Premium
Statistical Study of Atmospheric Turbulence by Thorpe Analysis
Author(s) -
Zhang Jian,
Zhang Shao Dong,
Huang Chun Ming,
Huang Kai Ming,
Gong Yun,
Gan Quan,
Zhang Ye Hui
Publication year - 2019
Publication title -
journal of geophysical research: atmospheres
Language(s) - English
Resource type - Journals
eISSN - 2169-8996
pISSN - 2169-897X
DOI - 10.1029/2018jd029686
Subject(s) - log normal distribution , turbulence kinetic energy , dissipation , middle latitudes , turbulence , atmosphere (unit) , meteorology , radiosonde , atmospheric sciences , environmental science , troposphere , latitude , statistical physics , physics , statistics , mathematics , thermodynamics , astronomy
The variability in turbulence is a crucial topic in the lower atmosphere. This study aims to validate the credibility of Thorpe analysis in the statistical study of atmospheric turbulence by applying this method on high‐resolution radiosonde data that are a composite of seven sites at midlatitudes from January 2012 to December 2016. The spatiotemporal variability of turbulent energy dissipation rate is also addressed from 2.4 to 30 km, thereby exhibiting remarkable spatial inhomogeneities and well‐defined seasonal variations. If the ratio between Ozmidov and Thorpe lengths (i.e., c 2 ) is a lognormal distribution, then energy dissipation rate can be simulated by the Monte Carlo method. The approximate proportionalities between the energy dissipation rates revealed by a constant c 2 and a lognormally distributed c 2 that is based on the results from Schneider et al. (2015) exist at middle, low, and high latitudes. At midlatitudes, energy dissipation rates by the simulated cases are quantitatively consistent with the radar results. These results suggest that, even for a highly variable c 2 , the Thorpe analysis can be statistically applied to the stably stratified free atmosphere, and the optimal value of c 2 is 0.54 at all heights and latitudes. We also speculate that the statistical application of a Thorpe sort is acceptable when c 2 simply follows a lognormal distribution.