Equatorial Coordinate System

The equatorial coordinate system is the address book of the sky. Every star, galaxy, quasar and pulsar in every modern catalogue is filed under two numbers — a right ascension $\alpha$ and a declination $\delta$ — and those numbers mean the same thing to an observer in Chile, in Hawaii, or in orbit. The trick is to stop referencing positions to the local horizon, which spins past every star each night and differs from town to town, and instead to inflate Earth's own latitude–longitude grid outward onto the celestial sphere. Earth's equator projects to the celestial equator, its rotation axis to the north and south celestial poles; declination becomes latitude on the sky and right ascension becomes longitude. Because this frame is bolted to the rotation axis rather than to the ground, a star's $(\alpha,\delta)$ is very nearly constant — it does not care where you stand or what time it is. Two design choices give the system its character. First, longitude needs a zero, and the sky's Greenwich is the vernal equinox — where the Sun, climbing the ecliptic, crosses the celestial equator going north each March, tying the origin to the geometry of Earth's orbit. Second, right ascension is measured not in degrees but in hours, because the sky turns once per sidereal day and a sidereal clock literally reads out the RA of whatever is crossing your meridian. The bridge between the epoch-fixed catalogue coordinates and what is up right now is the hour angle, $H = \mathrm{LST} - \alpha$ . The one caveat, and the reason the system carries a date stamp like J2000.0, is that the reference points are not truly fixed: lunisolar torque on Earth's bulge makes the axis precess, dragging poles and equinox slowly across the sky.

Key takeaways

The equatorial system fixes a sky position with two angles: right ascension $\alpha$ , measured eastward along the celestial equator, and declination $\delta$ , perpendicular to it — longitude and latitude projected onto the sky.
The frame is anchored to Earth's equator projected out (the celestial equator) and to the vernal equinox $\Upsilon$ , where the ecliptic crosses the equator tilted by the obliquity $\varepsilon \approx 23.44^\circ$ .
Right ascension is quoted in time units ( $24^{\mathrm{h}} = 360^\circ$ , so $1^{\mathrm{h}} = 15^\circ$ ) because the sky rotates once per sidereal day; declination runs $-90^\circ$ to $+90^\circ$ in arc.
The hour angle $H = \mathrm{LST} - \alpha$ turns fixed catalogue coordinates into an instantaneous position, feeding the spherical law of cosines for altitude, transit height, and rise/set times.
Because Earth's axis precesses with a $\sim\!25{,}800$ -year period, $\alpha$ and $\delta$ drift by seconds of time and arcminutes per century — so coordinates need an epoch like J2000.0, and modern practice pins the frame to distant quasars via the ICRS.
Key idea: right ascension $\alpha$ and declination $\delta$ are longitude and latitude on the celestial sphere, referenced to the celestial equator (Earth's equator projected out) and the vernal equinox (where the ecliptic crosses the equator going north); the coordinates are fixed to the stars only relative to a stated epoch.

The core insight

Right ascension $\alpha$ and declination $\delta$ are longitude and latitude on the celestial sphere, referenced to the celestial equator (the projection of Earth's equator) and the vernal equinox (where the ecliptic crosses the equator going north); the coordinates are fixed to the stars only relative to a stated epoch.

Where this lives today

The equatorial system is the live coordinate language of every observatory on and off the planet. The Gaia mission has catalogued nearly two billion stars in $(\alpha,\delta)$ with per-star precision approaching tens of microarcseconds — a level at which precession, nutation, aberration, and even relativistic light-bending by the Sun must be modelled explicitly. The fixed reference for those positions is the International Celestial Reference System (ICRS), realized by a few thousand distant quasars whose positions, fixed by very-long-baseline interferometry to better than a milliarcsecond, do not precess — there is no bulge for anything to torque. Every "GoTo" telescope still solves the astronomical triangle of equation (3) to slew from catalogue $(\alpha,\delta)$ to here-and-now altitude and azimuth, and time-domain surveys such as Vera Rubin's LSST report every transient as an $(\alpha,\delta)$ pair with an epoch — cross-matching them against Gaia and SDSS is precisely the great-circle separation of equation (6), $\cos\delta$ factor and all.

