Open AccessStatistics and Parameterizations of the Effect of Turbulence on the Geometric Collision Kernel of Cloud Droplets
Author(s) -
Charmaine Franklin,
Paul Vaillancourt,
M. K. Yau
Publication year - 2007
Publication title -
journal of the atmospheric sciences
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.853
H-Index - 173
eISSN - 1520-0469
pISSN - 0022-4928
DOI - 10.1175/jas3872.1
Subject(s) - turbulence , physics , radius , dissipation , collision , range (aeronautics) , mechanics , turbulence kinetic energy , collision frequency , flow (mathematics) , cluster analysis , statistical physics , computational physics , classical mechanics , statistics , mathematics , thermodynamics , plasma , materials science , computer security , quantum mechanics , computer science , composite material
Collision statistics of cloud droplets in turbulent flow have been calculated for 12 droplet size combinations in four flow fields with levels of the eddy dissipation rate of turbulent kinetic energy ranging from 95 to 1535 cm2 s−3. The flow fields were generated by using a direct numerical simulation technique and large numbers of droplets were explicitly tracked through the flow field for each experiment. The effect of turbulence on the collision kernel increases with both increasing radius ratio and eddy dissipation rate. These increases range from fairly modest values to almost 10 times the gravitational geometric collision kernel. The two physical processes responsible for these increases are the radial relative velocities and the preferential concentration or clustering of the droplets. The radial relative velocities increased by up to 3 times the corresponding gravitational value and the greatest increase in the clustering, as measured by the radial distribution function, is 4.5 times the value for a random distribution as for the gravitational case. Parameterizations have been developed for the effect of turbulence on the radial relative velocities and the clustering of the droplets. These models reduce the average root-mean-squared errors in the existing velocity parameterization of Saffman and Turner and Wang et al. by 32% and the clustering parameterization of Zhou et al. by up to 58%.