What makes sunrise and sunset times?

[conuly]

Oh, don't give me "earth's rotation" and "angle of the sun", I know that! But...

Look, I googled to find what time sunset was in NYC, found out that today it's 4:55. Sunrise was at 7:17 Okay, fine.

But in Anchorage it's sunrise at 9:52, but sunset only at 4:27. The sun rises two hours later there, but it only sets about thirty minutes earlier? I'd always assumed it was constant - if it rises an hour later, it sets an hour earlier, that sort of thing. And over in Honalulu the sun rose at 7:12 (barely earlier than here in NYC) but it doesn't set until 6:12. Maybe my problem is in viewing my own city as the default, and if I viewed these times as varying from the equator (or the North Pole) they'd make sense? I understand that summer and winter are more dramatic closer to the poles, and less dramatic close to the equator, but... like I said, I thought you took from both sides of noon more or less evenly wherever you were.

Comments

[ancarett]

Look at a map of time zones. See how Anchorage is about a full hour behind the time zone's boundary map by how far west it is -- that skews the linkage between when the time zone says it is X hour and what the local perception of time would be.

You and I both are on EST, I believe. Now, I live far north of you (another factor shortening the days this time of year) and our local sunset is at 5:08 (our sunrise was at 8:02). So that's partially a factor of our westward as well as our northern position within the EST boundaries.

[pne.livejournal.com]

Look at a map of time zones. See how Anchorage is about a full hour behind the time zone's boundary map by how far west it is -- that skews the linkage between when the time zone says it is X hour and what the local perception of time would be.

That's what I thought it might be, too.

Time zones mean that it's 12:00 in a wide swath of land, but of course the sun can only be directly overhead one point. So local solar noon will not usually be when the clock says 12:00, unless you live on your time zone's central meridian.

So if you're a bit west, then sunrise and sunset would (I imagine) be symmetrical around noon -- but not symmetrical around 12:00.

---

...although, come to think of it, sunrise and sunset are not perfectly symmetrical around noon. One consequence of this is that the day with the latest sunrise is not the day with the shortest period of daylight -- though the two days are close together.

If I remember correctly, the reason for this has to do with something called the "equation of time"; something about the sun at noon moving in a figure-eight-shaped path relative to you over the course of the year rather than straight up and down. Every time I tried to read up on it, though, it went over my head.

[ancarett]

If you're interested, there are some cool stories about the guy who invented standard time: Sir Sandford Fleming.

[oloriel.livejournal.com]

Tilted axis. And the earth is moving in an ellipsis instead of a circle, which probably contributes to the mess.

As in, the earth's axis isn't orthogonal to the sun, so the rotation and the angle are asymmetric, so it's no "even" constant but a mess. I know, you didn't want to hear "earth's rotation" or "angle of the sun", but sorry, that's the answer.

For the same reason, it's not like you add (say) a minute every day on either side of noon after the winter solstice to get the sunset/sunrise times just for one single place - no, it's a bit here and a bit there and a mess all over. Where I live, from late November on sunset stays virtually the same while the sun rises later and later until late December, when sunrise stays the same but the sun sets a bit later every day , until in January finally there's a little gain on either side - a minute here, three minutes there, etc etc. So even in this one place it's never a symmetric development.

[conuly]

As long as it's part of a complete answer, I'm cool with it. I just didn't want to hear "Well, Connie, as any third grader could tell you, the DAY AND NIGHT are caused by the earth's rotation and the SEASONS are caused by the earth's revolution around the sun, duh".

[redbird]

OK. This may not be complete, but I hope it's helpful. It does involve a little geometry; I hope that's okay.

The sun is a huge ball of very hot gas, producing light and heat (and X-rays and stuff). We can treat the sun as an unmoving object for these purposes.* The Earth is going around the sun in a stable, almost circular orbit.

The Earth's orbit and the sun lie in a plane (along with the orbits of most of the other planets). Think of the Earth as the usual spinning top. But it's not pointing straight up: it's tilted. Specifically, it's tilted compared to the orbital plane. The equator isn't in the orbital plane, and the north pole (or south pole) doesn't point straight up/down compared to the sun and the Earth's orbit.

This means that when the north pole is pointing away from the sun, areas near the north pole aren't in sunlight as long as areas near the south pole. Other parts of the Earth are in the way. The farther you are from the equator, the more extreme that is. Meanwhile, the parts of the Earth that are pointed toward the sun are getting more sunlight. In a couple of months, instead of pointing toward or away from the sun, the axis will be sort of parallel to it, so everyone gets the same amount of daylight.