The physical picture

Take the terrestrial latitude–longitude grid and inflate it like a balloon until it lands on the inside of the celestial sphere. The equator becomes the celestial equator, the poles the celestial poles, lines of latitude become circles of constant declination, and lines of longitude become the "hour circles" of constant right ascension running pole to pole. A star's declination $\delta$ is then its latitude on the sky, and its right ascension $\alpha$ its longitude, counted eastward from an agreed zero: the vernal equinox, where the celestial equator and the ecliptic (the Sun's annual path, tilted by the obliquity $\varepsilon \approx 23.44^\circ$ ) cross. The first figure lays the ecliptic across the equatorial grid so the whole scaffolding is visible at once.

Figure

The equatorial grid unrolled: right ascension α runs 0ʰ to 24ʰ eastward, declination δ from +90° to −90°. The celestial equator is δ = 0; the ecliptic (the Sun's path) crosses it at the vernal equinox Υ going north, reaching δ = +23.44° = ε at the June solstice. A sample star sits at α = 16ʰ, δ = +45°.

The catalogue $(\alpha,\delta)$ is a global label; what any one observer sees depends on where they stand, and the second figure connects the two. At latitude $\phi$ the north celestial pole sits fixed at altitude $\phi$ above the northern horizon — which is why Polaris hangs $\sim\!40^\circ$ up from the mid-north — while the celestial equator crosses the meridian at altitude $90^\circ-\phi$ in the south. Every star wheels on a diurnal circle parallel to the equator: those near the pole trace small circles that never set (circumpolar), while far-southern ones never rise. A star reaches its greatest height as it transits the meridian, at altitude $a = 90^\circ - |\phi-\delta|$ , and the whole quantitative apparatus reduces to spherical triangles drawn in this picture.

Figure

The observer's meridian at latitude φ = 40°: the north celestial pole sits at altitude φ = 40°, the celestial equator crosses the southern sky at altitude 90° − φ = 50°, stars with δ ≥ +50° are circumpolar (never set), and a star of δ = +23° transits the meridian at altitude a = 90° − (φ − δ) = 73°.

The core equations

Everything the system computes flows from a units convention, the hour-angle bridge, and a few consequences of the spherical law of cosines. We start with the convention that trips up newcomers — right ascension is an angle told as a time, since the sky sweeps $360^\circ = 24^{\mathrm{h}}$ per sidereal day.

1^{\mathrm{h}} = 15^\circ, \qquad 1^{\mathrm{m}} = 15', \qquad 1^{\mathrm{s}} = 15'' \qquad (1)

where: $^{\mathrm{h}},^{\mathrm{m}},^{\mathrm{s}}$ — hours, minutes, seconds of right ascension (time); $^\circ,',''$ — degrees, arcminutes, arcseconds of arc. Reading: $\alpha$ runs $0^{\mathrm{h}}$ to $24^{\mathrm{h}}$ and $\delta$ from $-90^\circ$ to $+90^\circ$ . A time interval in RA equals a true angle only on the equator — the caveat that returns as the $\cos\delta$ factor below. The bridge from the fixed catalogue to the turning sky is the hour angle: how far west of your meridian an object lies, in time units — the difference between local sidereal time and its right ascension.

H = \mathrm{LST} - \alpha \qquad (2)

where: $H$ — hour angle, west of the meridian in time units; $\mathrm{LST}$ — local sidereal time, the hour angle of the vernal equinox; $\alpha$ — right ascension. Reading: on the meridian $H = 0$ , so $\mathrm{LST} = \alpha$ — a sidereal clock displays the RA of whatever is transiting. With $H$ in hand, the astronomical triangle (pole–zenith–star) converts $(\delta, H)$ and the latitude $\phi$ into a height above the horizon — the master transformation of positional astronomy.

