Abstract
The Millionshchikov hypothesis of quasi-normal distribution of fourth-order moments fails for convective conditions where the probability density functions of temperature and vertical velocity fluctuations are skewed. This is shown for aircraft and large-eddy simulation (LES) data, and new closures for fourth-order moments that take the skewness into account are suggested. These new closures are in very good agreement with the data.
Corresponding author address: Dr. Jörg Hartmann, Alfred Wegener Institute for Polar and Marine Research, Postfach 12 0161, Bremerhaven 27515, Germany. Email: jhartmann@awi-bremerhaven.de