More sunlight means longer days, and it also means warmer ones, mostly for the same reason. The light also reaches you more directly in summer, so less of the energy is spread out in the upper atmosphere, which affects temperature but not day lengths; if there were no atmosphere, we'd still have longer days at some points in the year than others.

The closer to the solstices (a.k.a. first day of winter and first day of summer, about Dec. 21 and June 21), the more one pole points toward the sun and the other points away.

*This is legitimate: relativity says that we can choose any convenient object or location as a fixed point and define position and motion relative to it.

[I hope this helps; when I'm looking at this for work, I have illustrations. On the other hand, when I'm doing this for work I also have to use grade-school vocabulary and sentence lengths, rather than filling in something for a basically intelligent, educated grown-up. And I don't have the luxury of someone being able to say "okay, I got that part, but this isn't clear."]

[conuly]

I loathe geometry*, but I appreciate that I can't remove it from the world just because I'm not inherently good at it. (But maybe I was taught badly? I should study up on it again, it's not good to be ignorant.)

Now, this all explains why we have day and night and seasons (and I did know most of this more or less, somewhere in my mind), but not why it doesn't take away evenly from noon (solar noon, I suppose) as I'd just... assumed it must. I mean, I didn't think too deeply about it, but I always figured that's how it works, more or less.

Unless it does explain and I'm missing a step...?

*Algebra, now, is easy, just like the little ditty says it should be.

[oloriel.livejournal.com]

I think the problem is that noon is not (as it would be in your model) the precise middle of the sunlit period. It should be, but it isn't, not as we reckon time. (In the ancient Japanese calendar, for example, it used to be the point in time precisely between sunrise and sunset, just as midnight was precisely between sunset and sunrise, and the dark time was divided into six "hours" and the light time was divided into six "hours". Which meant that in summer you had really long day hours and really short night hours, and the other way round in winter. Don't ask me how they actually calculated this, I have no idea. At any rate it means that by our modern standards, noon was at a slightly different time each day...)
So as A Smith below says, the "not exactly 24 hours" bit may well have an impact on this. We say noon is exactly 12 hours into the day and exactly 12 hours from the end of the day, after all - but if a day actually lasts a bit longer than 24 hours, that means that it can only be 12 hours from one end, not from both at once.
Although that still fails to explain why the difference is in actual minutes, not mere fractions of seconds...

Part of the problem may also be applying things that are true for one part of the planet to the whole planet. Most parts of the planet, after all, don't actually have an "equinox" (i.e., night and day are of equal length) at the time that the calendar says so/ the new season begins...

I think the parts relevant to your question are mentioned in the explanation above (tilted axis; equator ≠ orbital plane), but if there was an actual explanation on just why days don't get longer/shorter equally on both sides of noon, I missed it, too.

[redbird]

The ellipse is a minor factor; Earth's orbit isn't that far off a circle (though not as close as Mercury)'s. If we were on Mars, the shape of the ellipse would also be relevant.

[elenbarathi.livejournal.com]

Okay, here ya go - what you need is the US Naval Observatory Data services (http://aa.usno.navy.mil/data/). Everything you want to know is there. Especially note the Duration of Daylight/Darkness Table for One Year feature - compare the table for Anchorage, AK with Miami, FL; you'll see how the difference works. There are good explanations for why it works that way in their Astronomical Information (http://aa.usno.navy.mil/faq/) section.

Have fun!

What is a day?

[(Anonymous)]

There is one fact that none of your other commentators have mentioned.
A day is not really 24 hours. It is actually a bit longer. Just like the Earth's rotation around the sun is not a true 365 days. We have to round things out to whole numbers for our feeble minds to comprehend or end up frustrated trying to figure out the exact time of day.

That extra day every four years compensates for a lot.

On a side note, I am not a LiveJournal user and haven't taken the time to figure out what an OpenID is so I just leave my posts in the default position. But if you must have a name for this occasional reader.

A Smith

Another side note, ever wonder why there are two words in those sign in verification things at the bottom of the reply section? I know, just wondering if a curious person like yourself has ever thought about it.

Re: What is a day?