\cos z = \sin\phi\,\sin\delta + \cos\phi\,\cos\delta\,\cos H \qquad (3)

where: $z$ — zenith distance, $z = 90^\circ - a$ with $a$ the altitude; $\phi$ — geographic latitude; $\delta$ — declination; $H$ — hour angle. Reading: this one equation contains rise, set, transit and everything between; the cases below are just special values of $H$ . Setting $H = 0$ (meridian crossing) collapses equation (3) to the height a star reaches at upper transit.

a_{\mathrm{transit}} = 90^\circ - |\phi - \delta| \qquad (4)

where: $a_{\mathrm{transit}}$ — altitude at meridian transit; $\phi$ — latitude; $\delta$ — declination. Reading: the star culminates south of the zenith if $\delta < \phi$ , north of it if $\delta > \phi$ ; the closer $\delta$ to $\phi$ , the higher it climbs. Setting instead $a = 0$ (the horizon) gives the hour angle at which an object rises or sets.

\cos H_0 = -\tan\phi\,\tan\delta \qquad (5)

where: $H_0$ — hour angle at the horizon; $\pm H_0$ mark setting and rising; $\phi$ — latitude; $\delta$ — declination. Reading: if $|\tan\phi\,\tan\delta| > 1$ there is no solution — the object is circumpolar ( $\delta \ge 90^\circ-\phi$ ) or always below the horizon ( $\delta \le -(90^\circ-\phi)$ ). Finally, comparing two positions requires the great-circle separation — another law of cosines, reducing for close pairs to a flat-sky Pythagoras with the $\cos\delta$ compression on the RA axis.

\cos\theta = \sin\delta_1\sin\delta_2 + \cos\delta_1\cos\delta_2\cos(\alpha_1-\alpha_2), \qquad \Delta\theta \approx \sqrt{(\Delta\delta)^2 + (\Delta\alpha\,\cos\delta)^2} \qquad (6)

where: $\theta$ — angular separation; $(\alpha_i,\delta_i)$ — the two coordinates; $\Delta\theta$ — small-angle separation; $\cos\delta$ — the factor shrinking an RA interval into a true angle away from the equator. Reading: near the poles a large spread in $\alpha$ is a tiny distance on the sky — why RA is ill-determined there. And because the frame slowly turns under precession, the coordinates themselves drift at a predictable annual rate — the reason every position carries an epoch.

\frac{\mathrm{d}\alpha}{\mathrm{d}t} = m + n\sin\alpha\tan\delta, \qquad \frac{\mathrm{d}\delta}{\mathrm{d}t} = n\cos\alpha \qquad (7)

where: $m = p\cos\varepsilon \approx 3.075^{\mathrm{s}}\,\mathrm{yr}^{-1}$ and $n = p\sin\varepsilon \approx 20.0''\,\mathrm{yr}^{-1}$ are the precession constants from $p \approx 50.29''\,\mathrm{yr}^{-1}$ , $\varepsilon \approx 23.44^\circ$ . Reading: RA creeps forward for essentially every star, faster near the poles; declination shifts fastest near $\alpha = 0^{\mathrm{h}}$ or $12^{\mathrm{h}}$ .

Derivation AAltitude of a star from its equatorial coordinates

The most important result in the subject is the transformation from a star's declination $\delta$ , the latitude $\phi$ , and the hour angle $H$ to its altitude above the horizon. Everything observable — when a star rises, how high it climbs, when it sets — is a special case of this one formula, so it is worth deriving cleanly. The whole argument lives inside a single spherical triangle with corners at the pole $P$ , the zenith $Z$ , and the star $X$ .

