Earth spins on an axis that is tilted 23.4°, and — this is the whole trick — that tilt points the same direction in space all year long. It does not swivel to follow the Sun. It just leans, stubbornly, at the same distant star.
So for half the year the northern half of the planet leans toward the Sun, and for the other half it leans away. Nothing about the Earth changed. What changed is which end of it is tipped into the light.
Watch the ring drawn around the globe in the space view — that's your latitude, carried all the way around by the spin of the day, with the green dot marking you at noon. The gold part of the ring is in daylight and the grey part is in night. Drag the day slider and watch it swing through the light. The gold fraction of that ring is your day length. In June most of it is gold; in December most of it is grey. That is the entire explanation of summer and winter, and everything else on this page is a footnote to it.
This is the most common wrong answer, and the fun part is that the truth is the exact opposite of it.
Earth's orbit is very slightly oval, so our distance to the Sun does change over the year — by about 3%, from 91.4 million miles to 94.5 million. But we are closest to the Sun in early January, around the 3rd, which is the dead of winter here. We are farthest in early July, around the 4th, in the middle of summer.
So the northern hemisphere actually gets about 7% less sunlight during its summer than the southern hemisphere gets during theirs. And yet July is still hot. The tilt wins, and it isn't close.
The tell: if distance caused seasons, the whole planet would have summer at the same time. It doesn't. When it's July in Hyak it's the middle of winter in Sydney, and both places are exactly the same distance from the Sun.
Two things at once, and they push in the same direction:
Hyak sits at 47.39°N. That's far enough north to matter a lot — slightly closer to the North Pole than to the equator, and only about 110 miles south of the Canadian border, which out here is the 49th parallel. Budapest is at essentially the same latitude, and so is the top of Mongolia.
| At Hyak | Sun's highest point | Daylight |
|---|---|---|
| Summer solstice, ~Jun 21 | 66.0° | 15 h 57 m |
| Equinoxes, ~Mar 20 & Sep 22 | 42.6° | 12 h 07 m |
| Winter solstice, ~Dec 21 | 19.2° | 8 h 27 m |
That is a swing of seven and a half hours between the longest day and the shortest. Near the equator the same swing is a few minutes and nobody bothers to notice.
Why the mountains make it worse: these numbers are for a flat horizon. At Snoqualmie Pass you're in a notch with ridges on both sides, so the Sun clears the actual skyline well after it technically rises and drops behind rock well before it technically sets. In December the difference can be the better part of an hour at each end.
Look closely at the year chart above, down at the bottom of December. The sunset curve bottoms out around December 10th and has already started climbing again before the solstice arrives on the 21st. The sunrise curve keeps getting later until about January 3rd. The shortest day is squarely between them, on the solstice, exactly where it should be — but neither end of it is at its own extreme on that date.
People notice this every year and assume the calendar is wrong. It isn't. Two things are going on at once, and they're both real:
Together those make solar noon — the moment the Sun is actually highest — drift away from clock noon by up to about a quarter of an hour either way over the course of a year. In December solar noon is getting later every day, which drags sunrise and sunset later together, and that drift is briefly stronger than the day shortening. So sunset starts recovering before the day does.
The chart is drawn in standard time on purpose. Daylight saving is a clock change, not a sky change — the Sun does not care. The shaded band shows when the clocks are pushed an hour ahead, which is why sunset appears to jump an hour in March and fall back an hour in November without anything happening outside. The dashboard shows real clock time, so during the shaded band its numbers are an hour later than this chart's.
There's a companion effect worth knowing: the days start getting longer on December 21st, but the coldest part of winter is a month or so later, and the hottest part of summer is a month after the June solstice. The ground and the ocean take weeks to catch up with the sunlight falling on them. Sunlight is the input; temperature is the slow, lagging output.
You are, at this moment, not moving at all as far as you can tell. Here is what you are actually doing, stacked one motion on top of another:
You feel none of it, because steady motion is not something a body can detect — only changes in it. That's the same reason a smooth aeroplane at 500 mph feels like a room. The Earth's spin does change direction constantly, but so gently that it takes a careful instrument to notice.
Point a camera at Polaris on a clear night at the Pass, leave the shutter open for an hour, and every other star draws a neat arc around it while Polaris stays put. That is a direct photograph of the axis in the space view above — the one that never moves. It is the same axis, the same lean, the same reason you get sixteen hours of light in June.
Slowly, though: the axis does eventually swivel, tracing a full cone every 26,000 years. In about 12,000 years the north star will be Vega instead. The ancient Egyptians used a different one. It's fixed for your purposes, and not for the pyramids'.
The sunrise, sunset, daylight and Sun-altitude numbers are computed properly and land within a couple of minutes of the real thing. The drawings take liberties:
The dashboard's own sunrise and sunset come from a proper ephemeris via Open-Meteo, so if the two ever disagree by more than a minute or two, trust the dashboard.