Research Staff: Eric Jensen, Andrew Ackerman, Leonhard Pfister

Over the past 20 years, many theories have been proposed to explain the extreme dryness of air in the tropical lower stratosphere. Recent observations suggest that the flux of air into the stratosphere may be dominated by slow ascent across the tropopause throughout much of the tropics. In this study, we use cloud model simulations to show that laminar, optically thin cirrus clouds (frequently observed near the tropopause) can effectively freeze-dry air entering the tropical stratosphere. We use our detailed ice cloud microphysical model coupled to a large-eddy simulation dynamical model for these simulations. As shown in the top panel of the figure, if no cloud forms, the slow ascent across the tropopause will eventually increase the water vapor mixing ratio above 5 parts per million by volume (ppmv). These values are much higher than observed water vapor mixing ratios. However, if we include cloud formation, then the slow ascent drives adiabatic cooling and nucleation of a small number of ice crystal (< 10/L). These crystals grow rapidly and precipitate out within a few hours. The ice crystal nucleation and growth prevents the relative humidity with respect to ice from rising above the threshold of ice nucleation (130-160%) and limits the water vapor mixing ratio above the tropopause to 3-4 ppmv (bottom panel of figure). The nucleation threshold depends upon the aerosol composition in the tropopause region, which is not well known. We have also run simulations including gravity waves propagating through the model. Temperature oscillations driven by the waves drive nucleation of larger ice number densities and more complete dehydration of the rising air. Hence, the effectiveness of upper tropospheric aerosols as ice nuclei and the climatology of waves in the tropopause region. We have gathered in situ tropical humidity observations from several field experiments. These measurements included accurate water vapor sensors mounted on the NASA ER-2 as well as balloon-borne instruments. The humidity observations provide a few examples of supersaturated air near the tropopause (as predicted by our model); however, further observations of water vapor and wave motions near the tropical tropopause are required to clarify the cloud formation and dehydration processed.

Point of Contact: Eric Jensen, 650/604-4392, ejensen@sky.arc.nasa.gov

Contours of water vapor mixing ratio (shading) are plotted versus time and height. The top panel shows that if no cloud forms, ascending air transport excessive amounts of water across the tropopause into the stratosphere. The bottom panel shows the dehydration of rising air if cloud formation is included in the model. The white contours show the cloud ice water content.