Szoke and Pietrycha

The landspout life cycle...maybe not as simple as previously thought

Ed Szoke* Albert Pietrycha
NOAA Forecast Systems Laboratory NCAR Research Applications Branch
Boulder, Colorado 80303 Boulder, Colorado 80301
email: szoke@fsl.noaa.gov
phone: 303-497-7395
*Joint collaboration with the Cooperative Institute for Research in the
Atmosphere, Colorado State University, Fort Collins, CO.

While there remains much uncertainty about how the tornado itself actually forms, it has generally been regarded that the most straightforward tornadogenesis mechanism occurs with the nonsupercell tornado, or "landspout". Earlier work based on radar studies by Brady and Szoke (MWR 1989) and Wakimoto and Wilson (MWR 1989) agreed on the following scenario for the development of a nonsupercell tornado: 1) a circulation (vorticity already in the vertical) develops along a stationary or quasi-stationary boundary, often before any clouds of consequence have developed with said boundary; 2) if a developing cumulonimbus cloud forms over the circulation, stretching associated with the developing storm's updraft concentrates the circulation into a tornado. While there may be other factors, such as a contribution to the vertical vorticity by the upward tilting of baroclinically generated low-level vorticity associated with the boundary, the most important implication of the scenario for landspout development was that the tornado would form in the active updraft stage of the developing storm, quite possibly before much, if any, precipitation had fallen. The process is deemed "simple" because the ambient vorticity is already in the vertical, compared to the more complex processes involved in getting low-level vertically oriented near-ground vorticity in a supercell storm.

Although the development of a typical landspout appeared then to be fairly straightforward; circulation develops along a boundary, cloud grows above circulation, tornado forms; the implications of the nonsupercell tornado life cycle for warning or short-term forecasting of the event were perhaps not so simple. Because the circulation formed from the near-surface and grew upwards, in the abscence of any midlevel circulation, any Doppler radar signal would first be present at LOW levels. These levels, in fact, might be SO low, and the size of the developing circulation so small (compared to a well-developed mesocyclone at midlevels in a supercell) that the effective range of being able to see the important features with a radar might only be 60 to 100 km. Still, as Brady and Szoke pointed out, one could use the fact that landspouts formed along a boundary in the early but generally rapid growing stages of a cell as a key to focusing on a fairly restricted area for potential landspouts.

This very forecasting stategy was employed by one of the authors along with his forecasting associate at the Denver Weather Forecast Office on the afternoon of 9 August 1996. Early on the forecaster's attention was focused on a stationary boundary that frequently forms near Denver, the Denver Convergence-Vorticity Zone (DCVZ), which on this day was positioned basically right near the Denver International Airport (DIA). The threat of a possible tornado was highlighted in the Nowcast as storms began to develop above the boundary, and then...nothing happened. At least nothing for the next 60+ minutes, as a line of showers and thunderstorms formed along the boundary. At the point when the forecasters essentially gave up on the possibility of a landspout, following the most likely time of tornado development in the landspout life cycle, three tornadoes formed just south of DIA. One of the authors captured the events on video, and the tornadoes occurred within a few miles of the 88D radar near DIA, with well-defined velocity signatures. The excellent radar and photographic coverage of this event should allow us to explore what happened on this day and why the tornadoes did not seem to follow the "typical" landspout life cycle. One hypothesis is that early circulations were too weak, and the eventual tornadoes that developed may have required a merger of circulations, as modeled by Lee and Wilhelmson (last SLS Conference). With the radar so close to the event we should be able to follow precursor circulations and determine if this hypothesis is true.