Observational Constraints on the Impact of Tropopause-Penetrating Convection on Lower Stratospheric Composition over the US in Summer (Invited Presentation)

Tropopause-penetrating convection provides a means of rapidly and irreversibly transporting lower tropospheric air and water vapor into the lower stratosphere. Given sufficient frequency, these small-scale processes can have a significant impact on the chemical and radiative properties of the upper troposphere and lower stratosphere (UTLS). Significant progress has been made in identifying regions prone to tropopause-penetrating convection with global satellite sensors. The region over the Great Plains of the U.S. in North America during summer (Jun, Jul, Aug) stands out globally both for the frequency of convection penetrating the local tropopause, and for the mean area of convective cores at the tropopause level. The development of high-resolution regional climatologies of these tropopause-penetrating convective events - utilizing long term records from ground-based (NEXRAD) and geostationary satellite (GOES) observing systems - provides a rich dataset for examining their frequency and geographic distribution as well as for determining their stratospheric penetration depths. The NEXRAD data show, for example, that on average approximately six thousand overshooting updrafts reach or exceed 390 K in potential temperature from June through August over the U.S. Concurrently, both satellite (Aura MLS) and in situ (Harvard Water Vapor) data have shown that tropopause-penetrating convective events can lead to localized enhancements in stratospheric humidity. While not all storms that penetrate the local tropopause level lead to significant hydration, the large number of observed events over the U.S. in summer suggest that the net impact of convection in this region and season may be significant, especially because these storms occur in the vicinity of the North American Monsoon upper-level Anticylone (NAMA), which acts to partially confine the convectively influenced air in the UTLS and increase its residence time over the region. Given the potential for dramatic changes to convective frequency and strength in response to anthropogenic climate forcing, it is important that the impact of convection on regional-scale stratospheric composition be well characterized for incorporation into prognostic chemistry and climate models.

In situ data provide a means of determining the quantity of water vapor delivered to the stratosphere by individual convective storm systems on a case-study basis, however, the net impact of tropopause-penetrating convection at larger spatial and temporal scales remains poorly quantified. In the present analysis, we utilize a range of approaches from simple back-of-the-envelope estimates to a more sophisticated calculation utilizing the full complement of in situ and satellite datasets to estimate the quantity of water vapor irreversibly delivered by overshooting convection to the lower stratosphere over the U.S. in summer. A preliminary analysis of the regional-scale datasets provided by NEXRAD and Aura MLS reveals 1) correlations between both sub-seasonal and inter-annual variability in the number of tropopause-penetrating storms over the conterminous U.S. and enhancements in mean water vapor at 100 hPa over the NAMA region; 2) similar correlations between storm overshoot frequency and the number of water vapor extrema observed by MLS in the lower stratosphere over this region; and 3) that the correlations are strongest for the deepest storms, i.e., those with radar echo top heights >19 km.

We will also discuss what we believe are the dominant regions for the convectively sourced moisture evident in the NAMA. GOES overshooting top (OT) data provide the observational coverage needed to evaluate convective contributions over North America. A preliminary examination of GOES data, produced for the summer of 2013, analyzed in conjunction with three-dimensional forward trajectory calculations using reanalysis winds reveals 1) that there are a few discrete zones in the NAMA region prone to frequent overshooting convection; 2) that the geographical distribution of convection shifts over the course of the summer, i.e., from May to October; 3) that the deepest OTs occur exclusively over the Central U.S.; and 4) that air within the NAMA is likely influenced not only by local OTs, but also by OTs that occur over the Southeastern U.S. and the Sierra Madre.