Anisotropy and multifractal analysis of turbulent velocity and temperature in the roughness sublayer of a forested canopy

TL;DR

This study investigates the anisotropy and multifractal properties of turbulent velocity and temperature in the roughness sublayer over a forest, revealing how surface effects influence turbulence characteristics at different heights.

Contribution

It introduces a combined analysis of anisotropy and multifractality in turbulence data over a forested canopy, highlighting the effects of surface roughness and thermal stratification.

Findings

Return to isotropy occurs at larger scales higher in the RSL.

Largest turbulent scales become more anisotropic with height.

Multifractal exponents' anisotropy decreases with extended-self-similarity.

Abstract

Anisotropy and multifractality in velocity and temperature time series sampled at multiple heights in the roughness sublayer (RSL) over a boreal mixed-coniferous forest are reported. In particular, a turbulent-stress invariant analysis along with a scalewise version of it are conducted to elucidate the nature of relaxation of large-scale anisotropy to quasi-isotropic states at small scales. As the return to isotropy is linked to nonlinear interactions and correlations between different fluctuating velocity components across scales, we study the velocity and temperature time series by using multifractal detrended fluctuation analysis and multiscale multifractal analysis to assess the effects of thermal stratification and surface roughness on turbulence in the RSL. The findings are compared so as to quantify the anisotropy and multifractality ubiquitous to RSL turbulent flow. As we go up in the RSL, (a) the length scale at which return to isotropy commences increases because of the weakening of the surface effects and (b) the largest scales become increasingly anisotropic. The anisotropy in multifractal exponents for the velocity fluctuations is diminished when we use the extended-self-similarity procedure to extract the multifractal-exponent ratios.