Water vapor in the stratosphere is primarily controlled by temperatures in the tropical upper troposphere and lower stratosphere. However, the direct impact of deep convection on the global lower stratospheric water vapor budget is still an actively debated issue. Two complementary modeling approaches are used to investigate the convective impact in boreal winter and summer. Backward trajectory model simulations coupled with a detailed treatment of cloud microphysical processes indicate that convection moistens the global lower stratosphere by approximately 0.3 ppmv (~10% increase) in boreal winter and summer 2010. The diurnal peak in convection is responsible for about half of the total convective moistening during boreal winter and nearly all of the convective moistening during boreal summer. Deep convective cloud tops overshooting the local tropopause have relatively minor effect on global lower stratospheric water vapor (~1% increase). A forward trajectory model coupled with a simplified cloud module is used to esimate the relative magnitude of the interannual variability of the convective impact during 2006-2016. Combing the results from the two models, we find that the convective impact on the global lower stratospheric water vapor during 2006-2016 is approximately 0.3 ppmv with year-to-year variations of up to 0.1 ppmv. The dominant mechanism of convective hydration of the lower stratosphere is via the detrainment of saturated air and ice into the tropical uppermost troposphere. Convection shifts the relative humidity distribution of subsaturated air parcels in the upper troposphere toward higher relative humidity values, thereby increasing the water vapor in the stratosphere.

Recently, Anderson et al. (2012, https://doi.org/10.1126/science.1222978, 2017, https://doi. org/10.1073/pnas.1619318114) and Anderson and Clapp (2018, https://doi.org/10.1039/C7CP08331A) proposed that summertime convectively injected water vapor over North America could lead to stratospheric ozone depletion through halogenic catalytic reactions. Such ozone loss would reduce the ozone column and increase erythemal daily dose (EDD). Using 10 years of observations over the North American monsoon region from the Aura Ozone Monitoring Instrument, we find that the column ozone and EDD has a ~0.8–0.9 spatial correlation with lower stratospheric water vapor measured by the Aura Microwave Limb Sounder. We show that this correlation appears to be due to the elevation of the monsoonal tropopause and associated monsoonal convection. The increase in tropopause altitude reduces the ozone column and increases EDD. We see no apparent evidence of substantial heterogeneous chemical ozone loss in lower stratospheric ozone coincident with the stratospheric monsoonal water vapor enhancement.