Understanding the Evolution of Anvil Top Characteristics with Geostationary Satellite Imagery

Deep convective storms with strong updrafts generate robust anvils and distinct cloud top structures. These structures, including overshooting tops (OTs) and above anvil cirrus plumes (AACP), can be routinely identified atop the most powerful and damaging storms on the planet, and are often indicative of the presence of severe weather including tornadoes and large hail. While satellite imagery allows us to observe such anvil cloud features, the dynamic evolution and relationships between these features are difficult to analyze in imagery with the human eye alone. As such, automated satellite-based methods have proven to be valuable for detecting and analyzing these anvil cloud features, and thus, useful for severe weather forecasting.

Previous attempts have been made to gather bulk statistics of anvil cloud top properties with 10-minute full disk observations of overshooting convective storms. However, the temporal resolution of such observations pushes the limit towards the established buoyancy frequency of typical OTs (10 minutes), thus under-sampling the detailed dynamic evolution of these features. The goal of this study is to derive key anvil cloud statistics from a robust database of tracked anvils and OTs with rapid scan geostationary satellite data­­­. As such, this work aims not only to characterize when and where OTs occur, but also how anvil cloud top features evolve throughout a storm’s lifetime.

To address our goal, overshooting cloud top features identified across several U.S. convective seasons observed with 1-minute GOES-R series imagery are tracked in space and time with the Tracking and Object-Based Analysis of Clouds (tobac) algorithm. By leveraging tobac’s new family tracking tool and a deep learning objective OT detection model, we are able to track the evolution of anvil and OT characteristics through each observed convective storm’s lifetime. The analysis of bulk anvil cloud statistics of and relationships between anvil and OT properties derived from the tracked OT database are presented. Properties include minimum brightness temperatures, OT area, OT lifetime, and number of OTs within each associated cloud anvil region. The implications of this analysis for our understanding of severe storm characteristics such as updraft strength and storm organization will also be presented. This work reveals novel information about not only the relationships between convective anvils and their cloud top structures, but also their evolution through space and time.