[conuly]

My mother would say it's so one can look after the other, and that it gives people with nothing to say something to talk about. (And she'd say it in French, too, so it took me a long time to understand what she was saying!)

Open ID is a... um... thingy set up by LJ so that if you use one of a variety of blogging services (or just like to post with a verified identity at blogs you read) you can use the same username everywhere and people know it's *you* and not somebody cannily pretending to *be* you.

To tell the truth, it's easy enough to see if you're being faked by most trolls, because most trolls are stupid in a few very important ways, but it's always nice to have a name to peg on to somebody. I can't explain it, but true anonymice are creepy. (And often spam, but that's beside the point.)

You think the discrepancy between the clock day/year and the actual length of the day/year is part of this?

Re: What is a day?

[(Anonymous)]

I think it is a combination of it all: Time zone association, the Earth's angle of axis in relation the position around the sun and distance during a given period of time (I think it takes 8 minutes from the light of our sun to reach us... or maybe I am confusing that with the light from the moon?), and a calendar and clock systems which we correct for every four years.

It gets even worst when you hear about how shockwaves from large volcanic eruptions, earthquakes and even atomic bombs have apparently shifted the planets axis.

Oh, and with those two word verification things... One word is a known word, a control word if you will, while the other is a word scanned in from a book that the computer cannot read properly because the print or page has been disrupted by creases, stains, etc... So the guy who created this verification system is also in charge of scanning in all these books for archiving, and instead of trying to fix all the unintelligible words with a small group of people he has enlisted all of us who use his verification system to help him one word at a time, assuming the human brain is able to read and understand what these distorted words are better than a computer.

I just thought you would get a kick knowing that everyone who replies to your postings is helping to restore a lost piece of literature, thus, you with every one of your postings is saving a book. How cool is that?

Re: What is a day?

[pne.livejournal.com]

(I think it takes 8 minutes from the light of our sun to reach us... or maybe I am confusing that with the light from the moon?)

No, the moon is on the order of one light *second* away. 8 minutes is about right for earth–sun.

a calendar and clock systems which we correct for every four years.

The calendar we correct every four years; the clocks we correct irregularly, with leap seconds. (For a while, leap seconds were inserted about every 18 months, but then there was a pretty long "dry spell".)

everyone who replies to your postings is helping to restore a lost piece of literature

Those CAPTCHAs are just shown to anonymous commenters; users with a LiveJournal account aren't shown one. (At least, this is my guess; it could also be that I don't see one because I'm on her friends list, but it seems more likely to turn on CAPTCHAs only for anonymous comments, not for all non-friend comments.)

Re: What is a day?

[lizziey.livejournal.com]

They're also shown to LJ users (even paid ones) if a post has 500 comments already. ;)

Re: What is a day?

[conuly]

I didn't know that.

Re: What is a day?

[lizziey.livejournal.com]

Yep. I know from being in sf_d and sfd_anon. :)

Re: What is a day?

[lizziey.livejournal.com]

And actually, I think it is 5000 comments, not 500. Sorry.

Re: What is a day?

[conuly]

Why, why, why do you frequent the drama comms? There's a reason people online look at me askance!

Re: What is a day?

[lizziey.livejournal.com]

They're fun and I'm an insomniac. :/

[peebs1701.livejournal.com]

Unless "sunrise" and "sunset" are measured as the time at which the sun rises above/sets below sea level, then local geography may have an effect on that as well.

[redbird]

I seem to have answered the wrong question.

Basically, the problem is time zones. [delete ramble about why time zones are useful.]

In theory, the planet is divided into 24 time zones, all the same width. In that theory, the lines are every 30 degrees, and the time for each zone is that of the center line. So, one zone from 15 east to 15 west (keeping Greenwich Mean Time) and so on around the world.

In practice, there are more time zones (for example, Newfoundland time is displaced by half an hour), and the lines are drawn for a variety of political reasons. They tend to go along state borders, not through the middle (though there are exceptions). Some years back, they decided to lump Alaska into fewer time zones (it had had three time zones, and now has two, one for the mainland and one for part of the Aleutians).

And sometimes the official time for a zone isn't even close to the midline: the entire People's Republic of China is on Beijing time.

(And, because there are exceptions to everything, the above does not apply to Saudi Arabia, where the legal time is local solar time. Computer operating systems ship with time zone instructions and, for Saudi Arabia, instructions to enter your longitude so the appropriate value can be calculated as an offset from GMT.)