Under unstable stratification, the Reynolds shear stress $-\overline{u^{\prime }{w}^{\prime }}$ in the atmospheric surface layer (ASL) reflects a multiscale superposition in which large-scale motions can contribute appreciably, motivating a scale-resolved quantification beyond standard similarity scaling. Using multiheight SLTEST velocity measurements under unstable conditions, we obtain an additive scale decomposition of $-\overline{u^{\prime }{w}^{\prime }}$ from momentum-flux cospectra and quantify cumulative contributions with an ogive $C\left(k\right)$, where $k$ is the streamwise wave number. Interpreting $C\left(k\right)$ as an empirical cumulative distribution in $lnk$, we parametrize it with a lognormal cumulative distribution function, yielding a compact two-parameter description $(\mu ,\sigma )$ of the dominant contributing scale and the breadth of scale allocation. Two indicators isolate the key mechanisms: the cumulative contribution from large scales and a counterstress fraction ${\eta }_{\mathrm{neg},1}$ that measures cospectral contributions opposite in sign to the net stress. Under weakly unstable conditions, normalized premultiplied cospectra collapse when plotted against the nondimensional wave number $kz$, where $z$ is the measurement height, with a peak at $kz\approx 0.5$, and the large-scale contribution increases with relative height, indicating enhanced large-scale influence aloft. For weak-cancellation records (${\eta }_{\mathrm{neg},1}<0.1$), ogives are accurately captured by the lognormal form, and $(\mu ,\sigma )$ reproduce the observed cospectral morphology across heights. With increasing instability, the dominant contribution shifts to larger wavelengths, the large-scale flux fraction increases, and scale-domain cancellation strengthens, with larger magnitudes and stronger sensitivity at higher relative heights. A variable-interval time-averaging-based burst analysis provides a consistent time-domain counterpart: events become sparser while the stress magnitude concentrates within fewer intervals, and a time-domain cancellation metric increases concurrently. These results establish a low-dimensional height-stability parametrization of momentum-flux scale allocation and show that unstable ASL transport is organized by coupled large-scale enhancement, strengthened cancellation, and increased intermittency.

Close

PrevNext

View figure in article

• Flow instability
• Structure & turbulence of boundary layers
• Turbulence