Scale-Dependent Droplet Clustering in Turbulent Clouds

Raymond Shaw

Department of Physics
Michigan Technological University

Abstract-
The most casual observer notices that atmospheric clouds are turbulent, yet we still struggle to understand the detailed role of turbulence in cloud physics. For example, current understanding of fundamental processes in clouds such as the collision and coalescence of droplets and the propagation of electromagnetic radiation, is based on the assumption that droplets are distributed in space in a perfectly random manner at small (e.g., sub-meter) scales. In fact, droplets in a cloud often do not follow what we might call the Poisson assumption, but instead are 'clustered' on various scales. Correlations in droplet positions can be caused by turbulent mixing of cloudy and clear air, by the inertial response of cloud droplets to fluid accelerations, or even by gravitational sedimentation in still air. To describe correlations at the small scales relevant to droplet-droplet interactions we set aside the notion of a continuous scalar variable, such as droplet number density, and instead adopt a language capable of describing a discrete, countable, scalar variable. The degree of spatial correlation is then quantified by the pair correlation function, and its scale dependence varies with such variables as droplet size, energy dissipation rate, and turbulence Reynolds number. Spatial correlations measured in turbulent clouds are observed to be strongest at centimeter scales and below and the pronounced clustering on these scales is consistent with the inertial clustering hypothesis.


GALCIT Home Page
2002-2003 Fluids Seminar Page


Maintained by: Michael Johnson
EMail: Michael Johnson
Last modified: September 10, 2002