Horizontal and Ecliptic Coordinate Systems

A celestial coordinate system is just a pair of angles that pins a direction to the sky, and the art of positional astronomy is choosing the right pair for the problem in hand. The equatorial system — right ascension and declination — is anchored to Earth's spin axis and rides with the stars, the natural catalogue frame. But two others are indispensable at the telescope and the ephemeris desk. The horizontal (alt-azimuth) system is glued to your horizon: altitude $a$ above it, azimuth $A$ around it. It is what a telescope on an alt-azimuth mount physically reads out, and it sets how much atmosphere the light must cross. Its price is locality — the same star sits at different altitudes for observers at different latitudes, and every star's $(a,A)$ slides continuously as the Earth turns. The ecliptic system is glued instead to the plane of Earth's orbit: longitude $\lambda$ along the Sun's annual path, latitude $\beta$ perpendicular to it. Because the planets and Moon orbit close to that plane, they hug $\beta\approx0$ , which is why Solar-System dynamics is cleanest here. All three frames share one celestial sphere and are linked by rotations through two characteristic angles: the observer's latitude $\phi$ ties horizontal to equatorial, and the obliquity of the ecliptic $\varepsilon\approx23.44^\circ$ ties ecliptic to equatorial. Every conversion below is a single application of spherical trigonometry to one triangle — the pole, the zenith, and the star.

Key takeaways

The horizontal system fixes coordinates to a particular observer's local horizon: altitude $a$ measures angle above the horizon, azimuth $A$ measures direction around it, and both change continuously as the Earth rotates.
Tied to place and instant, $(a,A)$ differ site to site and moment to moment — worthless without a stated observer location and time.
The ecliptic system uses Earth's orbital plane: longitude $\lambda$ eastward from the vernal equinox, latitude $\beta$ perpendicular; it is natural for the Sun, Moon, and planets, which sit near $\beta=0$ .
One angle links each pair of frames: latitude $\phi$ for horizontal↔equatorial, obliquity $\varepsilon\approx23.44^\circ$ for ecliptic↔equatorial.
Every transformation is the spherical law of cosines applied to the astronomical triangle (pole–zenith–star): $\sin a=\sin\phi\sin\delta+\cos\phi\cos\delta\cos H$ .
Key idea: each system picks a fundamental plane — horizon, celestial equator, or ecliptic — and one frame is carried into another by a rotation about the line of nodes through a single characteristic angle: $\phi$ for horizontal↔equatorial and $\varepsilon$ for ecliptic↔equatorial.

The core insight

Each coordinate system picks a fundamental plane (horizon, celestial equator, or ecliptic); coordinates in one frame are obtained from another by a rotation about the line of nodes through a single characteristic angle — latitude $\phi$ for horizontal↔equatorial and obliquity $\varepsilon\approx23.44^\circ$ for ecliptic↔equatorial.

Where this lives today

The astronomical triangle runs in the pointing computer of every large telescope built in the last fifty years. Modern giants (the VLT, Keck, the coming Extremely Large Telescope) sit on alt-azimuth mounts because a two-axis horizontal mount is far cheaper and stiffer than an equatorial one for a many-tonne mirror. The cost is that a star's $(a,A)$ change nonlinearly, so the control system solves $\sin a=\sin\phi\sin\delta+\cos\phi\cos\delta\cos H$ many times a second and — because the field rotates on an alt-az mount — spins a de-rotator to hold the image fixed. The same $z=90^\circ-a$ and $X\approx\sec z$ drive nightly scheduling at surveys like the Vera Rubin Observatory, which chases each field near culmination to minimize airmass and extinction. In the ecliptic frame, the Sun–Moon–planet ephemerides that predict eclipses and occultations are computed in $(\lambda,\beta)$ because those bodies hug $\beta\approx0$ ; and because Earth's orbital motion lies in the ecliptic, the parallax ellipses Gaia measures to microarcsecond precision are cleanest when referred to the ecliptic pole.

