Monitoring overshooting tops in the tropical cyclone eyewall using bispectral satellite imagery

Abstract

Vigorous convective towers in tropical cyclone (TC) eyewalls likely play a significant role in TC intensification. The current satellite remote sensing tools used by TC analysts and forecasters have some shortcomings in their ability to identify and monitor such convective towers, though there exists an easily derivable product that, in theory, may alleviate these issues: the difference between brightness temperatures of the infrared (IR) minus water vapor (WV) channels from the GOES satellite imagers. The ability of this difference (hereafter called IRWV) to identify vigorous eyewall convection was tested on four TCs using subjective interpretation and pixel counting schemes, with both compared to overshooting tops identified in visible and IR. The subjective interpretation scheme produced dramatically varied results, including one TC that had negative Gilbert and Heidke skill scores. IRWV images that correctly identified overshooting tops in the subjective interpretation scheme were frequently marred by IRWV signals that would impair interpretation without any corroborating visible imagery, thus defeating the purpose of IRWV. The pixel counting scheme results showed that counts of IRWV ≤-3 K pixels near the TC core were unable to distinguish images containing overshooting tops from those with none. An examination of soundings near the TCs showed that a steeper lower-stratospheric inversion increases the number of IRWV < 0 pixels per image, resulting in geographical inconsistency in the performance of the IRWV product. The IRWV product is extremely inconsistent in its ability to correctly identify overshooting tops in TC eyewalls, and should not be depended on in this function.