Weather radars are indispensable tools for probing the winds inside tornadoes and revealing the structure of storms producing tornadoes. Weather radar data are critical to short-term forecasts of tornadoes, and an improved understanding of radar signatures of tornadoes and tornadic storms can improve operational warnings and risk communication and mitigate threats to life and property. An important radar signature for operational applications is the dual-polarization tornadic debris signature (TDS). Debris produce a unique dual-polarization radar signature that can be used to remotely detect debris and characterize how much and how high debris are lofted. Our past research had shown that tornado debris signatures reaching higher altitudes tended to be associated with higher Enhanced Fujita scale rating. 

This project examined the relationships among lofted tornado debris, the associated dual-polarization radar signature, and the three-dimensional winds in tornadoes. While several observation-based hypotheses had been proposed, these had not been rigorously tested with physical modeling. To accomplish this, a unique numerical simulation approach was employed that permitted controlled studies and hypothesis testing of different variables impacting tornado debris signatures by combining a dual-polarization radar simulator with a numerically simulated tornado. The dual-polarization radar simulator initializes user-specified types, sizes, and concentrations of debris, which are then tracked in the tornado's flow using aerodynamic modeling. In addition to different types of debris, a wide range of simulated tornadoes were available as inputs (e.g., weak, strong, and violent tornadoes), which enabled studies of how tornado wind speeds impact the tornado debris signature. 

One of the primary findings from the radar simulation studies was that debris characteristics, such as size or type, can be inferred from some dual-polarization radar measurements. For example, higher returned power (reflectivity) and reduced similarity between the horizontal and vertical signals on receive (correlation coefficient) occurred with larger debris sizes. An increase in the areal extent of the debris signature and a reduction in signal similarity between horizontal and vertical channels was also found as tornado wind speeds increased. This finding supports operational use of the tornado debris signature for characterizing not only damage severity, but the intensity of the tornado. 

A second major component of the radar simulation studies was the development of a technique to correct errors in Doppler velocity measurements in tornadoes. The debris within a tornado move at a different speed than the air motion, with larger differences occurring as debris size increases and often as tornado wind speeds increase. Since the radar measures the speed of objects in the air rather than the air itself, this leads to large wind speed errors, especially since radars typically measure the speeds of the largest debris. To address this problem, a technique was developed using the dual-polarization radar measurements to sort the motions of rain and debris within a radar sampling volume using the raw radar signal. Since the rain drops move at a velocity closer to the air speed, the debris velocities are filtered. Using both radar observations and simulations, the technique was demonstrated to improve the accuracy of Doppler velocities. 

In addition to the simulation-based studies, data from research radars were used to study tornadoes and their debris signatures. An analysis of two simultaneous TDSs was conducted and revealed unique patterns of storm-scale debris lofting and fallout. Using rapid-scan radars, analyses of major tornado outbreaks were conducted to study how rapid changes in the TDS relate to surface damage patterns in populated areas, as well as how debris signatures can develop within intense mesocyclones prior to tornado formation. The project also leveraged advanced, rapid-scanning radars called phased arrays to study the formation and dissipation of tornadoes. In a phased array study of tornado dissipation, a two-stage process of decay was documented including a top-down broadening and weakening of the tornado vortex.  

The project provided unique research opportunities for undergraduate and graduate students, who worked with an interdisciplinary team of meteorologists and engineers. A total of 2 Ph.D., 2 Master's, and 12 undergraduate students worked on the project, including 7 students from underrepresented groups in the atmospheric sciences. The project team participated in the National Weather Festival in Norman, Oklahoma, which involves 5,000 participants each year. During COVID, a virtual tour of the OU Radar Innovations Lab was created with a feature segment on the laboratory used in this study (anechoic chambers) and tornado debris signatures. The project team also developed a small exhibit at the National Weather Museum in Norman, OK on the Atmospheric Imaging Radar, describing how this phased array radar worked, how we used it for field experiments, and what new scientific knowledge was gained from using its data.

Last Modified: 10/26/2023
Modified by: David J Bodine