The commonly accepted Poisson model of clouds places droplets and raindrops in space and in time as evenly as randomness allows. This means than no ``blobs'' or ``patches'' are allowed on any length scale and statistical independence holds. While this applies to, for example, counting statistics of molecules in an ideal gas, counter-examples include bunching of photons and clustering of galaxies (both of which violate the statistical independence assumption).

Cloud droplets are suspended in turbulent air. Turbulence is a quick but not a thorough mixer so that ``blobs'' of high and low concentration remain. This implies correlations (lack of statistical independence) on some spatial scales and, therefore, super-Poissonian statistics (larger than the Poissonian variance of number density). This chain of reasoning depends only on the presence of turbulence. Hence, super-Poissonian distributions are likely to be ubiquitous and apply to all kinds of cloud particles and aerosols. Implications of such ``enhanced fluctuations'' might be rather wide-ranging: from droplet collisions and growth to radar meteorology and radiative transfer of visible and IR radiation.

Refreshments will be served