Effects of horizontal and vertical grid spacing on mixing in simulated squall lines and implications for convective strength and structure

Abstract

The sensitivity of an idealized squall line to horizontal Δh and vertical Δv grid spacing is investigated using a new approach. Simulations are first performed at a horizontal grid spacing of 1 km until the storm reaches its mature stage. The model output is then interpolated to smaller (and larger) grid spacings, and the model is restarted using the interpolated state plus small thermodynamic perturbations to spin up small-scale motions. This framework allows an investigation of the sensitivity of the storm to changes in Δh without complications from differences in storm initiation and early evolution. The restarted simulations reach a quasi steady state within approximately 1 h. Results demonstrate that there are two Δh-dependent regimes with the transition between regimes occurring for Δh between 250 and 500 m. Some storm characteristics, such as the mean convective core area, change substantially for Δh >250 m but show limited sensitivity as Δh is decreased below 250 m, despite better resolving smaller-scale turbulent motions. This transition is found to be independent of the chosen Δv . Mixing in the context of varying Δh and Δv is also investigated via passive tracers that are initialized 1 h after restarting the simulations (i.e., after the spin up of small-scale motions). The tracer field at the end of the simulations reveals that entrainment and detrainment are suppressed in the simulations with Δh ≥ 500 m. For decreasing Δh, entrainment and detrainment are substantially more important, limiting the flux of low-level tracer to the upper troposphere, which has important implications for modeling studies of convective transport from the boundary layer through the troposphere.