Time Dimensions: Astronomy & Beyond

[DaveC426913]

This is really a conceptual question about time dimensions bit it came up in the context of astronomy so I'll put it here.

I was gazing up at the sunset sky the other night and saw a very bright (twinkling) star. I was about to make a note about its location so I could look it up later. I realized that using the alt-azimuth system, I needed four coordinates to look it up: 2 spatial and 2 temporal.

I needed the altitude and the azimuth, obviously, and obviously the time of day, But that was not enough - I also needed the day of the year.

I realize, technically, that those are both the same time dimension, just at different scales (hours versus days) but that is not very useful or convenient. It makes much more sense to specify a day of the year, along with the time of the day, since it's (almost) in the same location every day, and the hours of the day are cyclical.

Does it - er - qualify to be a second (abstract) dimension - or at least a second time coordinate?

 

[Halc]

Time of day and day of year are needed to convert to sidereal time of day. But yes, it's still just one dimension.

 

[DaveC426913]

Time of day and day of year are needed to convert to sidereal time of day.

Conversion? I'm simply talking about the process of pinpointing a star in the sky (say, to show to someone or look it up in a virtual telescope). You don't need to convert anything; you simply need the two sky coords and the two time coords.

 

[Vanadium 50]

To point to a star you need three numbers - a direction, and elevation and when to look. Two spatial and one time dimension.

 

[DaveC426913]

...and when to look...

But "when to look" can be as simple as "10PM". That'll be sufficient for quite a few weeks in either direction of "now". In that sense, one does not really need a specific date of the year.

< analogy >
It's kind of like specifying a specific house on the Earth using a Solar System coordinate system. Technically, "all" you need is 1] radial distance, 2] inclination from the ecliptic and 3] azimuth of the SS.

But that's a really awkward and uninformative system. All of your coordinates would be teeny tiny fractions:
- 1AU ±0.00001AU radius (the width of Earth)
- 0 degrees ±0.0002 degrees inclination (the pole-to-pole height of Earth), etc.
- 87 degrees clockwise from galactic centre ±0.0002 degrees

An example might be:
- radial distance: 1.000083709AUs,
- inclination: 0.00015 degrees N,
- azimuth: 87.00001 degrees.

It would be much more informative to specify the large scale location of the Earth that way, followed by simple Lat, Long and Alt.

That's six coords.
< /analogy >OK, so I'm not really talking about "dimensions" except in the abstract sense. but, as a database entry I would certainly specify six coordinates. It would b way more useful.

I realize this is not so much rigorous thinking, physics-wise, but I'm pretty sure it's valid mathematically, though specific "*ology" escapes me at the moment.

 

[collinsmark]

The time is one dimension. It will save some trouble though, if when you record it, you convert it to universal coordinated time (UTC). Otherwise, the person to whom you are communicating this information to, will have to do the conversion themselves. Basically, they'll need to know what time zone you are in, and if you were in daylight saving time or not when the time was recorded.

Once you have that figured out, just record the time in the following format:
yyyy-MM-dd-hh:mm:ss.<fractional second>

So that's one dimension, but you will need more. Specifically, you'll need:
Azimuth of object
Altitude of object
Your latitude
Your longitude
(We can ignore your elevation since we're discussing things farther away than say the moon.)

And with those pieces of information, there is enough to convert that to celestial coordinates in terms of right ascension (RA) and declination (dec).

Assuming the object is far away (such as farther away than the moon), and is not moving quickly like an astroid or comet, right ascension and declination are enough to tell any astronomer on Earth where to point their telescope.

 

[DaveC426913]

right ascension and declination are enough to tell any astronomer on Earth where to point their telescope.

Sure. Except Right Asc/Declination coords are not very conducive to visually spotting a star in the evening sky with the naked eye (see OP scenario). Alt-Az is better.

And for that, time-of-day is a required coord.

 

[collinsmark]

Sure. Except Right Asc/Declination coords are not very conducive to visually spotting a star in the evening sky with the naked eye (see OP scenario). Alt-Az is better.

And for that, time-of-day is a required coord.

