Effects of radar range and azimuthal resolution on tornadic shear signatures: applications to a tornado detection algorithm

The current Tornado Detection Algorithm (TDA) used with the Weather Surveillance Radar-1988 Doppler (WSR-88D) network utilizes an input velocity field that is often noisy and subject to de-aliasing errors. The current TDA also depends on azimuthal shear calculations, which are affected by noisy velocity data and can degrade significantly with range. Because of these and other data accuracy issues, the current TDA is prone to producing false detections and inaccurate circulation tracks.

Coincident with the advent of new radar-derived products and ongoing research involving new weather radar systems (e.g., Phased Array Radar; PAR), the National Severe Storms Laboratory is developing an improved TDA. A primary component of this algorithm will be the local, linear least squares derivatives (LLSD) azimuthal shear field. The LLSD method uses rotational derivatives of the velocity field and is less affected by noisy velocity data in comparison to the more traditional “peak-to-peak” azimuthal shear calculations.

Initial detections will be made on a field of maximum low-level LLSD shear and diagnosed for potentially tornadic characteristics. Although LLSD shear is less range-dependent than peak-to-peak shear, some range dependency is unavoidable. A preliminary study of 31 tornadoes indicated that the threshold LLSD shear value needed to detect tornadoes was moderately dependent on range from the radar. A regression analysis was completed to determine the relationship between range and shear values so that range-corrected shear values could be estimated.

In addition to range, azimuthal sampling is an important consideration in tornado detection. Of particular interest for this work is the azimuthal resolution of the National Weather Radar Testbed PAR in Norman, Oklahoma. The beamwidth of the PAR increases smoothly with increasing angle from boresight, ranging from 1.5° at boresight to 2.1° at an angle of 45° from boresight. Although overlapped sampling is applied to the PAR to increase the azimuthal resolution, the PAR does not currently reach the super-resolution capabilities in use with the WSR-88D network. A two -dimensional Rankine vortex model was used to demonstrate the effects of azimuthal resolution and range on peak-to-peak and LLSD shear calculations. Simulated Rankine vortices were sampled with azimuthal resolution mimicking that of the PAR and a typical WSR-88D radar and results were compared.