A Study of Spectra, Structure and Correlation Functions and Their Implications for the Stationarity of Surface-Layer Turbulence

Abstract

The related concepts of stationarity and the existence and values of integral timescales are central to the ability of analyzing micrometeorological data within theframework of Monin–Obukhov similarity theory. Not only does the theory strongly hinge on the stationarity assumption, the estimation of turbulence moments and their accuracies are dependent on the values of the correspondent integral time scales. In spite of the general importance of these concepts, there are relatively few studiesconcerned with them. Moreover, although each turbulence variable has its ownintegral scale, this fact is often overlooked when numerical values are estimated.In this work we study three daytime events of surface inversion formation, that is,events where a nonstationary period is clearly present. Our analysis reveals alow-frequency component in the temperature data that is not totally removed bya simple (but often used in turbulence data analysis) first-order recursive filter.This component has to be filtered out in the frequency domain, after which we areable to recover similarity between temperature and humidity statistical descriptors(in this case, the structure function). After applying a simple criterion to estimatenumerical values of the integral time scales, we are able to assess the relationshipsbetween the existence of integral scales and the stationarity of the correspondingprocess. Finally, we find out that in the case of second-order moments the Sarmanovtheorem does not always apply. The implications for accuracy estimates of thesemoments are then briefly discussed.