The formation of small-scale atmospheric vortices via horizontal shearing instability

Abstract

Motivated by recent high-resolution observations of small-scale atmospheric vortices along near-surface boundaries, this study presents a series of simulations that attempt to replicate the dynamics of the dryline and surrounding boundary layer with special emphasis on misocyclones. The first simulation was a real data case, initialized and forced through time-dependent lateral boundary conditions via analyses of temperature, moisture, and momentum from the 22 May 2002 IHOP dataset. The second series of simulations were barotropic runs, initialized with a north-south oriented constant vorticity shear zone and north-south periodic boundary conditions. The third series of simulations were baroclinic, where the shear zone also contained and east-west temperature gradient. The barotropic and baroclinic simulations had varying magnitudes of shear and shear zone widths (corresponding to differing initial vorticity values) across the runs. Additionally, several barotropic simulations were rerun with moisture included to assess preferred could formation regions. The real data simulation produced several misocyclones with characteristics consistent with those observed along near-surface boundaries in the atmosphere. Several of these misocyclones also had features resembling those observed in many laboratory studies and other numerical studies. Many of these features were also found in the barotropic simulations (i.e. instabilities developed into elliptical cores that precess, contain pressure perturbations in their centers, and evolve with cores connected by vorticity braids). To assess the instability mechanism, the results were compared to linear theory. Excellent agreement was found between predictions from linear theory in terms of wavenumber of maximum growth as a function of shear zone width and growth rate as a function of shear zone vorticity, suggesting to a very good first approximation, horizontal shearing instability (HSI) is responsible for the growth of initial small perturbations. It was also found that predictions of linear theory tend to extend well into the nonlinear regime. The baroclinic simulations were more complicated and allowed for tilting and stretching of vorticity not seen in the barotropic simulations. As the shear zones contract due to frontogenesis, vorticity increases, thus increasing the growth rates and the wavenumber of maximum growth. An attempt was made to model the contraction and apply a “modified linear theory” to the results, by allowing linear theory to have a time-varying shear zone width. This modified model provided excellent agreement with the simulated results in terms of growth rate and wavenumber of maximum growth. Finally, an attempt was made to assess preferred regions of cumulus formation by including moisture in the real data case and in several barotropic simulations. It was found that maximum updrafts and simulated cumuli tend to form along the periphery of cores and/or along the braided regions adjacent to the cores. Due to the important modulating effect of misocyclone development via HSI and subsequent moisture transport, cumulus spacing and size/depth was also dependent on the shear zone width and vorticity.