OK, so here’s a question that has been bugging me for years. March 20 is the beginning of astronomical spring, correct? That means that on this day, the entire planet should have exactly 12 hours of day and 12 hours of night.
So, why, in my almanac for my eastern North Carolina town, is TODAY, March 17, the day that has exactly 12 hours of sun and 12 of night. The sun rises at 6:18 am and sets at 6:18 pm. By the time we get to March 20, the sun rises at 6:14 am and sets at 6:21 pm, meaning the day is 7 minutes longer than the night. And I’ve double checked this in more than one almanac. If March 20 is really the “equinox”, what gives???
Almanacs and newspapers use a somewhat liberal definition of “sunset” and “sunrise.” Typically they’ll say sunrise is when the sun first becomes visible, and sunset is when it is no longer visible. The result is that the day, using these measurements, is slightly too long.
The astronomical definition of sunrise is that it occurs when exactly one half of the sun’s disc is visible above the horizon, and sunset occurs when exactly one half of the sun’s disc is visible above the horizon also. That length of time will be very close to 12 hours on the Equinox.
Thanks Friedo, and that’s a point that I hadn’t thought of before. If you count sunup as the first bit of the sun over the horizon, to sundown as the last glint of evening sun to go below the horizon, that could actually add 5 to 7 minutes to the “day”. But apparently this is pretty universal problem. I just checked with the U.S. Naval Observatory’s website, and (again for my eastern NC town) they’re already showing us with our day-length at 12 hours, 4 minutes…on March 18.
I think the technical definition of “equinox” is the moment when the sun’s declination is zero, that is, the sun is right on the celestial equator. (A more PC definition in terms of heliocentrism would be the moment when the earth is at one of the two points on its orbit where the plane of the orbit intersects the plane of the earth’s equator.)
Using a simplified geometric model, this situation does indeed produce equal lengths of day and night at all latitudes. But IIRC the physics screws it up a bit, so the day on which the zero solar declination occurs generally does not have exactly twelve hours of daylight.
Not only that, but the atmosphere refracts the sun’s light, so that when you see the sun at the horizon, it’s actually already past the horizon (by about a half-degree or a minute IIRC), and you only see it because the atmosphere is like a lens.
WeatherGeek68, it seems this explains the effect you’re looking for. It sounds like your second post is saying that this effect doesn’t explain something, so I’m confused.
Heh-heh…yeah, I had to dis-abuse a coworker of that mythical notion yesterday. He was red-faced and spittle-flecked, especially after another coworker balanced one on the counter in the break room. CONFESSION: we shook it first.
So yeah, the “physics that screws up” the simplified geometrical identity between “equinox” and “exactly twelve-hour day” is the fact that the Sun’s disk is bigger than a point, and (as Curt noted) atmospheric refraction.
Aha! Hadn’t thought about refraction. I think that settles it for me.
Sorry if the second post was confusing. I was working off of Friedo’s earlier post, that perhaps the definitions of sunrise and set were too liberally interpreted by almanacs and such, and that THAT was giving me the extra five or so minutes. In other words measuring sunrise from first glint over the horizon, to sunset as last glint to sink, could give a longer day than a more uniform measurement, such as sunrise is full sun up and sunset is complete sun below the horizon.
But at any rate, I think the mystery is solved for me. Not really a big deal, but one of those little mysteries that’s always puzzled me. Thanks to all…
WG68
