Comparing Tropopause-Penetrating Convection Identifications Derived From NEXRAD and GOES Over the Contiguous United States

Overshooting tops (OTs) are a well-known indicator of updrafts capable of transporting air from the troposphere to the stratosphere and generating hazardous weather conditions. Satellites and radars have long been used to identify OTs, but the results have not been entirely consistent due to differences in sensor and measurement characteristics. OT detection approaches based on satellite infrared (IR) imagery have often been validated using human-expert OT identifications, but such datasets are time-consuming to compile over broad geographic regions. Despite radar limitations to detect the true physical cloud top, OTs identified within multi-radar composites can serve as a stable reference for comprehensive satellite OT analysis and detection validation. This study analyzes a large OT data set compiled from Geostationary Operational Environmental Satellites (GOES)-13/16 geostationary IR data and gridded volumetric Next-Generation Radar (NEXRAD) reflectivity to better understand radar and IR observations of OTs, quantify agreement between satellite and radar OT detections, and demonstrate how an increased spatial sampling from GOES-13 to GOES-16 impacts OT appearance and detection performance. For nearly time-matched scenes and moderate OT probability, the GOES-13 detection rate (similar to 60%) is similar to 15% lower than GOES-16 (similar to 75%), which is mostly attributed to coarser spatial resolution. NEXRAD column-maximum reflectivity and tropopause-relative echo-top height as a function of GOES OT probability were quite consistent between the two satellites however, indicating that efforts to account for differing resolution were largely successful. GOES false detections are unavoidable because outflow from nearby or recently decayed OTs can be substantially colder than the tropopause and look like an OT to an automated algorithm.

Automated summary

The study analyzes a large dataset of OT compiled from GOES-13/16 geostationary IR data and gridded volumetric NEXRAD reflectivity. It aims to better understand radar and IR observations of OTs, quantify agreement between satellite and radar OT detections, and demonstrate the impact of increased spatial sampling from GOES-13 to GOES-16 on OT appearance and detection performance. The research finds that GOES-13 detection rate is similar to 15% lower than GOES-16 due to coarser spatial resolution, but efforts to account for differing resolution were largely successful in maintaining consistency between the two satellites.