Analysis of MCV Tornadoes through Storm-scale Data Assimilation and Simulations

The 3DVAR system of the Advanced Regional Prediction System (ARPS) is used to assimilate data from the CASA (Center for Collaborative Adaptive Sensing of Atmosphere) and WSR-88D radars together with more traditional forms of data such as ASOS, RAOBs, wind profiler data, and observations from the Oklahoma Mesonet, for the case of a mesoscale convective vortex (MCV) that developed on 9 May 2007 in southwest Oklahoma during the CASA Spring 2007 Experiment. This convective system, containing a number of multi-cell storms, produced several small tornadoes.

Data assimilation and forecast experiments are conducted on two one-way nested grids that have 2 km and 400 m grid spacings. Data assimilation is performed on both grids. On the 2 km grid, radar data are assimilated every 5 minutes over a 1 hour assimilation window. On the 400 m grid, radar data are assimilated every 5 minutes for assimilation window lengths of between 40 and 80 minutes. In the subsequent forecasts, the model predicts the timing and location of the MCV with great accuracy. Furthermore, forecasts on the 400 m grid generate two tornado-like vortices (one near Minco, OK and another near El Reno, OK) that track within 5 km/10 min of the confirmed tornado location/time.

The model simulations as well as observational data indicate a strong resemblance between the structure and evolution of the MCV and that of a mid-latitude wave-cyclone. An area of persistent shear and convergence associated with the MCV is found to be responsible for the generation and enhancement of the vortices that become tornadic. At later stages of low-level vortex intensification, the simulated vertical velocity field shows structures that are similar to those of rear-flank/occlusion downdraft in supercell thunderstorms. Further analysis on the tornadogenesis processes will be reported.