Abstract
Abstract A number of studies in recent years have used wind fields derived from portable Doppler radars in combination with the ground-based velocity track display (GBVTD) technique to diagnose the primary (tangential) and secondary (radial and vertical) circulations in tornadoes. These analyses indicate very strong vertical motions in the vortex core, in some cases with updrafts and downdrafts exceeding 100 m s−1. In addition, many of the analyses indicate strong radial outflow at low levels and in the vicinity of the low-level tangential wind maximum. This paper shows that strong outward motion at this location cannot be consistent with a tornado circulation that lasts more than a few minutes. In addition, using data from numerical simulations as truth, it is shown that using observed radial velocities to diagnose vertical velocities greatly overestimates the intensity of downward motion in the core for two reasons: neglect of the mass flux into the core through the swirling boundary layer, and the likely positive bias in low-level radial velocities due to the centrifuging of debris. Possible methods for accounting for these errors are briefly discussed.