Step 1. Identify the three sides as arcs of great circles. The arc $PZ$ (pole to zenith) is the co-latitude $90^\circ - \phi$ , since the pole stands at altitude $\phi$ ; the arc $PX$ (pole to star) is the co-declination $90^\circ - \delta$ ; the arc $ZX$ (zenith to star) is the zenith distance $z = 90^\circ - a$ . Why: recasting latitude, declination and altitude as arcs lets a purely geometric identity do the physics.

Step 2. Identify the included angle at the pole. The angle $\angle ZPX$ between the meridian (through $P$ and $Z$ ) and the star's hour circle (through $P$ and $X$ ) is precisely the hour angle $H$ . Why: the hour angle is the one time-dependent quantity, and it enters at exactly one vertex, cleanly separating "where the star is catalogued" from "what time it is".

Step 3. Apply the spherical law of cosines to side $ZX$ , opposite the angle $H$ :

\cos z = \cos(90^\circ-\phi)\cos(90^\circ-\delta) + \sin(90^\circ-\phi)\sin(90^\circ-\delta)\cos H

Why: the spherical law of cosines is the sphere's analogue of $c^2 = a^2 + b^2 - 2ab\cos C$ , relating one side to the other two and their included angle — exactly the data we have.

Step 4. Collapse the co-functions with $\cos(90^\circ-x) = \sin x$ and $\sin(90^\circ-x) = \cos x$ :

\cos z = \sin\phi\,\sin\delta + \cos\phi\,\cos\delta\,\cos H

Since $a = 90^\circ - z$ , this equivalently gives $\sin a$ . Why: the co-function identities turn a formula in polar distances into one written directly in the tabulated $\phi$ and $\delta$ . *Check: at meridian transit $H = 0$ , $\cos z = \cos(\phi-\delta)$ , so $z = |\phi-\delta|$ and $a = 90^\circ - |\phi-\delta|$ — reproducing the transit-altitude result exactly.*

Go deeper · Derivation B · 1 minThe condition for a star to be circumpolar

Derivation BThe condition for a star to be circumpolar

A star is circumpolar for an observer if it never sets — if even its lowest point of the night stays above the horizon. Which declinations qualify tells you at a glance which part of the sky is permanently visible from your latitude, and it explains why Polaris is a northern fixture while the Southern Cross is invisible from Europe. The argument is a one-line application of the geometry at lower culmination, when a star crosses the meridian below the pole.

Step 1. Write the altitude at lower transit. At lower culmination the star sits on the meridian below the pole, so its altitude is the pole's altitude $\phi$ minus the star's polar distance $(90^\circ - \delta)$ :

a_{\mathrm{lower}} = \phi - (90^\circ - \delta) = \delta + \phi - 90^\circ

Why: the star's diurnal circle is centred on the pole with angular radius $90^\circ-\delta$ , so its lowest point sits that far below the pole.

Step 2. Impose the never-set condition $a_{\mathrm{lower}} \ge 0$ , i.e. $\delta + \phi - 90^\circ \ge 0$ . Why: "never sets" means the deepest point of the nightly circle stays at or above the horizon.

Step 3. Solve for declination to get the circumpolar band:

\delta \ge 90^\circ - \phi

By the mirror-image argument, a star never rises if $\delta \le -(90^\circ - \phi)$ ; both conditions flip hemisphere for a southern observer. Why: the two thresholds carve the sky into a circumpolar cap, a never-visible cap, and a band that rises and sets — the observability map for any site. *Check: at the North Pole ( $\phi = 90^\circ$ ) the condition becomes $\delta \ge 0^\circ$ — exactly the northern celestial hemisphere is circumpolar, as it must be when the pole is overhead.*

Go deeper · Derivation C · 1 minThe precession drift of equatorial coordinates

Derivation CThe precession drift of equatorial coordinates

Because the frame is defined by the celestial equator and equinox, and both move as Earth's axis precesses, the coordinates of a genuinely fixed star crawl year after year. The annual drift rates show why a position is worthless without an epoch, and quantify the effect: seconds of time in $\alpha$ and arcminutes in $\delta$ per century, enough to miss a target in a narrow-field instrument. We treat precession as a slow westward rotation of the equinox along the ecliptic and project it onto the equatorial axes.