Sure. But if you're going to communicate where to look to another, in terms of their altitude and azimuth, for somebody else who happens to be at a different location on Earth, You'll need some additional information. At least one of you will need to know the following:

• Time of your observation (this includes the date. Also it must include time zone and daylight saving status; OR convert your local time to UTC.)
• Your observed altitude of object
• Your observed azimuth of object
• Your longitude
• Your latitude
• The other person's longitude
• The other person's latitude
• The other person's local time information, such as time zone and daylight saving status. This piece of information is not necessary if the time communication is in UTC.

You'll then need to do some coodinate transformations for the conversions. Only then can you tell the other person, "look at this altitude and this azimuth at this time," for them to see the object of interest.

Somebody has to do the math. Your observed altitude, azimuth and time will be different from their altitude, azimuth and time. And if there are 50 different people around the globe you'll have to do the math 50 different times.

Or you could just convert to declination and right ascension, tell the other people those two values, and let the others convert it to their own altitude and azimuth (and time) themselves.

 

[Mark44]

I needed the altitude and the azimuth, obviously, and obviously the time of day, But that was not enough - I also needed the day of the year.

I realize, technically, that those are both the same time dimension, just at different scales (hours versus days) but that is not very useful or convenient.

Yes, they are both components of the same point in time, just with greater or lesser precision. Specifying the year and day gets you to a 24-hour period, and then specifying the hour, minute, and second gets you to a specific one-second interval.

 

[Vanadium 50]

One needs to specify when to look. The fact that once can also see it at other times too is not evidence for a second time dimension.

 

[Tom.G]

Probably more than you want to know, but here it is.

If you look up Star Atlas you will find that astronomers use the Julian Date.

from: "Norton Star Atlas and reference handbook seventeenth edition" pg.14
Published 1978, ISBN 0-582-98898-5

The Julian Period isused to calculate the exact interval between dates at long intervals apart. It started on Jan. 1, 4713 B.C at noon. The Julian Date (J.D.) is the number of days that have elapsed since the beginning of the Julian Period. The Julian Day is given in decimal form, and not in hours and minutes. It is reckoned from Greenwich noon. Thus 1971 Jan. 1, 9:00 p.m. is J.D.2440953.375. Any astronomical time less than 12 hours (0.5 day) still belongs to the Julian Day preceeding the civil date.

The Modified Julian Date is found by subtracting 2,400,000.5 from the Julian Day numbers.

e.g. 1964 Dec. 31 midnight is J.D. 2438395.5​

. . . . . . . . . . . . . . . . . . . . . . . .M.J.D. 38395.0​

It seems the Astronomers beat us to finding a Universal time reference.

Cheers,
Tom

 

[collinsmark]

Probably more than you want to know, but here it is.

If you look up Star Atlas you will find that astronomers use the Julian Date.

[...]

It seems the Astronomers beat us to finding a Universal time reference.

Cheers,
Tom

Yes, it's true. Julian Day is the used by astronomers and is the preferred format for pretty much all astronomy programs (and calendar programs), under the hood. It allows you to represent date and time in a single floating point number format. It makes calculations way easier, if you're doing calculations involving absolute time (by "absolute" I mean both the date and time).

I've written some astronomy programs before myself, and yes, all dates and times are lumped into Julian Day everywhere except the user interface.

-- I suggested converting to UTC earlier in this thread only because, well, Julian Day conversions are probably not something you want to do in your head.

 

[sophiecentaur]

Sure. Except Right Asc/Declination coords are not very conducive to visually spotting a star in the evening sky with the naked eye (see OP scenario). Alt-Az is better.

And for that, time-of-day is a required coord.

Measurements of position and time are always relative to an origin. Just because astronomers choose to use a particular system it doesn't involve any other dimensions. Any Earthly astronomer will need to compensate for the Earth's rotation and tilt. If he moved into deep space and wasn't rotating relative to the galaxy then he would only need three numbers in order to point at any star he wanted. Even then, the stars (and galaxies) aren't actually fixed so he'd also need to have the information for compensating for that. But I don't think we would call that "introducing extra dimensions".
I have to agree with you about manual finding with Equatorial mounts being fiddly, though. The net advantage used to be clear when people didn't have goto because a single motor (or useful friend) could track the RA. Alt Az fails (a bit) due to rotation of the sky but that could be compensated for by rotating the camera / focuser or the stacking software.

 

[DaveC426913]

Yeah, I should reiterate that I didn't really mean physical dimensions; I guess I meant coordinates.