I understand that Right Ascension is a longitude-like celestial coordinate that varies from 0-24hrs, taken from a reference point of the vernal equinox. More specifically, for star maps that are based on the epoch 2000.0, this specifically means when the vernal equinox occurred in the year 2000. Even more specifically, the vernal equinox is a specific time and date, and location of the exact point in time, when sunrise occurred, of when the day and night were exactly equal in length. Here are my questions:

1) If the Right Ascension is referenced from “Vernal Equinox” position “0hr”, in the year 2000, then why does nobody ever talk about the equivalent physical location? ie: if the right ascension (sunrise of the vernal equinox) occurred in the year 2000 in a particular city, or longitude, why doesn’t anyone ever mention that. ie: (just an example) Right Ascension is always measured from longitude -80deg (close to Miami, Florida, USA), as this was a city the vernal equinox occurred in the year 2000?

2) If the star maps do not change much within a period of a few decades, (eg: take for example, the right ascension and declination of the sirius star is pretty much the same in 2019 as it was in the epoch 2000, or 1990 for that matter), then these RA and dec values should not change much within a decade or so. Ipso facto, we can deduce that the sunrise time for the vernal equinox should not change very much from year to year. ie: Since RA is measured from a reference based on the sunrise of the vernal equinox in the year 2000, then the sunrise of the vernal equinox in the year 2019 should not be very much different (assuming RA is the same today as it was in 2000). But it isn’t!!! If you track the sunrise of the vernal equinox from year to year, it varies a lot! I am obviously missing something… it’s driving me nuts.

Update below:

From Norton’s 2000.0 star atlas, please see the following definitions:

Norton's 2000 Star Atlas Excerpt 1/3
Norton's 2000 Star Atlas Excerpt 2/3
Norton's 2000 Star Atlas Excerpt 3/3

In the last link, It says the 0 line of right ascension is equivalent to the Greenwich meridian on Earth. In a sense, this is the ultimate answer to my question #1, although it begs more details. In Astronomy, why is this rarely mentioned? If it was, star maps would be much easier to comprehend by anyone. (Ie: RA of any star is simply referenced from the Earth equivalent of longitude 0…).

Does this also mean, that during the vernal equinox, the point in time where the sun crosses from the Southern Hemisphere to the Northern hemisphere, longitude 0 meridian on Earth is directly at that intersecting point too? So, as Mike G said, “sunrise" is irrelevant on the day of the vernal equinox. I understand this part. But the exact moment of the vernal equinox does matter… because from any point on Earth, the stars all move in the sky as time progresses, hour by hour. If you’re going to base a map of where the stars are, it has to be referenced from one moment in time, and from once longitudinal reference (and one declination reference).

Now, bringing it all back to star maps: @Mike G: Would you say that simply, any star map based on J2000 epoch is basically where all the stars were on Jan 1, 2000 (at 12 noon greenwich time)? However, you also mentioned the J2000 epoch does NOT depend on the equinox. This is where I go fuzzy. From reading the reference links above, it seems like it does depend on the equinox. If the J2000 epoch star maps are indeed referenced from 12 noon, then in the year 2000, at the time of vernal equinox, the greenwich meridian intersection occurred at 12 noon?

Update 2

OK, I think I have some of my basic misconceptions cleared. RA 0hr simply cannot be measured relative to any point on Earth. It is measured against the first point of Pisces (It used to be Aries a long time ago). Specifically, for J2000 star maps, the RAs would be calculated based on relative positions from Pisces in the year 2000 around March 21 (vernal equinox). (Norton's Atlas -2nd link above- describes RA as around this date. I found no reference for January 1 -please correct me if this is not the case).

So to summarize what I understand, if you're only given an RA and declination for a particular star and you wanted to find it in the sky, you would need to innately know where Pisces is. From there, you can roughly work out approximately how many degrees of "longitude" away from Pisces your star is. (So, you would also need to know where due North and due South are). You would also have to know what latitude you are at on Earth, so you can work out roughly where your star's declination places it on it's meridian in the sky. For example, if you knew you lived at +40deg latitude on Earth, and you looked straight up at the zenith, you could estimate that the horizons are at about +90deg and -90deg away from that +40deg position. For there, you could interpolate where on the meridian (declination) your star lies.

Update 3

Thanks Mike G for the wiki reference to Celestial Coordinates. This shows that the star atlas J2000 is NOT referenced to the vernal equinox, but simply references to a snapshot in time of the stars on January 1, 2000, noon terrestrial time (or 2000 January 1, 11:58:55.816 UTC, which was the old greenwich mean time). The RAs and Decs in the atlas are referenced to Pisces at that moment in time.