Step 1. State the driving motion. Lunisolar torque on Earth's bulge makes the axis trace a cone, carrying the equinox westward along the ecliptic at $p \approx 50.29''$ per year, tilted from the equator by the obliquity $\varepsilon \approx 23.44^\circ$ . Why: precession is a rotation about the ecliptic pole, so its effect on equatorial coordinates must be resolved through the obliquity that separates the two poles.

Step 2. Project onto the two equatorial axes to define the precession constants:

m = p\cos\varepsilon \approx 46.1''\,\mathrm{yr}^{-1} \approx 3.075^{\mathrm{s}}\,\mathrm{yr}^{-1}, \qquad n = p\sin\varepsilon \approx 20.0''\,\mathrm{yr}^{-1} \approx 1.336^{\mathrm{s}}\,\mathrm{yr}^{-1}

The component $m$ is the equinox's slide along the equator (an offset added to every star's RA); $n$ scales the direction-dependent terms. Why: an eastward longitude like $\alpha$ picks up the along-equator projection $m$ uniformly, while the tilt mixes in position-dependent terms through $n$ .

Step 3. Assemble the standard rigid-rotation result, valid over the decades-to-century spans where the functions of $(\alpha,\delta)$ can be held fixed:

\frac{\mathrm{d}\alpha}{\mathrm{d}t} = m + n\sin\alpha\tan\delta, \qquad \frac{\mathrm{d}\delta}{\mathrm{d}t} = n\cos\alpha

Why: the $\tan\delta$ term is why RA runs away fast near the poles, while $\cos\alpha$ makes declination drift vanish at $\alpha = 6^{\mathrm{h}}, 18^{\mathrm{h}}$ and peak at $\alpha = 0^{\mathrm{h}}, 12^{\mathrm{h}}$ . *Check: the full precession period is $360^\circ/p = 1{,}296{,}000''/50.29'' \approx 25{,}770$ yr — the classical $\sim\!25{,}800$ -year Platonic year.*

Limiting cases

The system's behaviour in extreme locations and directions is the fastest sanity check on the formulae.

Case	Result	Meaning
Observer at a pole	$\phi = \pm 90^\circ$	The celestial pole sits at the zenith; declination circles coincide with altitude circles and no star rises or sets — hour angle loses meaning as azimuth is undefined.
Observer at the equator	$\phi = 0^\circ$	The celestial equator passes through the zenith; every star rises and sets vertically, and all declinations are accessible over a year.
Object on the celestial equator	$\delta = 0^\circ$	The projection factor $\cos\delta = 1$ , so $1^{\mathrm{s}}$ of RA is a true $15''$ — RA intervals and arc coincide.
Object near a celestial pole	$\delta \to \pm 90^\circ$	Hour circles converge, $\cos\delta \to 0$ , and $\alpha$ becomes ill-determined; a tiny motion swings the right ascension hugely.

Common trapRight ascension is in time units, and a second of RA is not a second of arc

The classic error is to treat $\alpha$ as if it were in degrees. It is in time: $1^{\mathrm{h}} = 15^\circ$ , $1^{\mathrm{m}} = 15'$ , $1^{\mathrm{s}} = 15''$ . Worse, even after converting, a second of RA equals $15''$ of true angle only on the celestial equator. At declination $\delta$ the hour circles crowd together, so one second of time in RA subtends only $15''\cos\delta$ on the sky — forgetting this over-estimates separations near the poles by a factor of two at $\delta = 60^\circ$ , and without bound as $\delta \to 90^\circ$ . It is also why proper motions are quoted as $\mu_\alpha\cos\delta$ rather than $\mu_\alpha$ : only the product is a true angular rate.

