Intermittent transitions between turbulent and non-turbulent states are ubiquitous in the stable atmospheric surface layer (ASL). Data from two field experiments in Utqiagvik, Alaska, and from direct numerical simulations are used to probe these state transitions so as to (i) identify statistical metrics for the detection of intermittency, (ii) probe the physical origin of turbulent bursts, and (iii) quantify intermittency effects on overall fluxes and their representation in closure models. The analyses reveal three turbulence regimes, two of which correspond to weakly turbulent periods accompanied by intermittent behavior (regime 1: intermittent, regime 2: transitional), while the third is associated with a fully turbulent flow. Based on time series of the turbulence kinetic energy (TKE), two non-dimensional parameters are proposed to diagnostically categorize the ASL state into these regimes; the first characterizes the weakest turbulence state, while the second describes the range of turbulence variability. The origins of intermittent turbulence activity are then investigated based on the TKE budget over the identified bursts. While the quantitative results depend on the height, the analyses indicate that these bursts are predominantly advected by the mean flow, produced locally by mechanical shear, or lofted from lower levels by turbulent ejections. Finally, a new flux model is proposed using the vertical velocity variance in combination with different mixing length scales. The model provides improved representation (correlation coefficients with observations of 0.61 for momentum and 0.94 for sensible heat) compared to Monin–Obukhov similarity (correlation coefficients of 0.0047 for momentum and 0.49 for sensible heat), thus opening new pathways for improved parametrizations in coarse atmospheric models.

The probability density function p(k) of the turbulent kinetic energy k is investigated for diabatic atmospheric surface layer (ASL) flows. When the velocity components are near-Gaussian and their squared amplitudes are nearly independent, the resulting p(k) is shown to be gamma-distributed with exponents that vary from 0.8 to 1.8. A non-linear Langevin equation that preserves a gammadistributed p(k), but allows linear relaxation of k to its mean state, is proposed and tested using multiple ASL data sets. The three parameters needed to describe the drift and non-linear diffusion terms can be determined from the ground shear stress and the mean velocity at height z. Using these model parameters, the Langevin equation reproduces the measured p(k) with minimal Kullback-Leibler divergence.

Particularly challenging classes of heterogeneous surfaces are ones where strong secondary circulations are generated, potentially dominating the flow dynamics. In this study, we focus on land‐sea breeze circulations (LSBs) resulting from surface thermal contrasts, in the presence of increasing synoptic pressure forcing. The relative importance and orientation of the thermal and synoptic forcings are measured through two dimensionless parameters: a heterogeneity Richardson number (measures the relative strength of geostrophic wind and convection induced by buoyancy), and the angle α between the shore and geostrophic wind. Large eddy simulations reveal the emergence of various regimes where the dynamics are asymmetric with respect to α. Along‐shore cases result in deep LSBs similar to the scenario with no synoptic background, irrespective of the geostrophic wind strength. Across‐shore simulations exhibit a circulation cell that decreases in height with increasing synoptic forcing. However, at the highest synoptic winds simulated, the circulation cell is advected away with sea‐to‐land winds, while a shallow circulation persists for land‐to‐sea cases. Scaling analysis that relates the internal parameters Qshore (net shore volumetric flux) and qshore (net shore advected kinematic heat flux) to the external input parameters results in a succinct model of the shore fluxes that also helps explain the physical implications of the identified LSBs. Finally, the vertical profiles of the shore‐normal velocity and shore‐advected heat flux are used, with the aid of k‐means clustering, to independently classify the LSBs into four regimes (canonical, sea‐driven, land‐driven, and advected), corroborating our visual categorization.This article is protected by copyright. All rights reserved.