so, is it winter yet?

well, there's no snow. i was just outside on the grass, and the ground is not really frozen yet - although it's a little firm. that should soften back up entirely in the next few days. we haven't yet had an extended period of subzero temperatures. we briefly made it down to -10 overnight on tuesday morning. but, that's not uncommon, here, in october. and it was only for a few hours.

they're forecasting three days of cold weather starting on monday, but the numbers have been revised upwards repeatedly. a few days ago, they were forecasting a low on tuesday of -16. that's been changed to -11. we'll see if we actually get there. but it looks like we'll see the same thing: a few hours around -10. that's not an extended period below freezing. "normal" nighttime lows here this time of year are around -10. we're barely hitting that, this year, at all. we may end up with something like ten hours below -10 for the whole year.

i'm at this point willing to suggest that we've merely had an extended autumn, and that it looks set to transition directly into spring. like i say: there's only so much winter can do to assert itself, if it can't manage to freeze the ground or establish a snow cover.

i acknowledge it's largely a definitional issue. some people will claim that winter starts at the moment of the first flurry, but then you're extending winter most years from september to may - and that's outrageous. winter in canada requires more stringent conditions: an extended period of subzero temperatures that is at least able to freeze the ground and is accompanied by a visible snow cover. that seems unlikely. so, winter still appears to be cancelled down here, in this little southern crook of canada - although i know this is not the case elsewhere in the country.

"The Swedish meteorological institute (SMHI) define winter as when the daily mean temperatures go below 0 °C (32 °F) for five consecutive days"

i think that this is pretty good, and i do think we've nearly met this definition. but, what if you have four consecutive days three times, each separated by several days? that might be what happens, here.

i also think you need a bit more: you also need a frozen ground. that should be nearly automatic after five actual cold days, but not if the temperature is just hovering around zero, which is really what we've seen - we only had a few hours of actual cold, and will only get a few hours more next week. so, maybe a better mean temperature is -2.5, being halfway to -5 - and then we're not there.

and, if there are further ambiguities, the question of snow cover must be the final determinant that trumps everything. if we end up in this highly ambiguous position, but there's no snow? no winter.

maybe a better definition is a 20 day period where the total average temperature - not an average of the averages, just a flat average, hourly - is below zero. but, then you wouldn't know you had winter until it was done - or at least halfway done.