NavList:

Paul Saffo, you wrote:
"I thus would argue that the Solstice is nothing other than the Solstice. It is not the start of summer, nor mid-summer."

Yes, that's fair. Defining the start of summer and winter as the solstices and the start of spring and autumn/fall as the equinoxes is traditional and semi-official in many parts of the world, but it's arbitrary. So why is it so popular? I think mostly, and paradoxically, it's because those dates don't really work, and as a result they give us yet another way to have small-talk about the weather. How many times in the past few years have you heard someone say "wow, it's so cold, and it's still two weeks until winter!" or in California it's more likely to be "ugh, it's so hot, and still two weeks to summer!" People enjoy calling out the weather for misbehaving. In addition the astrological seasons (they are more astrology than astronomy) have the advantage of being precisely timed. It's the sort of science that the "average schmo" can appreciate. Schmo gets a cheap thrill from saying, "Spring just began ten seconds ago!"

So what should we use for the seasons instead? We have to start with some common ground, so let's agree (for now) that the seasons should be three months each, dividing the year up into quarters. They should be based in some approximate way on the weather, and they should try to align in some fashion with the traditional solar seasons. 

In the northern tier of the continental US, if you look at average climatology (daily records of mean high and low temperatures), you'll find that the hottest day of the year averages about July 21 and the coldest around January 21. In other words, the "middle" of summer and winter by that standard is about a month after the solstices, on a broad average. By our normal accounting where summer and winter start on the solstices, the "middle of the two seasons would be two weeks after those extreme (average) temperature days. When people say to me in late February things like, "I can't deal with four more weeks until Spring finally starts", I try to reassure them that the average weather is two weeks ahead of those hoary old definitions. The end of winter in my book is counted from March 7 every year. Likewise the beginning of summer is June 7 and the beginning of fall is September 7.

What do we "really" use for our practical seasons in the US? The beginning of summer is variously measured by Memorial Day or by high school graduation or similar, which fall very roughly "around" June 7, and the practical end of summer is almost universally counted as Labor Day, which is anywhere from Sep 1 to Sep 7. So, in fact, our practical measures of summer are pretty darn good.

Meteorologists, at least in the US culture of meteorology, count the seasons as starting just a bit earlier, six days before the dates that I usually use. Meteorological Summer begins on June 1. Meteorological fall starts on Sep 1, etc. That's not bad either.

Seasons in the tropics are much less pronounced, and they bear little connection with our solar seasons. There's not much point in knowing the exact day when winter begins. You want to know how long until the rainy season starts. That's the seasonal variation that matters. As a good rule of thumb, "daylight saving time" is useful (or useless and hated despite the practice) in the same regions where traditional solar seasons are useful, and for the same reasons. The traditional seasons are not much help in the tropics, and daylight saving time is rarely adopted in the tropics. This is also true in another extreme pair of regions of the world. Solar seasons are much different at the north and south pole, but at least here we can see the merit of calling the solstices the "mid-summer" dates. 

There are a couple of senses for the solstices as "mid-summer" and "mid-winter" that make good sense. Not quite at the poles, but in the arctic and near-arctic, the length of the day has a more dramatic impact on daily life. And everywhere away from the equator, the length of the day is symmetrical about the solstices with implications for solar power and also more mundane issues like UV exposure. Years ago when tanning was a measure of wealth and even health (what could be more 20th century that a wealthy white person tanned to dark brown while smoking cigarettes all day?), the days that mattered were those centered on the summer solstice. May, June, July were the prime tanning months, while August and September were the tail end.

I don't think the traditional dates for the seasons in the US, Canada, and other regions with the same definitions are going anywhere, but the discussion is certainly useful. :)

Frank Reed