In the United States, the upcoming autumnal equinox is marked on most calendars as the "first day of autumn." Similarly the solstices are commonly called the "first day of summer" and "first day of winter." However in most other languages -- and even in older discussions in English -- the solstices have names like "Midsummer's Day" and "Midwinter's Day."

Is there some sound reason (e.g. typical temperatures, asymmetrical sunrise/sunset times, thermal inertia, etc.) to consider a solstice as the first day of a new season, rather than the middle?

Also (perhaps less on-topic) is there any record of when the change in nomenclature took place in the U.S.? Are there similar discrepancies in other English-speaking countries or in other cultures?

  • 1
    $\begingroup$ "In many cultures, solstices/equinoxes traditionally determine the midpoint of the seasons, which can be seen in the celebrations called midsummer and midwinter. In this vein, the Japanese celebrate the start of each season with an occurrence known as Setsubun. The cumulative cooling and warming that result from the tilt of the planet become most pronounced after the solstices, leading to the more recent custom of using them to mark the beginning of summer and winter in most countries of Central and Northern Europe, as well as Canada, USA and New Zealand." en.wikipedia.org/wiki/Solstice $\endgroup$
    – f.thorpe
    Commented Sep 15, 2014 at 22:17
  • 5
    $\begingroup$ This question was asked in the English stack exchange. In fact... your question is really about language not science. english.stackexchange.com/questions/117183/… $\endgroup$
    – f.thorpe
    Commented Sep 15, 2014 at 22:18
  • $\begingroup$ Related: earthscience.stackexchange.com/q/2133/725 $\endgroup$
    – Gimelist
    Commented Dec 14, 2014 at 6:06

3 Answers 3


There are three main ways to reckon seasons:

  • Solar. The 3 months with the greatest insolation are designated summer, so the solstices and equinoxes fall in the middle of the season. In the Celtic calendar, summer started on 1 May; in the traditional Chinese calendar it is 5-7 May. The cultural 'Midsummer' and 'Midwinter' festivals reflect this.
  • Astronomical. The solstices and equinoxes are the starts of the seasons, with summer starting on about 21 June (it varies). Its use dates back at least to the Romans, and it's used in most western cultures today (apparently not in Russia, Australia, or New Zealand).
  • Meteorological. This is what professional meteorologists use (e.g. UK Met Office), and is based on the prevailing weather on land at mid-latitudes. Seasons are 3 months long and summer starts on 1 June.

I've tried to illustrate the relationships with insolation and temperature here:

Different ways to reckon seasons

There are some other ways too:

  • Ecological. Scientists who study the behaviour of organisms (hibernation, blooming, etc.) adapt to the local climate, sometimes using 6 seasons in temperature zones, or only 2 in polar and tropical ones.
  • Agricultural. This would centre around the growing season and therefore, in North America and Europe at least, around frost.
  • Cultural. What people think of as 'summer', and what they do outdoors (say), generally seems to line up with local weather patterns. In my own experience, there's no need for these seasons to even be 3 month long; When I lived in Calgary, summer was July and August (hiking), and winter was December to March (skiing). Here's another example of a 6-season system, and a 3-season system, from the Aboriginal people of Australia, all based on weather.

Why do systems with later season starting dates prevail today? Perhaps because at mid-latitudes, the seasonal lag means that the start of seasonal weather is weeks later than the start of the 'insolation' period. In a system with no heat capacity, there would be no lag. In systems with high heat capacity, like the marine environment, the lag may be several months (Ibid.). Here's what the lag looks like in three mid-latitude cities:

Seasonal lag. Licensed CC-BY-SA by EOEarth

The exact same effect happens on a diurnal (daily) basis too — the warmest part of the day is often not midday (or 1 pm in summer). As with the seasons, there are lots of other factors too, but the principle is the same.

These aren't mutually exclusive ways of looking at it — there's clearly lots of overlap here. Cultural notions of season are surely rooted in astronomy, weather, and agriculture.

