The El Reno Tornado: Mobile Polarimetric Radar Data and Photogrammetric Analysis

On 31 May 2013, a large and intense tornado formed near El Reno, Oklahoma. This destructive event occurred only 11 days after the Moore tornado and was noteworthy since it resulted in the deaths of several storm chasers. The size and motion of the tornado made it challenging to strategically place mobile Doppler radars in a position to collect high-resolution data on the circulation. The focus of this study is on the volumetric data collected by the rapid-scanning X-band polarimetric Doppler radar (RaXPol) that was located a few kilometers from the El Reno tornado. In addition, a series of still photographs and a high-resolution video simultaneously recorded the visual features of the funnel cloud. Photogrammetric techniques were used to merge vertical cross sections of the radar reflectivity and single-Doppler velocity data through the center of the circulation onto the pictures. Unique in the present study is the incorporation of the differential reflectivity (Zdr) and horizontal-vertical correlation coefficient (ρhv) data into the photogrammetric analyses. The latter two variables have been shown to be important in defining the characteristics of the debris lofted by the tornado. This is believed to be the first time that the debris fields defined by the polarimetric data were superimposed onto pictures of a tornado.

A weak-echo hole (WEH) was evident within the hook echo with minimum reflectivities <-5 dBZ. WEHs have been shown to result from centrifuging of hydrometeors and/or debris within the intense vortex producing a small echo-free region. Vertical cross sections of single-Doppler velocities revealed two rotational couplets. The large couplet was separated by ~2 km and was associated with the mesocyclone circulation. The smaller couplet associated with the tornado was separated by a few hundred meters. The photogrammetric analysis placed that the strongest speeds (>110 ms-1) at the lowest elevation scans at the visible edge of the funnel cloud. The ρhv fields reveal that the boundary of the strongest debris signature (ρhv<0.5) was collocated with the southern edge of the funnel cloud. The northern limit of the debris signature, however, extended far to the north of the tornado circulation suggesting that debris was being carried aloft and, subsequently, falling to the surface in regions where the updrafts were weaker. A prominent V-shaped wedge of high ρhv values that encircled the WEH was resolved during one of the analysis times. Very low values of ρhv indicating the presence of debris were located inside and outside of the V-shaped wedge. A V-shaped region was also noted in the Zdr field associated with values ranging from 2-4 dB. These combined observations suggest that the wedge was comprised of precipitation particles and is an example of the complexity of the centrifuging process of debris and hydrometeors

The photographs and videos recorded on this day also revealed areas distant from the tornado funnel that were characterized by pockets of dust/debris that were lofted above the surface. Analysis of the ρhv fields revealed that these dust/debris areas were associated with values between 0.8-0.9 and is the first visual confirmation of such features that have been previous noted in the literature. In contrast, these pockets of debris were not detected in any of the Zdr plots. It was possible to track dust features in the high-resolution video images to estimate the radial and vertical velocities at the periphery of the El Reno. Updrafts near the funnel edge were estimated to be ~16 ms-1 and the inward radial velocities were ~36 ms-1. These speeds can be quantitatively compared with future Doppler radar estimates (e.g., the GBVTD methodology) to determine the contamination of the velocity data to centrifuging of debris.