Scale Localization of Cloud Particle Clustering Statistics

Abstract

ecent work has examined the spatial distribution of droplets within a cloud. Experimentally, most studies analyze interevent times from static probes flown linearly through a cloud, allowing the spatial information to be conveyed through a time series of particle detections. Analysis of these data has shown unequivocally that most clouds have nontrivial spatial structure. There is still debate as to which of many different possible statistical descriptions is most appropriate for characterizing this spatial structure and for identifying spatiotemporal scales of physical significance from the data record. This paper seeks to clarify some pervasive misunderstandings and to outline more carefully the range of validity for several of these statistical tools. Simulations are used to explore the relative scale-localizing capabilities of various commonly used statistical tools.