A Definitive Approach to Turbulence Statistical Studies in Planetary Boundary Layers

Abstract

A method for comprehensive statistical data analysis is presented. It includes a procedure for selection of segments within a data record, where data can be considered stationary, allowing the computation of statistical parameters (from data over selected segments) and their confidence intervals (estimating the statistical reliability of the parameters). The method is illustrated by analysing the vertical component of the wind velocity data collected by research aircraft flights over Lake Michigan during Project LESS (Lake Effect Snow Studies). The results indicate that for skewness and kurtosis, the requirements for acceptable ?averaging distances? are less limiting than the required distances for corresponding third- and forth-order moments.