Premium
Determination of non‐stationarity in the surface layer during the T‐REX experiment
Author(s) -
Večenaj Ž.,
De Wekker S. F. J.
Publication year - 2014
Publication title -
quarterly journal of the royal meteorological society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.744
H-Index - 143
eISSN - 1477-870X
pISSN - 0035-9009
DOI - 10.1002/qj.2458
Subject(s) - intermittency , series (stratigraphy) , parametrization (atmospheric modeling) , turbulence , similarity (geometry) , boundary layer , mathematics , statistical physics , surface (topology) , statistics , econometrics , computer science , meteorology , physics , geology , mechanics , paleontology , geometry , quantum mechanics , radiative transfer , artificial intelligence , image (mathematics)
Stationarity is a fundamental assumption in the statistical investigation of turbulence and the development of similarity functions that are widely used in surface‐layer parametrization schemes. Non‐stationary time series should therefore be removed from the analysis before assessing turbulence statistics used in similarity functions. Many approaches have been developed over the years to determine non‐stationarity of means and (co‐)variances, but there has been no systematic investigation of the differences and similarities between the approaches. In this article, we contrast several frequently used approaches, including two statistical tests to determine trends, the determination of a non‐stationarity ratio, and the determination of differences in turbulence statistics calculated for different averaging times. We apply these approaches to near‐surface time series of wind and temperature collected during the Terrain‐induced Rotor Experiment (T‐REX). Our results show that the degree of non‐stationarity varies considerably with the approach used. Investigation of the time series that are declared stationary by all the above approaches simultaneously reveals that in many cases such time series still show a behaviour indicative of intermittent turbulence, both for stable and unstable conditions. When these approaches are combined with an additional condition for the intermittency level, a rigorous approach for the detection of non‐stationarity is developed.