Key points are not available for this paper at this time.
We have added convective ice lofting to a Lagrangian trajectory model of tropical tropopause layer (TTL) water vapor (H 2 O) and its stable isotopologue, HDO. The Lagrangian model has been previously shown to accurately simulate H 2 O in the TTL and lower stratosphere. We show here that the model does a poor job reproducing the observed HDO depletion ( δ D) in the TTL. When convective ice lofting to altitudes below the cold point (the point where air experiences its lowest H 2 O saturation mixing ratio) is added to the model, there is little change in H 2 O in the lower stratosphere, but a large change in δ D throughout the TTL that brings the model into better agreement with measurements. Thus convective ice lofting has the capacity to improve the model's δ D simulation while not significantly degrading the agreement between simulated and measured H 2 O. Convective ice lofting to altitudes above the cold point, on the other hand, has a large effect on lower stratospheric H 2 O, suggesting that changes in convection reaching these altitudes could drive changes in lower stratospheric H 2 O. This suggests a mechanism by which lower stratospheric H 2 O trends may be at least partially decoupled from tropopause temperature trends. Such a disconnection was suggested by previous observations of simultaneously increasing stratospheric H 2 O and a cooling tropical tropopause.
Dessler et al. (Tue,) studied this question.