Radar analysis of the airflow over geographic features that may affect mesocyclone intensity and tornadogenesis

It may be shown that horizontal gradients in roughness length, with some component normal to the low level wind, produce quasistationary areas of horizontal shear, and associated vertical vorticity and circulation. The vorticity is positive (negative) when the roughness gradient is directed toward the left (right) looking down wind. In a neutral boundary layer, this vorticity may be detected at significant heights AGL. The intensity of mesocyclones passing across these gradients in friction may be altered, in some cases leading to tornadogenesis.

In addition, the channeling or tunneling of flow in river valleys, gorges, and other such topographic features may also locally change the wind direction and speed, affecting the storm-relative helicity and/or producing quasistationary regions of vertical vorticity. Bosart et al. (2004) and LaPenta et al. (2005) have examined the effects of wind channeling in tornado cases in the northeastern United States.

In this paper, Doppler radar data will be used to examine quasistationary regions of positive or negative vertical vorticity or storm-relative helicity associated with horizontal gradients in friction and others associated with flow channeling. Where available, dual-Doppler synthesis will be used. However, dual-Doppler coverage is not widespread at the present time, and a method has been developed for examining quasi-stationary perturbations to the background flow, and associated vertical vorticity, using a single Doppler radar. It may be shown using the horizontal momentum equation, in the case of along-wind gradients in friction, the wind speed adjusts to the change in friction much more rapidly than the wind direction does, due to the small magnitude of the Coriolis force relative to the drag force. Therefore, VAD wind profiles are used to determine the wind direction at each range and height over the lowest elevation scan of a radar, and wind speeds and vectors are calculated assuming this wind direction estimate. These wind fields are then averaged over multiple volume scans to remove noise and the effect of individual convective elements, producing a map of the average wind at low levels, usually over a 2-3 hour period. Areas of background vorticity may then be located, and the effect upon storms interacting with these areas can be examined. For wind tunneling effects, a similar method is used, except available surface wind data and DEM elevation data are used to estimate the wind direction in channeling areas, and regions of background vorticity are located.

Several case studies will be presented, including: 1) weakening of rotation in convective storms approaching the coast during the landfall of Hurricane Katrina (2005), 2) tornadogenesis at Panama City Beach, Florida near a coastal lagoon and bay during the landfall of Hurricane Ivan (2004), 3) effects of an inland river valley during the Super Tuesday tornado outbreak (2008), and 4) potential effects of flow channeling on tornadogenesis just south of Huntsville, AL (2009).