The physical picture

Stand on a flat field at night. The bowl of the sky meets the ground at the horizon; straight up is the zenith, straight down the nadir. To say where a star is, you give "that high" (altitude) and "over that way" (azimuth). That is the horizontal system — vivid, immediate, and utterly personal: valid only for your spot and this instant, because Earth's rotation drags every star along a diurnal circle and two friends a thousand kilometres apart disagree about the numbers. To make them computable you connect them to the star's fixed equatorial coordinates $(\alpha,\delta)$ and the sky's clock, the local sidereal time. The bridge is a single spherical triangle whose corners are the celestial pole, your zenith, and the star; naming its three sides and one included angle is the whole transformation, and the first figure is that triangle.

Figure

The astronomical triangle for an observer at latitude φ = 40°: pole P, zenith Z and star S have sides 90°−φ, 90°−δ and z = 90°−a, and the included angle at the pole is the hour angle H, so the spherical law of cosines gives sin a = sin φ sin δ + cos φ cos δ cos H

Now let the year turn instead of the night. Follow the Sun against the background stars and it traces one great circle through the zodiac, tilted $23.44^\circ$ to the celestial equator. That circle is the ecliptic — Earth's orbital plane projected onto the sky — and the tilt is the obliquity $\varepsilon$ . Measuring along it (longitude $\lambda$ ) and away from it (latitude $\beta$ ) is the ideal bookkeeping for the Sun, Moon, and planets, which all orbit near this plane. The single number $\varepsilon$ does everything: rotate the equatorial frame by $\varepsilon$ about the equinox line and you land on the ecliptic frame, and the Sun's declination — the driver of the seasons — is just its ecliptic longitude fed through that tilt. The second figure shows the tilt and the declination swing it produces.

Figure

The obliquity ε = 23.44° tilts the ecliptic against the celestial equator; the Sun rides the ecliptic, so its declination swings as sin δ = sin ε sin λ, reaching +23.44° at the June solstice and −23.44° at the December solstice

The core equations

Everything below is one identity applied twice — once with latitude $\phi$ , once with obliquity $\varepsilon$ . We begin with the two definitions that make the horizontal frame usable, then the master conversion, then the ecliptic pair. The zenith distance is the horizon frame's second natural angle: the complement of altitude, and the quantity that actually governs how much atmosphere the light traverses. It is measured from the zenith downward.

z = 90^\circ - a \qquad (1)

where: $z$ — zenith distance, angle from the zenith to the object; $a$ — altitude above the horizon. Reading: $z=0$ at the zenith, $z=90^\circ$ at the horizon; small $z$ is a high, good target. The airmass measures the path length through the atmosphere in units of the straight-up value. For a flat, plane-parallel atmosphere the geometry gives a clean secant, exact enough for $z\lesssim60^\circ$ and the workhorse of photometric reduction.

X(z)\approx \sec z = \frac{1}{\cos z} \qquad (2)

where: $X$ — airmass; $z$ — zenith distance. Reading: $X=1$ at the zenith, $X=2$ at $z=60^\circ$ ; near the horizon the true value stays finite ( $\approx38$ ) while $\sec z$ diverges, so the formula breaks down there. The hour angle is the sky's clock hand: how far west of the meridian the star has swung. It is what converts a fixed right ascension into a live pointing angle, using the local sidereal time.

H = \theta_L - \alpha \qquad (3)

where: $H$ — hour angle, westward from the meridian; $\theta_L$ — local sidereal time; $\alpha$ — right ascension. Reading: $H=0$ at transit, increasing $15^\circ$ per hour as the Earth turns. This is the master equation of the article — the altitude of any object from its declination, the observer's latitude, and the hour angle. It is the spherical law of cosines applied to the pole–zenith–star triangle, and an alt-azimuth telescope evaluates it continuously to track a star.

