SMHI/FMI temperature-based seasonal classification

Lommaren  Swedish summer definitions are better optimized for Sweden's own climate, which is rather different from that of the Bay Area.

The seasons are defined as follows:

Spring: 7 consecutive days with means >0'C
Summer: 5 consecutive days with means ≥10'C
Fall: 5 consecutive days with means <10'C
Winter: 5 consecutive days with means ≤0'C

The beginning of each season retroactively gets assigned to the first day in the consecutive streak.

The reason SMHI defines seasons to begin with is to get a general sense of whether seasons started early or late, and has a lot to do with everyday people who wonder whether, say, spring arrived early 2014 or not. Unlike USA which has a massive array of different climates, Sweden is relatively homogenous, enough so that a unified definition is relatively manageable. 

In Umeå, all of these definitions work well except fall.
Spring in Umeå:
7 consecutive days above 0'C tends to coalign neatly with the thawing away of most or all of the snow, and lots of flowers in the flowerbeds, and apart from the outlying freak February in 2015, it's never happened early enough that a major backlash of say 5 or 7 consecutive winter days have come afterwards. Sure, once spring is sprung, there are always going to be backlashes, but if late April sees a massive snow storm it's hardly fair to say spring started after that.

Summer in Umeå: 5 consecutive days with means above 10'C coalings well with everything starting to be green and usually happens mid-May to early June. 5 consecutive "summer days" means a week of sunny days with 15-25'C highs with everybody walking around in shorts and t-shirt and sunbathing all day, and everything is green and everyone's got a summery vibe. Then even if June is really cold and gloomy, it feels more like shitty summer rather than spring because everything has already sprung and you've already got into summer mood.

Fall in Umeå: Here's where it gets tricky. Whereas summer is always going to start with a sunny heatwave, it's probably going to end with a rainy and cloudy low pressure, and so even if there's a week of 13/9'C and rain in late August, it's still going to count as summer because the mean is above 10'C. My suggestion here would be to base the summer definition upon highs rather than lows (maybe 5-7 consecutive days with highs above 15-17'C for the start of summer and conversely 5-7 consecutive days with below 15'C highs for fall.

Winter in Umeå: 5 consecutive days below 0'C usually coalligns well with my perception of winter. Usually it will mark a prolonged snow cover, and lakes will freeze over. Sometimes it does happen prematurely in early November or possibly late October which can then be followed by at least a month of very mild and drizzly fall-like weather. It certainly feels like winter after those 5 "winter days", but after enough mild days, you do forget that it was winter. I would suggest that the winter pronouncement get retroactively removed if, after winter's been declared, there are 10 consecutive days of fall.

The definitions for Sweden as a whole:

Sweden is more homogenous than, say, USA, but due to spanning over 15 latitudes, any national definition is going to be a compromise. For example, most recent years, southernmost Sweden has gone directly from fall to spring without ever having a winter. And conversely, some stations in the far north have some years gone straight from spring to fall without ever having a summer. A national definition would work pretty perfectly in a country like Estonia where all climates are virtually the same. It can't for Sweden; the extreme ends are always going to suffer. Also, the fall definition is ridiculous. It's not summer in October in southernmost Sweden. The definition of a summer day should be based off of highs and not lows. Means work for the other seasons, although preferably, there should be a work-around for winter-less years in the south. I don't know how exactly, though.

How is a 10C mean summer like though. Our Aprils are like that..

No I’m not having that. Even for cold climates in Sweden that seems like a poor system considering the averages. I wouldn’t consider summer starting with 5 days of 15C means

Only problem with Finland's system is it isn't automized. It requires human judgement.

For comparison, in Estonia four additional inter-seasons have been defined to more accurately describe the transitions around winter and take care of some of the ambiguity.

Spring-winter: the time of melting snow when temps start to crawl above freezing and snow thickness decreases.
Early spring: the time after the snow has melted but there's no vegetation activity.
Spring: starts with first vegetation activity which coincides with average temperatures permanently staying above 5°C.
Summer: average temperature permanently above 13°C.
Autumn: average temperature permanently below 13°C.
Late autumn: average temperature permanently below 5°C, vegetation activity ceases.
Pre-winter: the time after first snow or the first ice day and the start of a permanent snowpack and/or permanently below freezing temperatures. This can drag on for weeks or be skipped entirely if the first snow is permanent.
Winter: permanent snowpack and/or below freezing temperatures.

I think a seven day period is used for the "permanent" part but EMHI is quite secretive and I couldn't find a source on that right now. I do remember reading that seasons can flip back and forth or be skipped entirely if warmth or cold persists long enough.

From earlier Shoutbox conversation


-

-

-

-

-