  • $\begingroup$ What exactly is "the sunrise of the vernal equinox"? If you showed some numerical examples that demonstrate what you mean by "from year to year, it varies a lot!" What varies, and how much is "a lot"? That might be a big help. Thanks! $\endgroup$
    – uhoh
    Commented Oct 19, 2019 at 7:12
  • 1
    $\begingroup$ You asked exactly the same question at the physics StackExchange. Except now you have even more misconceptions than you did earlier. $\endgroup$ Commented Oct 19, 2019 at 12:20
  • $\begingroup$ Further reading: en.wikipedia.org/wiki/Equinox_(celestial_coordinates) $\endgroup$
    – Mike G
    Commented Oct 22, 2019 at 17:39
  • $\begingroup$ @WillK Based on your Update 3, I think that you are confusing the "vernal equinox" between a date/time in March versus a location on the sky. The position of 0 hours right ascension is based on the location on the sky where the ecliptic and equator intersect. This point is often abbreviated as the vernal equinox (as in the "location of it") by astronomers. Because of precession, this location changes with time. This position is different on Jan 1 2000 compared to the Mar 20 2000 (the "date" of the vernal equinox). When referring to coordinates, the date of the system needs to be given. $\endgroup$
    – JohnHoltz
    Commented Oct 24, 2019 at 17:16
  • $\begingroup$ Yes, I believe that was indeed the confusion. As you say, star maps should be more clear. Norton's J2000 is particularly confusing because they talk a lot about the vernal equinox, but they don't clearly state is that their maps are based on a Julian date Jan 1, 2000 -not the date of the Vernal Equinox. Now that I understand that, I can sleep better at night. ;) $\endgroup$
    – WillK
    Commented Oct 27, 2019 at 3:35

2 Answers 2


Fortunately things aren't as entangled as you thought. The J2000 epoch does not depend on the equinox; it is simply 12:00 TT on January 1, 2000. The time of vernal equinox is unrelated to sunrise anywhere; it is just when the Sun appears to cross the celestial equator northward as seen from Earth.

The vernal equinox point ♈ is located at the intersection of the equator and the ecliptic. Precession changes the equatorial plane over time, shifting its intersection with the ecliptic plane by ~50" per year or ~1.4° per century. The J2000 equinox is where ♈ was at the J2000 epoch.

The variation in equinox times is an artifact of approximating the tropical year's odd quarter day with a whole day every four years.

The Greenwich meridian and the 0h hour circle have analogous roles in their respective coordinate systems, but one rotates with the Earth and the other doesn't. They align at 0:00 Greenwich sidereal time each sidereal day. Local sidereal time corresponds to the right ascension aligned with the local meridian at that moment.

As the Sun appears to move ~1 degree per day along the ecliptic, sidereal time runs faster than solar time by ~4 minutes per day or 1 day per year. Solar and sidereal times match when the Sun is at the autumnal equinox ♎ (RA = 12h) and differ by 12 hours at the vernal equinox.

The difference between local sidereal time and a star's right ascension is the star's hour angle east or west of the local meridian. Using this, any star of known identity and right ascension can serve as a reference for locating other stars by coordinates. Equatorial setting circles rely on this.


There are several meanings to "vernal equinox", and I think you have not picked the right sense here.

The vernal equinox is here taken to be a point on the celestial sphere. It is the point where the projection of the Earth's equator to the celestial sphere and the projection of the Earth's orbit intersect. (There are of course two points, the other is the autumnal equinox). This is also called the "point of Aries" (despite the fact that it is currently located in Picies) This point always exists and only moves very slowly. It is a point in the sky, not a point on the Earth.

The other meaning is the time when the Earth moves from the the portion of it's orbit that is in the south, to the part when it is in the North. This occurs around the 21st of March each year. (The exact date varies, due more to the calendar than to irregularities in the orbit). There is no place on Earth that is at the vernal equinox. In this sense it is a time, not a place.

Right ascension is measured from the Point of Aries, and in particular the location of the Point of Aries at the start of January of the year 2000 (hence 2000.0) The point of Aries moves slowly, but quickly enough for there to be a need to update stellar coordinates every 50 years.

The right ascension is not measured from the sunrise at the vernal equinox in 2000. It is measured from the Point of Aries in January 2000.0

In the Norton sky atlas is says "the zero line of right ascension is the equivalent of the Greenwich meridian". This is correct in the sense that they play the same role in there respective coordinate systems. It is not the case that there is any other relationship between the two lines. The zero line of right ascension is not the projection onto the sky of the Greenwich meridian.

I'm not sure what you mean by "rarely mentioned". It is implict in any map of the celestial sphere, which would show lines of constant right ascension as great circles from one pole to the other. This point is not deep or complex.

It does not mean that longitude 0 is at any particular point at the moment of equinox. The exact time of equinox can occur at any point in the day-night cycle. At that time, the Greenwich meridian could be pointing in any direction, relative to the stars or to the sun.

The stars do move in the sky, but they do not move (much) on the Celestial sphere. The definition of the celestial sphere means that the stars are almost fixed and I can talk about the coordinates of a star. Even as the stars appear to travel from East to West, their celestial coordinates remain unchanged.

Stars only appear to move slowly due to the precession of the equinoxes, nutation, the actual movement of the star in space and the slight wobble of parallax. Distant stars and galaxies don't have any measurable actual motion or parallax.

So when I give coordinates of a star position, I need to define my frame of reference. One way to do that is use the actual position of the Point of Aries at the current time. However the very slow precession means that even distant star's coordintes will change over the years.

Alternatively I can use the position of the Point of Aries on Jan 1 2000. This means that distant stars will be completely fixed in this coordinate system.

To put it another way: The J2000.0 system has a point of aries (or vernal equinox) that is fixed in relative to distant stars.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .