Aerial Damage Survey and High Resolution Dual-Polarization Radar Analysis of the 2013 El Reno Tornado

On 31 May 2013, a large and deadly EF3 tornado produced a U-shaped damage swath over the rural area south of El Reno, Oklahoma. Details of the damage swath are documented in a comprehensive aerial and ground damage survey. The damage path is approximately 23 km long and 6 km wide in one section. In addition to WSR-88D measurements, three mobile Doppler radars collected high-resolution radar data at close range to the tornado. They include the rapid-scanning X-band polarimetric Doppler radar (RaXPol), and the Doppler on Wheels dual-polarization (DOW6) and rapid-scan phased array (DOW8) radars. In this study, we integrate the damage survey and radar data to document the complex and detailed structure of the El Reno tornado damage path.

The U-shaped path contains many examples of non-linear damage markings that are only visible from the air. A cusp is observed in the damage path as a region of matted vegetation. Small-scale suction vortex paths are evident in many fields. A looping path is observed near the location of veteran storm chaser fatalities in the aerial imagery and rapid scan radar couplet locations. Further evidence of multiple vortices is evident as a wavenumber 4 pattern in the RaXPol data. Enhanced damage is collocated with two of the sub vortices. An anticyclonic tornado is observed both in the damage survey and radar data to the south of the El Reno tornado.

The challenges of surveying a multiple vortex tornado in a rural region will be discussed. In particular, we will argue that areal surveys are critical when mapping out details in the damage path. Much of the damage may be to vegetation that is not obvious on the ground. Furthermore, areal imagery taken at multiple viewing angles is paramount as subtle field markings are only visible at particular viewing angles relative to the sun.

Finally, some results of a comprehensive areal and ground survey of the 2013 EF5 Moore Oklahoma tornado are discussed. Aerial and ground survey teams independently rated over 4000 structures (78% are residential buildings) damaged by the Moore tornado. For the first time, we are able to compare the aerial and ground surveys to identify biases. It will be shown that the areal survey contains a low bias in the ground rated EF0 and EF1 structures. This is attributable to the difficulty in observing weak damage to residential from overhead aerial photos. The need for aerial ortho and oblique imagery for future surveys will be discussed.