On the statistical distribution of dust devil diameters

Dust devil diameters, like many features in nature, have skewed distributions. I summarize, and present in a unified manner, literature values of dust devil optical diameters from seven terrestrial surveys, three rover-based surveys on Mars and two orbital surveys on Mars. The problems of appropriately treating these data are analogous to those for impact craters, and similar display and binning approaches are suggested. Remarkably, the Mars dust devil population remains better-known than Earth’s. The theoretical justifications for possible log-normal, and (truncated) exponential and power-law descriptions of dust devil properties are discussed, and the challenges of discriminating between these candidate distributions with finite (and often, coarsely-binned) observation sets are noted: the best-sampled datasets so far appear well-described by power laws. Data required for advances in model discrimination are discussed: data binned in four or fewer ranges are useless for this purpose. Caution must be exercised in applying the notion of an ‘average’ dust devil and in calculating population-integral properties such as dust flux.

► Presents a unified treatment of 12 different dust devil surveys on Earth and Mars.
► Examines binning and counting statistics, and approaches for graphical presentation.
► Considers theoretical basis, and quality, of power-law and exponential fits to data.
► Shows the importance of power-law characteristics in calculating dust raising.
► Assesses the statistical requirements for future surveys.