\sin a = \sin\phi\,\sin\delta + \cos\phi\,\cos\delta\,\cos H \qquad (4)

where: $a$ — altitude; $\phi$ — geographic latitude; $\delta$ — declination; $H$ — hour angle. Reading: on the meridian ( $H=0$ ) this collapses to the transit altitude $a=90^\circ-|\phi-\delta|$ . Altitude fixes how high the object is; azimuth fixes which way to face. The companion formula comes from the same triangle via the sine rule, with a cosine-rule partner needed to resolve the quadrant unambiguously.

\sin A = -\frac{\cos\delta\,\sin H}{\cos a},\qquad \cos A = \frac{\sin\delta-\sin\phi\,\sin a}{\cos\phi\,\cos a} \qquad (5)

where: $A$ — azimuth, measured from north through east; the other symbols as above. Reading: use both expressions (an `atan2`) to place $A$ in the correct quadrant; a single $\arcsin$ is ambiguous. Now the ecliptic pair. Rotating the equatorial frame by the obliquity about the equinox line turns ecliptic coordinates $(\lambda,\beta)$ into equatorial ones. The declination component reads directly off the rotation.

\sin\delta = \sin\beta\,\cos\varepsilon + \cos\beta\,\sin\varepsilon\,\sin\lambda \qquad (6)

where: $\delta$ — declination; $\beta$ — ecliptic latitude; $\lambda$ — ecliptic longitude; $\varepsilon$ — obliquity. Reading: on the ecliptic ( $\beta=0$ ) this is $\sin\delta=\sin\varepsilon\sin\lambda$ — the Sun's declination wave. The right-ascension component of the same rotation is a ratio of the two in-plane coordinates; take it as a two-argument arctangent to keep the quadrant.

\tan\alpha = \frac{\sin\lambda\,\cos\varepsilon - \tan\beta\,\sin\varepsilon}{\cos\lambda} \qquad (7)

where: $\alpha$ — right ascension; $\lambda,\beta,\varepsilon$ as above. Reading: numerator is $y_{\text{eq}}$ , denominator is $x_{\text{eq}}$ ; the signs of both place $\alpha$ in $0$ – $360^\circ$ . Finally, the obliquity itself is not a constant: gravitational torques from the Moon and planets grind it down slowly, so precise work carries its epoch dependence.

\varepsilon \approx 23^\circ 26' 21'' - 46.8''\,T \qquad (8)

where: $\varepsilon$ — mean obliquity; $T$ — Julian centuries from J2000.0. Reading: at J2000 ( $T=0$ ), $\varepsilon=23^\circ26'21''=23.4393^\circ$ , decreasing $\approx47''$ per century.

Derivation AThe master altitude equation from the astronomical triangle

The four transformation formulas of this article all descend from one object: the spherical triangle whose vertices are the north celestial pole, the observer's zenith, and the star. The goal is to derive the master altitude equation, $\sin a=\sin\phi\sin\delta+\cos\phi\cos\delta\cos H$ , by naming the three sides of that triangle and applying the spherical law of cosines exactly once — turning a picture into the formula an alt-azimuth telescope evaluates many times a second to track a star.

Step 1. Name the three sides as great-circle arcs. Pole to zenith is the co-latitude $90^\circ-\phi$ , because the celestial pole sits at altitude $\phi$ above the north horizon. Pole to star is the co-declination $90^\circ-\delta$ , since declination is measured from the equator. Zenith to star is the zenith distance $z=90^\circ-a$ . Why: spherical trigonometry works with arcs between points, and each of the three fundamental angles reappears as one arc of this one triangle.

Step 2. Identify the included angle at the pole. Between the arc to the zenith (which lies on the observer's meridian) and the arc to the star (the star's hour circle) the angle is exactly the hour angle $H$ . Why: the hour angle is defined as the dihedral angle at the pole between the meridian and the star's hour circle, so it is the triangle's natural pole angle with no conversion needed.

Step 3. Apply the spherical law of cosines for the side $z$ 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 law of cosines is the one identity relating three sides and one angle of a spherical triangle, and we have exactly those four quantities.

