On the Micrometeorology of the Southern Great Plains 1: Legacy Relationships Revisited

Abstract

Data from a 32-m tower located near Ocotillo, Texas ($32.12050^{\circ }$N; $101.37555^{\circ }$W), provide an opportunity to examine the relevance of standard micrometeorological flux–gradient formulations to observations made in an area characteristic of a large portion of the central USA, within the Southern Great Plains. Comparison with data obtained at a greater height (80 m) reveals that the velocity distributions change substantially between the lower set of observations and the upper, with the former being constrained at the low wind-speed end. In the early morning, sensible heat-flux divergence correlates well with the measured rate of change of temperature with time within the surface layer of air sampled by the tower, but this association disappears when the depth of the mixed layer extends beyond the reach of the tower. As in the case of all previous examinations of flux–gradient relationships, the overall dependence of the dimensionless wind and temperature gradients $\phi _\mathrm{m}$ and $\phi _\mathrm{H}$ on stability is characterized by considerable scatter, with the familiar relationships best describing the average. For conditions of stable stratification, there is indeed the expected close proximity of $\phi _\mathrm{m}$ and $\phi _\mathrm{H}$, however, describing either $\phi _\mathrm{m}$ or $\phi _\mathrm{H}$ in terms of the classical stability index $Z/L$ (where $Z$ is the height above the zero plane and L is the Obukhov length scale of turbulence) then appears questionable because the dependence of $\phi _\mathrm{m}$ on the measured sensible heat flux is not always single-valued, especially near the surface. For unstable stratification, support is found for the conclusions of early workers that free convection initiates at about $Z/L \approx -0.03$, and that the general behaviour is then compatible with the concept of a moving air mass from which momentum is continuously extracted, embedded within freely convective cells. It is concluded that legacy descriptions of the relationships between fluxes and gradients apply to averages that might occur rarely, that a dominant factor is likely the chaotic nature of the processes that control the variables considered in these relationships, and that the net consequence of the original randomness is that the levels of predictability theoretically attainable might never be realized in practice.