Based on the most recent decade being pretty skimpy, any linear regression trend on White Xmas around here is going to be negative.
But if you throw another 2000s type decade on the dataset, then it might change it…we had a really brutal stretch of lean White Xmas in the 1930-early 1950s period…so if you started from there and ended in a year like 2010, you’d have a positive trendline.
I don’t think it’s going to become more common than the 20th century mean though…a huge component of a White Xmas is keeping existing snowpack from melting out before that date and warmer temps simply make that harder. Even if you partially offset it with somewhat more frequent sizable events from working with more moisture.