Worked examples

1 · Observing Betelgeuse from mid-northern latitudes (fully worked)

Betelgeuse has $\alpha = 5^{\mathrm{h}}\,55^{\mathrm{m}}$ , $\delta = +7.4^\circ$ (J2000.0). For an observer at $\phi = +40^\circ$ , find (a) its RA in degrees, (b) whether it is circumpolar or ever-invisible, (c) its transit altitude, and (d) how long it stays up.

Step 1. Convert the RA. With $1^{\mathrm{h}} = 15^\circ$ and $1^{\mathrm{m}} = 0.25^\circ$ : $\alpha = 75^\circ + 13.75^\circ = \mathbf{88.75^\circ}$ .

Step 2. Observability class. The circumpolar threshold is $\delta \ge 90^\circ - \phi = 50^\circ$ , the never-rises threshold $\delta \le -50^\circ$ . At $\delta = +7.4^\circ$ Betelgeuse falls between them, so it is neither circumpolar nor permanently hidden — it rises and sets each day.

Step 3. Transit altitude. From equation (4), $a_{\mathrm{transit}} = 90^\circ - |40^\circ - 7.4^\circ| = \mathbf{57.4^\circ}$ , reached as it crosses the meridian south of the zenith (since $\delta < \phi$ ).

Step 4. Time above the horizon. From equation (5), $\cos H_0 = -\tan 40^\circ\,\tan 7.4^\circ = -0.109$ , so $H_0 = 96.3^\circ = 6.42^{\mathrm{h}}$ . The star is up for $2H_0 = 12.8^{\mathrm{h}} \approx \mathbf{12^{\mathrm{h}}\,50^{\mathrm{m}}}$ each sidereal day — a little over half the time, as expected for a near-equatorial star seen from the mid-north.

2 · The RA–declination compression, made concrete

Two stars share a declination $\delta = +60^\circ$ and differ in right ascension by exactly $\Delta\alpha = 1^{\mathrm{h}} = 15^\circ$ . What is their true angular separation?

Step 1. Small-angle form. Treating RA as a true angle would give $15^\circ$ , but hour circles crowd together away from the equator. With $\Delta\delta = 0$ , equation (6) gives $\Delta\theta \approx \Delta\alpha\,\cos\delta = 15^\circ \times \cos 60^\circ = \mathbf{7.5^\circ}$ — half the naïve value.

Step 2. Exact check. The full formula gives $\cos\theta = \sin^2 60^\circ + \cos^2 60^\circ\cos 15^\circ = 0.9915$ , so $\theta = \mathbf{7.48^\circ}$ — the small-angle result is good to a percent. The lesson: a full hour of RA spans only $7.5^\circ$ on the sky at $\delta = 60^\circ$ .

3 · Why coordinates need an epoch: precession over a lifetime

A star sits near $\alpha = 0^{\mathrm{h}}$ , $\delta = +20^\circ$ . How far do its coordinates drift over $T = 50$ years of precession?

Step 1. Right ascension. At $\alpha = 0^{\mathrm{h}}$ , $\sin\alpha = 0$ , so the direction-dependent term vanishes and $\mathrm{d}\alpha/\mathrm{d}t = m = 3.075^{\mathrm{s}}\,\mathrm{yr}^{-1}$ . Over $50$ yr, $\Delta\alpha = 153.8^{\mathrm{s}} \approx \mathbf{2^{\mathrm{m}}\,34^{\mathrm{s}}}$ .

Step 2. Declination. At $\alpha = 0^{\mathrm{h}}$ , $\cos\alpha = 1$ is maximal, so $\mathrm{d}\delta/\mathrm{d}t = n = 20.0''\,\mathrm{yr}^{-1}$ . Over $50$ yr, $\Delta\delta = 1000'' \approx \mathbf{16.7'}$ .

Step 3. Interpretation. Nearly three minutes of time and a quarter-degree in half a century dwarfs most instruments' fields of view. Quoting $(\alpha,\delta)$ without an epoch like J2000.0 is therefore meaningless — without the date it does not point anywhere in particular.

