Updated & corrected: Rationalised sub-intervals used for plotting; 1957 had a July zero
While we’re waiting for some real snow, I’ve updated the season start chart (you might need a browser refresh to see the latest version):
Snow depths have been recorded weekly by Snowy Hydro Limited at Spencers Creek, midway between Perisher Valley and Thredbo, NSW, Australia, since 1954. The graph shows the season start-up, by plotting the percent of years with no snow at dates through April-May-June¹. As before, it’s no snow at the date, not by the date. Some years will have had snow earlier, but lost it.
There is a suggestion of a slightly later season start in more recent times, but not by much — by about two weeks at the 50th percentile, less elsewhere. That’s in stark contrast to the melt phase of our snow season, which now happens a whole month earlier than it did 50 years ago:
Notes
- The graph was made by running a 14-day window over the weekly resampled snow depth record and computing the percentage of zeros in that across the record lengths shown. Most years have two weekly depths in each 14-day window, both of which are counted in the percentage, which is plotted centred. The “whole record” curve is smoothed with a 5-day FWHM Gaussian; the others are unsmoothed.
Leave a Reply