Field observations validate LES-derived scaling of vertical-velocity variance during the afternoon transition of the CBL

The convective velocity scale is a cornerstone for describing vertical-velocity variance in the bulk of the convective boundary layer driven by surface heating, and is widely used in observations, theory, and operational meteorological models. This scaling relies on the quasi-equilibrium assumption that the surface heat flux varies slowly compared to the adjustment time scale of the large scale convective eddies (or the eddy turnover time). This quasi-equilibrium assumption typically holds over land, from late morning to early afternoon, but breaks down in the late afternoon. Recent idealized large-eddy simulations (LES) of free convection reported departure from the classical convective scaling due to quasi-equilibrium breakdown, and identified the relevant parameters to describe the vertical-velocity variance during the late afternoon transition. Both the characteristic time scale of the surface heat flux decay, τ = ΙH^-1dH/dtΙ^-1 (H is the surface heat flux), and the convective eddy turnover time, t_* = z_i / w_* (z_i is the boundary-layer depth and w_* is the convective velocity scale) should be taken into account. During the early afternoon transition when the parameter r = τ /  t_* is larger than 1, the quasi-equilibrium assumption is satisfied. The departure starts during the late afternoon transition when r approaches 1, and a new scaling regime emerges when r is smaller than 1. In this study, we evaluate these scaling predictions using extensive field observations spanning 264 days. Despite substantial day-to-day variability (due to random errors in the estimation of turbulence statistics or the sensitivity of the turbulent flow to unknown local disturbances not included in the analysis), averaging across multiple days reveals a good agreement between observations and the LES-derived scaling, supporting its validity for describing vertical-velocity variance in this regime.