Physics of the portrait

The equatorial system is old enough that its history is its physics: the reason we date coordinates at all was discovered before the telescope, by the astronomer who first pinned the stars down with numbers.

Origins · 1 minHipparchus and how this was discovered

Hipparchus

c. 190 – c. 120 BCE · founder of positional astronomy

Hipparchus of Nicaea is the reason the equatorial system carries a date. Around 129 BCE he compiled the first known star catalogue of the Western world — roughly 850 stars with measured positions and a brightness scale that survives, essentially unchanged, as today's magnitude system. Checking his positions against older Babylonian and Greek records, he found a systematic shift: the stellar longitudes had all increased by about $2^\circ$ over a century and a half, as though the reference grid — the equinox itself — had slid westward against the fixed stars. This was the discovery of the precession of the equinoxes, the most important caveat of the whole system; he estimated its rate at not less than $1^\circ$ per century (the true value is about $1.4^\circ$ ). He had no dynamical explanation — that waited nearly two millennia for Newton's lunisolar torque — but he established the crucial fact that celestial coordinates are defined only relative to a moving frame, and therefore relative to a time. Ptolemy preserved the catalogue in the Almagest; Copernicus recast the drift as a wobble of the moving Earth; modern astrometry finally sidestepped the equinox by nailing the frame to quasars. The discipline of writing an epoch beside every position is his.

History

129 BCE

— Hipparchus compiles the first star catalogue and discovers precession of the equinoxes

150

— Ptolemy's Almagest tabulates over 1000 stars and codifies the equinox-based system

1543

— Copernicus recasts precession as a slow wobble of the moving Earth's axis

1687

— Newton derives precession dynamically from lunisolar torque on Earth's equatorial bulge

1998

— the ICRS replaces the dynamical equinox with a quasar-defined extragalactic frame at J2000.0

Going further

Texts. Carroll & Ostlie, An Introduction to Modern Astrophysics, 2nd ed., §1.3 · Binney & Tremaine, Galactic Dynamics, 2nd ed., Appendix B · Ryden, Introduction to Cosmology, 2nd ed., Ch. 2 · Meeus, Astronomical Algorithms, 2nd ed., Chs. 12–13 and Ch. 21 (precession) · Smart, Textbook on Spherical Astronomy, 6th ed., Chs. 2–3. The key results in this article are: the altitude formula $\cos z = \sin\phi\sin\delta + \cos\phi\cos\delta\cos H$ (from the spherical law of cosines on the pole–zenith–star triangle), with limits $a = 90^\circ - |\phi-\delta|$ at $H=0$ and $\delta \ge 90^\circ-\phi$ at $a=0$ ; the precession constants $m \approx 3.075^{\mathrm{s}}\,\mathrm{yr}^{-1}$ , $n \approx 20.0''\,\mathrm{yr}^{-1}$ from $p \approx 50.29''\,\mathrm{yr}^{-1}$ , $\varepsilon \approx 23.44^\circ$ , giving a precession period of $\approx 25{,}770$ yr; and the worked numerical results $\alpha = 88.75^\circ$ , $a_{\mathrm{transit}} = 57.4^\circ$ , $2H_0 = 12.8^{\mathrm{h}}$ , $\Delta\theta = 7.48^\circ$ , $\Delta\alpha \approx 2^{\mathrm{m}}34^{\mathrm{s}}$ , $\Delta\delta \approx 16.7'$ .

Prerequisites. The Celestial Sphere. Leads to. Horizontal and Ecliptic Coordinate Systems · Precession, Nutation and Epochs · Stellar Parallax and the Parsec.

Ask the tutor: Give me any star's $(\alpha,\delta)$ and your latitude, and I will tell you whether it is circumpolar, how high it transits, and when it is up tonight — or walk you through the astronomical triangle step by step.

