies, we find that the impact of convection on TTL cirrus occurrence decreases with increasing height in the TTL. In the upper TTL, model simulations without direct convective input can reproduce the observed cirrus occurrence frequency and regional distribution. We further address the processes affecting cirrus occurrence downstream of convection by examining the temperature/relative humidity history in the time between convective detrainment and downstream cirrus occurrence. We find that much of the TTL cirrus likely results from reformation of ice in plumes of moist air downstream of deep convection rather than persistence of ice produced by the convection directly.