The Calm before the Storm: Comparing the Initial Stages of Tornadic versus Nontornadic Supercells

In Oklahoma alone, severe weather occurs on average 52.5 days per year across the state. Occurring most frequently in late spring and early summer, this weather takes the form of violent thunderstorms such as rotating multi-cell thunderstorms known as supercells that can produce threats such as large hail, damaging winds, and tornadoes (Ahrens and Henson, 2019). Due to the frequency of these severe weather conditions and the dangers posed to people, accurately predicting the onset of this weather is crucial for the continued safety of residents of Oklahoma and across the country.

In order to better understand what conditions might influence the formation of tornadoes I studied both tornadic and non-tornadic supercells. This work was done primarily through the use of computational models using the Cloud Model (CM1) code and the high performance computing system at Oklahoma State University. CM1 is a program that was designed to model small-scale atmospheric processes and can be customized to simulate many instances of severe weather. For this project we used CM1 to model a nontornadic supercell and a tornadic supercell thunderstorm using a combination of preconfigured code and soundings collected by the National Weather Service in Norman, Oklahoma during a storm that took place El Reno on May 31, 2013.

In order to study the atmospheric conditions prior to the formation of supercell storms we first exained data collected by the Oklahoma Mesonet system during supercell storms occurring in 2013 and 1998. From comparing the atmospheric pressure and wind speeds for these two storms we found that though the pressure dropped during each storm it was significantly higher overall and closer to standard atmospheric pressure for the non-tornadic supercell than for the tornadic supercell.

Next, we used the high performance computing system and CM 1 code to simulate a non-tornadic supercell to output pressure and wind speed close to the ground. In comparing this output to the non-tornadic supercell data from the Mesonet and found that the simulated data behaved similarly to the data from the observations, in which the pressure dropped but was close to standard atmospheric pressure throughout.

Over the course of this project, primary findings indicated that lower pressure at the surface prior to supercell development may indicate tornados are more likely. There is much more research to do in this area, primarily studying more supercells storms to see if this of higher pressure compared to tornadic supercell holds. If this research is performed it may present greater capabilities to determine whether or not a tornado will form as a supercell develops.