Comparing Tropopause-Penetrating Convection Identifications Derived from NEXRAD and GOES over the Contiguous United States

Abstract

In this paper overshooting top (OT) refers to a convective updraft that extends above the primary lapse¬rate tropopause. OTs are a result of powerful updrafts that rapidly transport air from the lower troposphere to the lower stratosphere. They can initiate gravity waves and mixing of air across the tropopause. Within the last decade remote¬-sensing observations from satellites and radars have been used to estimate the altitudes of OTs, but the results have not been consistent. To understand the reasons for the differences and to improve the accuracy of tropopause ¬overshooting top identifications, this study compares overshooting convection identified using NEXRAD radar reflectivity data with two approaches that use GOES IR data. The study region covers a large part of the contiguous United States during selected active convection dates in 2004 and 2006. NEXRAD can estimate the altitude of echo tops directly from the radar pointing geometry. The GOES methods use IR brightness temperature as a proxy to estimate the cloud top relative to the tropopause and surrounding cloud. GOES IR Version 1 retrievals of OTs are heavily weighted towards horizontal temperature gradients within the cloud, while GOES IR Version 2 gives more weight to temperature differences between the potential top and tropopause. We found Version 1 overestimates the frequency of OTs, particularly in the southeast US. The frequency is overestimated because too much weight is given to horizontal temperature gradients within the cloud and not enough weight is given to the temperature difference between the cloud top and tropopause. GOES IR Version 1 erroneously tags events that are warmer than the tropopause and located outside of strong updraft regions as overshoots. Version 2 agrees much better with NEXRAD OT retrievals, particularly when the spatial matching conditions are slightly relaxed.