i'm still holding out hope on a winter cancellation in the very south-western reaches of ontario. i think it's a definitional problem, though: what defines winter in an empirical sense? is it temperature dependent? because it's not unknown in canada to have short, disconnected bursts of cold weather at any time of the year. so, then it should be sustained. see, i'm not convinced we can get a sustained level of sufficiently cold weather here this year to qualify, but what is the temperature, anyways? zero seems reasonable. but, then is it zero for the overnights or zero for sustained highs? i don't see a sustained period of subzero highs here in windsor, so by that metric it seems doubtful.

but, it's probably not the best to measure it by sustained highs, as we're dealing with a continuous rather than a discrete problem (quantizing things isn't always useful, and is pretty much always really just a shorthand guess when it is). so, if we get a sustained continuous period of near-zero temperatures with only moments above freezing, that may be close to qualifying as winter. but, then how near-freezing is it, really, and what are the actual observable changes? that happens sometimes in november, or even october, and we consider that to be fall.

i think a better metric is probably related to snow cover. that is a clear indication of winter. and, i think that's what i'm holding out on not happening at all this year in windsor - that we're able to avoid any serious cold snap below -10, and that we're able to avoid any snow cover at all. if that happens, winter will have been cancelled for the year.

www.theweathernetwork.com/news/articles/january-pattern-change-to-make-mildness-a-distant-memory/61773/