Abstract
During the winters of 1961–62 and 1963–64, measurements of light scattering by falling snow provided a continuous record of snowfall rates on the campus of McGill University in Montreal. This permitted an analysis of the time variations of intensity during a snowstorm, with a resolution not possible when a heated tipping bucket is the measuring device. It was found in the course of this study that a major portion of the variance in snowfall rate measured at the ground during a single storm can be explained in terms of a simple Markov process, in which the snowfall rate at each time interval appears to be affected by its value one time step before, but is independent of the rate more than one time step before. Under certain circumstances, this may provide a method of predicting total snowfall accumulations.

This publication has 0 references indexed in Scilit: