Sensitivity of tornadogenesis in very-high-resolution numerical simulations to variations in model microphysical parameters

Idealized numerical simulations are performed using the Advanced Regional Prediction System (ARPS), in which convection is initiated using a sounding associated with the May 20, 1977 tornadic supercell that occurred near Del City, Oklahoma. Thirteen simulations are conducted at 1 km and seven at 100 m horizontal grid spacing, using various values of the drop size distribution (DSD) intercept parameters for rain, snow, and hail, and of hail densities that are part of a single-moment ice microphysics scheme. Influence of these parameters on storm development and tornadogenesis is analyzed.

It is found that the cold pool intensity and low-level storm dynamics are most sensitive to rain and hail intercept parameters, with little sensitivity to snow intercept parameter or hail density. Using DSD parameters that favor small hydrometeors, particularly small raindrops, leads to storms with a more intense cold pool, due to mainly the enhanced evaporative cooling over a larger area. Using DSD parameters favoring larger hydrometeors produce relatively weaker cold pools. Two 100 m simulations produced long-lived low-level tornadic circulations, with maximum intensity of F2 and a duration of about 4 minutes for the control (default) microphysical settings and a duration of 10 minutes for a case of reduced rain intercept parameter. In the cases with strong cold pools, the storm updraft is tilted rearward above the surface gust front, which is located several kilometers ahead of the main mid-level updraft. Cases with weaker cold pools usually feature strong, sustained, vertical updrafts that are located more directly over the gust front — a scenario that is found to be more favorable for tornadogenesis.

Additional diagnostic analysis of the simulated tornado will be also be presented.