I’ve been wondering this for a while, and I’m hoping some of you can help me out. The winter solstice marks the shortest day of the year, December 21. So when is the longest night of the year–the night preceding it or the night following it?
Thanks,
Amanda
That’s a good question but it’s kind of tricky. The winter solstice is not a date, but an exact time. For instance, in 2002 the solstice occurs on Dec 21 at 8:01 PM EST (cite). If you look at the chart on that page, you’ll see that the solstice happens at a different time each year, but it’s always on Dec 21 or 22. Now, I may be oversimplifying things, but I believe that whatever day is closest to the solstice will be the shortest day, and whatever night is closest to it will be the longest night. So, for instance, in 2002, in the Eastern Time Zone, Dec 21 would be the shortest day, and the following night would be the shortest night. But looking ahead to 2004, the solstice occurs on Dec 21 at 7:41 AM. So in this case, the night before Dec 21 would be the shortest night.
Thanks! I assumed it was actually a specific time, but I hadn’t been able to find a table like the one you linked to. My wedding is on 12-21 (this year), so I’d been wondering whether the night before it or the wedding night itself would be the longest of the year. 
I appreciate the information.
-Amanda
** Achernar (the hottest of the top ten) ** [sub](yeah, I know I always do that)[/sub]is correct on a first approach but the question is trickier than that. First, the local time of ephemeris depends on geographical location so what happens on one date for one observer may happen on the previous or next date for an observer located elsewhere. To give any universal meaning to the question you would have to pick a location with some universal meaning (or just calculate for your own particular location). You could choose Latitude 0° 0’ 0", Longitude 0° 0’ 0" (on the corner of Equator and Prime Meridian Streets) as a significant location. . . except nobody lives there and seasons at the equator are non existent as the days and nights there are practically equal all year long (except for minor variations caused by the equation of time). So a better place to pick might be Greenwich Observatory (Latitude N 51° 28’ 38", Longitude 0° 0’ 0").
Besides geographical location I also suspect the longest night may not necessarily be that one closest to the solstice due to the effects of the equation of time, but I would have to confirm this by calculating it for a number of years and I am too lazy to do it right now.
You would also have to decide whether you take atmospheric refraction into account etc. All these are very minor corrections of a few seconds which have no practical effect in real life. Whether the weather is clear or overcast locally is going to have a much greater practical effect on the time length of darkness.
http://greenwichstar.com/solstice/index.htm
Ok I see you are in Georgia and you want to know for this year only. My calculations show for year 2002, N33 W84, the night following the 21 is about one second longer than the previous and the following.
12/20/02 - 14:02:04
12/21/02 - 14:02:05
12/22/02 - 14:02:04
Surely I’m not the only poster with a star named after him or her? Anyway, mmeblueberry, you should trust sailor’s uber-capable assessment of this situation (and not just because he agreed with me).
>> Surely I’m not the only poster with a star named after him or her?
But I am the only one with a moon named after me!
What? You never heard of Sailor Moon?
mmeblueberry, I can understand why you would want your wedding night to be a long one . . . or maybe not. Your fiance doesn’t have any problems in the, ahem, “astronomical” aspect, does he?
Be careful later and avoid saying “My wedding night, now that was a looooong night!”
Doesn’t the E of T affect sunrise and sunset by the same amount? It doesn’t seem to me that it should affect the length of night, just when exactly night occurs. And refraction, angular size of the Sun, etc., should have the same effect on all nights (or at least a random effect, for refraction).
And I’ve got a planet, in Latin translation, at least. sirius has never posted, nor has rigel. Vega has posted twice. No Altair, Deneb, or Betelgeuse, though.
Chronos, I understand what you mean and in general terms you are correct but the EoT does not add or subtract a constant amount every day. The EoT as you know it is the integral of the time added or subtracted daily so you can see this amount varies slowly over time. Suppose the length of night without accounting for EoT would be during several days in a row:
day 1 12:00:00
day 2 12:00:02
day 3 12:00:00
the longest night is day 2. Now if the EoT daily effect is a constant 2 sec then you have
day 1 12:00:02
day 2 12:00:04
day 3 12:00:02
and day 2 is still the longest night. But if the effect of the EoT is increasing by 3 sec/day, after accounting for the EoT you could have a result like this:
day 1 12:00:02 +3 =12:00:05
day 2 12:00:04 +6 =12:00:10
day 3 12:00:02 +9 =12:00:11
I have not studied the EoT closely enough now to see at what times this would happen and at what rates. It certainly would be quite a coincidence if it happened at the solstice so, in that sense you are correct but it is not tru the year around.
Another thing that happens is that even though after the solstice the days start getting longer sunrise may occur later for a few days due to the EoT. In other words, the latest sunrise is not the closest to the solstice.
We would also have to define “night”. When does it begin and end? When the center of the sun geometrically crosses the horizon? When the upper limb geometrically crosses the horizon? When the center apparently crosses the horizon? When the upper limb apparently crosses the horizon? When the center is a certain angle below the horizon? (Nautical twilight is defined as the moment when the sun is below the horizon a certain angle).
Anyway, I think the OP has been sufficiently answered and she doesn’t need all this level of precission.
But is the effect
I think I see what you’re getting at, sailor The equation of time varies over the course of the year, which means that it will (in general) be slightly different at sunrise and sunset. One might call this the differential equation of time, I suppose. That’s a rather high-order effect, of course, since the E of T itself is already pretty small, and the differential E of T would be bringing in another order in (day/year). Then again, though, near the solstices, the change in day/night length is also a high-order effect (quadratic in separation from the true solstice), so I’m not sure which small effect is larger. I imagine that it would depend on latitude.