The Genesis of Three Nonsupercell Tornadoes Observed with Dual-Doppler Radar

Abstract

Dual-Doppler radar analyses of three tornadoes associated with a multicellular line of storms are presented. The F2?F3 intensity tornadoes occurred on 15 June 1988 near Denver, Colorado, during the Terminal Doppler Weather Radar (TDWR) Project. These tornadoes developed from misocyclones of no larger than 2 km in diameter that formed along the collision of two surface outflows. The misocyclones were observed to build in height and intensify with time, coincident with rapid storm growth overhead. All three misocyclones were clearly associated with the maximum storm updrafts. Downdrafts and associated outflows did not play a role in the formation of one of the tornadoes, but may have contributed to the genesis of the other two tornadoes. It is clear that a downdraft is not a necessary condition for the formation of a nonsupercell tornado, but when present, likely plays a role in determining the timing and intensity of the tornado. This is achieved by the downdraft and outflow causing an increase in the magnitude of the low-level convergence and updraft. Vertical vorticity production terms were examined for each tornado. Given the close proximity in time and space of the tornadoes, there was surprising variability in the magnitudes and locations of the stretching, tilting, and advection terms for each tornado. In general, however, the predominant contribution to positive vertical vorticity and tornadogenesis was from vorticity stretching in the 0.2?2.0-km layer resulting from intensification of low-level convergence and storm updrafts. Above 2.0 km, increased vertical vorticity resulted from a redistribution of low-level vorticity vertically. Small areas of positive vorticity tilting were found within the regions of large streamwise vorticity just prior to tornadogenesis but not during the formative stages of the mesocyclones, amplifying the already strong contributions to tornadogenesis from vertical stretching of the vortices. The spatial resolution of the data presented here is as high as any documented in tornado literature. However, limitations in what features are actually resolvable became strikingly apparent and are discussed in the paper.