Numerical Simulation of a Mini-Supercell over Kanto Plain on 19 September 1990

On the evening of 19 September 1990, nine mini-supercells were observed over the Kanto plain, which covers the area of 150km by 150km. Three of them spawned a tornado. The Kanto plain was located in the northeast quadrant of Typhoon 9019 moving northeastward. An upper-air sounding at Tateno near the center of the Kanto plain indicated that a strong veering shear existed in the lowest 2km AGL. The characteristics of the storms were observed by a single Doppler radar ( Suzuki et al., accepted for Mon. Wea. Rev.).

A storm-scale numerical simulation in which the upper-air sounding at Tateno is given as the horizontally uniform basic state has been performed with ARPS Ver. 4.2.1 developed by the Center for Analysis and Prediction of Storms (CAPS), University of Oklahoma. The simulated storm lasted for more than 160 min before it approached the boundary of the calculation domain. It exhibited a cyclic generation every one hour. The characteristics of the observed storms such as a small horizontal dimension, a low cloud-top, a north-south elongated echo pattern with a hook at the south edge, a mesocyclone, a weak echo region has been well reproduced in the simulation. The temperature difference across a simulated gust front is found to be at most 2K, which also agrees with the surface observation, and is much smaller than that found in a classic supercell observed in the Great Plains in the United States.

The generation mechanism of vertical vorticity of the mature low-level mesocyclone in the simulated storm is examined by a vorticity budget analysis along back trajectories during 20 min. The trajectories coming into the low-level mesocyclone are found to consist of two major storm-relative paths: One originates near the surface of the north of the mesocyclone and flows straightly southward into the mesocyclone. The other originates at about 1000m AGL to the east or northeast of the mesocyclone, descends to the low levels while turning cyclonically, and finally flows into the mesocyclone from its west.

A new method to accurately evaluate a fractional contribution of baroclinically-generated horizontal vorticity to the vertical vorticity of the mesocyclone has been invented. For the former path, tilting of baroclinically-generated horizontal vorticity and subsequent vertical stretching explains about a half of the vertical vorticity of the mesocyclone. This occurs because the small temperature difference across the gust front is concentrated in a relatively narrow region, so that the horizontal temperature gradient is nearly comparable to that in a classic supercell. A majority of the trajectories, however, belongs to the latter path, for which baroclinically-generated horizontal vorticity is of little importance and the vertical vorticity is generated principally through tilting of the horizontal vorticity in the environmental wind and subsequent vertical stretching. The vorticity dynamics in the present mini-supercell is discussed by comparing with that of a simulated Del City storm.