Impact of Convectively Detrained Ice Crystals on the Humidity of the Tropical Tropopause Layer in Boreal Winter

Deep convection detraining in the uppermost tropical troposphere is capable of transporting water vapor and ice into the tropical tropopause layer (TTL), but the impact of deep convection on the global and regional TTL water vapor budget remains uncertain. In particular, the role of convectively detrained ice crystals that remain suspended after active convection has subsided is not well understood. These ice crystals represent aging cirrus anvils detached from the convective core. We use a cloud microphysical model that tracks individual ice crystals throughout their lifetimes to quantify the impact of detrained ice on the humidity of the TTL during boreal winter. Convective influence of air parcels near the wintertime cold point tropical tropopause is determined by tracing thousands of backward trajectories through satellite‐derived, global, 3‐hourly convective cloud‐top altitude fields. Detrained ice, most of which is found over the tropical western Pacific, experiences cooling on the order of 1 K day −1 downstream of convection. Downstream cooling increases relative humidity and explains the observed supersaturated TTL over this region. Vapor in excess of saturation condenses onto the detrained ice, which ultimately brings the relative humidity down to saturation. Thus, convectively detrained ice crystals in aging anvils predominantly dehydrate the TTL, but the effect is small (0.01 ppmv). Moistening by active convection (0.30 ppmv), including the rapid sublimation of convectively lofted ice crystals near the tops of core anvils, overwhelms the dehydration by aging anvil ice crystals detrained from the core. The net effect is moistening by convective core anvils during boreal winter.