Why is right ascension measured in hours instead of degrees?

Because the sky turns once per sidereal day, and it is enormously convenient to have the coordinate match the clock. As Earth rotates, objects cross your meridian in order of increasing right ascension, and a sidereal clock reads out the RA of whatever is transiting. Dividing the $360^\circ$ of the equator into $24$ hours makes $1^{\mathrm{h}} = 15^\circ$ , so you can tell at a glance how long until a target culminates — subtract its RA from the current sidereal time. Declination has no such rotational meaning, being a fixed latitude, so it stays in ordinary degrees, arcminutes and arcseconds.

What exactly is the vernal equinox, and why is it the zero of right ascension?

The vernal equinox is a direction, not a date, though it shares its name with one. It is the point where the Sun, moving north along the ecliptic, crosses the celestial equator — the ascending node of the two great circles, reached around March 20. Longitude on any sphere needs an origin; the sky's is this crossing point, a natural, physically defined direction tied to the geometry of Earth's orbit rather than an arbitrary star, and right ascension is counted eastward from it. The catch is that the equinox is not fixed: precession slides it westward along the ecliptic at about $50''$ per year, which is exactly why coordinates referenced to it must carry an epoch.

If catalogue coordinates are "fixed", why do they still change with precession?

The coordinates are fixed relative to the frame — the celestial equator and equinox — but the frame itself rotates. Precession is the slow coning of Earth's axis under lunisolar torque on its equatorial bulge, period about $25{,}800$ years. As the axis moves, the poles and equator move with it and the equinox slides along the ecliptic, so a genuinely fixed star acquires slowly changing $(\alpha,\delta)$ : a few seconds of time in RA and up to $20''$ in declination per year. Over a lifetime that is arcminutes of drift, far larger than modern pointing tolerances — so every position carries an epoch such as J2000.0 to say which orientation of the moving frame it refers to.

How do I tell whether a star will be visible from my location tonight?

Two quick tests from your latitude $\phi$ . First, the observability limits: if $\delta \ge 90^\circ - \phi$ the star never sets, and if $\delta \le -(90^\circ-\phi)$ it never rises. Second, for stars in between, $\cos H_0 = -\tan\phi\tan\delta$ gives the time above the horizon, $2H_0$ in sidereal time centred on transit. To know when, use $H = \mathrm{LST} - \alpha$ : the star transits when $\mathrm{LST} = \alpha$ , reaching altitude $a = 90^\circ - |\phi-\delta|$ . Those three formulae answer "is it up, how high, and when".

What is the difference between the equatorial and horizontal (alt-az) systems?

They answer different questions. The equatorial system gives an object a nearly permanent, observer-independent address $(\alpha,\delta)$ — perfect for cataloguing, since the same numbers work everywhere and for years. The horizontal system gives its instantaneous position for you, as an altitude and azimuth that change minute by minute and differ for every observer. The two are linked by the astronomical triangle: feed a star's $\delta$ , your latitude $\phi$ , and hour angle $H = \mathrm{LST} - \alpha$ into $\cos z = \sin\phi\sin\delta + \cos\phi\cos\delta\cos H$ , turning the fixed catalogue address into the here-and-now pointing that aims a telescope.

Why does the ICRS replace the equinox with quasars if the equinox has always worked?

Because the equinox moves and quasars effectively do not. The classical frame is dynamical: its axes follow Earth's equator and orbit, both of which precess, nutate and wobble, so realizing it means continuously modelling Earth's motion — and any model error corrupts every star's position. The International Celestial Reference System instead fixes its axes to a few thousand extragalactic radio sources (mostly quasars) so distant their proper motions are unmeasurable; their VLBI positions, good to sub-milliarcsecond, give a frame that does not rotate at all. The axes were aligned with the J2000.0 mean equator and equinox so the change is seamless in practice, but conceptually it is decisive: from a frame that turns with the Earth to one anchored to the distant universe.