| Abstract | This dataset contains the simulation output and analysis code used by Henry and Markowski (2026) for their article, "Microscale variability of convective storm forecasting parameters in a numerical simulation of a strongly sheared, turbulent environment." Output from a large-eddy simulation (LES) of a strongly sheared, supercell-favorable environment is used to quantify microscale variability - and the associated representativeness error - of common sounding-derived severe-weather parameters. Convective available potential energy varies relatively little, with departures from the mean (~3600 J kg-1) generally being less than 100 J kg-1. However, convective inhibition (CIN) departures of ~10 J kg-1 from the mean (~8 J kg-1) are common and potentially problematic for assessing convection initiation potential. Kinematic parameters tied to the surface layer [0-1- and 0-3-km storm-relative helicity (SRH) and the 0-6-km bulk wind difference] organize into streamwise streaks with ~500-m spacing and exhibit relatively large variability (e.g., approaching 150 m2 s-2 in the case of 0-1-km SRH). Pseudosoundings constructed by advecting balloon-like trajectories are unbiased relative to instantaneous profiles, but individual SRH errors can exceed 130 m2 s-2, underscoring the difficulty of inferring near-surface kinematics from a single drifting profile. Finally, 30 supercell simulations initialized with randomly selected instantaneous soundings from the LES all produce long-lived supercells, yet tornadolike vortex intensity spans nontornadic to EF4. In total, the results show that, even in the absence of mesoscale heterogeneity, boundary-layer turbulence alone can introduce nontrivial uncertainty in sounding-derived parameters and in idealized case-study simulations initialized from proximity soundings. |
