Trends in Northern Hemispheric Snow Presence
Jon Woody, Mississippi State University
This talk develops a mathematical model and statistical methods to analyze snow presence/absence. The methods are applied to satellite based observations which are part of the Climactic Data Record during the 1967-2020 time period. A two-state Markov chain modeling snow absence/presence is equipped with periodic dynamics to account for seasonality. The model allows one to characterize the autocorrelation structure of the 0-1 data. Uncertainty estimates permit statistical testing of trends. Among trustworthy grids, snow presence is seen to be declining in almost twice as many grids as it is advancing.