  • $\begingroup$ How did you create the illustration of the 3 main ways of defining a season? $\endgroup$
    – G. Gip
    Commented Sep 26, 2016 at 21:57
  • $\begingroup$ Upvote: this makes my point much better than I did: the insolation is highest on the solstice, not the temperature. $\endgroup$
    – user967
    Commented Aug 6, 2018 at 17:25

On the summer solstice, we receive more heat from the Sun than on any other day. The next day, we receive less heat from the Sun, but we still receive more heat than we lose. Thus, the temperature increases.

At some point (roughly August 2nd, "Midsummer's Day"), we receive as much heat as we lose. That represents the maximum temperature. After that, we lose more heat than we receive and it gets colder.

Mathematically, the summer solstice represents the maximum of the derivative of temperature, not the maximum temperature itself.

  • $\begingroup$ As a maths nerd, this is such a great answer. $\endgroup$
    – naught101
    Commented Oct 10, 2014 at 3:07
  • 1
    $\begingroup$ Note I've had it explained to me recently (somewhere I thought around here, though I can't find it, but it was convincing) that the seasonal lag is actually basically caused by the delay caused by heat storage... and is thus basically 99% the ocean's fault. It being cool still in early summer, keeps us cooler than we should be. So for a while it is still increasing in stored heat, until it reaches the point where it stops gaining. And if it weren't for the oceans, the maximum temperature would be very near after solstice. $\endgroup$ Commented Aug 6, 2018 at 7:33
  • $\begingroup$ And I'd suggest max derivative is probably more in Mar or Apr (though my mid-latitude bias may be showing). Most of the hemisphere has rapidly shifted out of winter well before solstice, many gaining upwards of 30 degrees... and most only gain another 10-20 degrees max after. (usclimatedata.com/climate/barrow/alaska/united-states/usak0025 suggests even in AK, the max derivative may indeed be May). Don't feel bad, I certainly thought it was more straightforward math than fiddling effect. Proves even a meteorologist (one who used to give a quiz on such misconceptions!) can be wrong! $\endgroup$ Commented Aug 6, 2018 at 7:39
  • 1
    $\begingroup$ @JeopardyTempest You are correct. I should have said "the maximum heat received", which isn't the same thing: the hotter it is, the faster the Earth loses heat, so amount of heat received isn't proportional to temperature change. And, if you want to be picky, I'm assuming a cloudless, hazeless day, where the Sun is up for the longest period of time in the year and at the highest elevation with nothing blocking its rays from the ground. $\endgroup$
    – user967
    Commented Aug 6, 2018 at 17:18

This all depends on the context in which the word is used.

From an astronomical perspective, the solstices are midsummer and midwinter, and the equinoxes are the middle of autumn and the middle of spring.

However, the seasons, as perceived in terms of weather, lag somewhat behind the astronomical seasons (presumably this is due to the thermal mass of the planet?). For example, most people of my acquaintance think of northern-hemisphere summer as occupying approx. June/July/August, putting the June solstice near the start of it. Therefore, colloquially, the solstices and equinoxes seem to roughly match up with the starts of the seasons, and hence are often treated that way.

Why does this apply to the equinoxes more than the solstices? (citation needed - but purely from personal observation, many people think of the summer solstice as "midsummer" but the spring equinox as "the start of spring"). I'm speculating, but my guess is that midsummer and midwinter are really obvious to anybody keeping an eye on the sky (when the sun reaches its greatest or least north/south extent, or "when the days start getting shorter", or whatever). The equinoxes are slightly harder to judge, and so aren't noted in quite the same way.

  • 2
    $\begingroup$ Meteorological seasons (NH) start on 1 Mar (spring), 1 June (summer), 1 Sep (autumn) and 1 Dec (winter). $\endgroup$
    – casey
    Commented Sep 17, 2014 at 19:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.