Step 4. Simplify with the co-function identities $\cos(90^\circ-x)=\sin x$ , $\sin(90^\circ-x)=\cos x$ , and $\cos z=\sin a$ . This yields $\sin a=\sin\phi\sin\delta+\cos\phi\cos\delta\cos H$ . Why: rewriting complements as their co-functions collapses the raw geometric relation into the compact form used in practice, with altitude appearing directly.

Step 5. The azimuth follows from the same triangle: the spherical sine rule gives $\sin A=-\cos\delta\sin H/\cos a$ (the sign fixing the north-through-east convention), and the cosine rule for the angle at $Z$ supplies $\cos A$ so the quadrant is unambiguous. Why: altitude alone does not say which way to point, and both trig forms are needed because a lone arcsine loses a quadrant. *As a quick check: on the meridian $H=0$ gives $\sin a=\cos(\phi-\delta)$ , so $a=90^\circ-|\phi-\delta|$ ; at $\phi=40^\circ$ the Sun at $\delta=20.15^\circ$ transits at $70.15^\circ$ .*

Go deeper · Derivation B · 1 minWhich stars are circumpolar and which never rise

Derivation BWhich stars are circumpolar and which never rise

Evaluating the altitude equation at the two extreme hour angles turns it into a pair of visibility rules that depend only on the observer's latitude and the star's declination. The goal is to derive the transit altitude $a_{\text{transit}}=90^\circ-|\phi-\delta|$ and, from the lowest point of the diurnal circle, the sharp thresholds $\delta>90^\circ-\phi$ for a star that never sets (circumpolar) and $\delta<-(90^\circ-\phi)$ for one that never rises — the geometry behind why Polaris is always up and Canopus is invisible from mid-northern latitudes.

Step 1. Set $H=0$ (upper culmination). Then $\cos H=1$ and $\sin a=\sin\phi\sin\delta+\cos\phi\cos\delta=\cos(\phi-\delta)$ , so the transit altitude is $a=90^\circ-|\phi-\delta|$ . Why: a star is highest as it crosses the meridian, so the maximum altitude is the natural anchor for its diurnal circle and for whether it clears the horizon at all.

Step 2. Set $H=180^\circ$ (lower culmination). Now $\cos H=-1$ and $\sin a=\sin\phi\sin\delta-\cos\phi\cos\delta=-\cos(\phi+\delta)$ , the lowest altitude reached during the night. Why: whether a star ever sets is decided at its faintest, lowest moment, which occurs on the meridian below the pole.

Step 3. Impose $a_{\min}\ge0$ , i.e. $-\cos(\phi+\delta)\ge0$ , which requires $\phi+\delta\ge90^\circ$ , that is $\delta\ge90^\circ-\phi$ . Such a star is circumpolar — it never dips below the horizon. Why: if the minimum of the diurnal circle is still above the horizon, the star is up all night every night, which is precisely the circumpolar condition.

Step 4. By north–south symmetry, stars with $\delta\le-(90^\circ-\phi)$ have their maximum altitude below the horizon and so never rise for that observer. Why: reflecting through the equator swaps "never sets" for "never rises," giving the southern cutoff for free. *As concrete examples: at $\phi=40^\circ$ the threshold is $|\delta|=50^\circ$ ; Polaris ( $\delta=89.3^\circ$ ) is circumpolar with transit altitude $90^\circ-|40^\circ-89.3^\circ|\approx41^\circ\approx\phi$ , while Canopus ( $\delta=-52.7^\circ<-50^\circ$ ) never rises — invisible north of latitude $37.3^\circ$ .*

Go deeper · Derivation C · 1 minRotating the equator into the ecliptic and the Sun's declination

Derivation CRotating the equator into the ecliptic and the Sun's declination

Because the ecliptic and equatorial frames share the same origin and the same $x$ -axis (the line to the vernal equinox), one is carried into the other by a single rotation through the obliquity $\varepsilon$ about that axis. The goal is to obtain the ecliptic-to-equatorial formulas (6) and (7) from that rotation, then, by setting $\beta=0$ , to prove that the Sun's declination is $\delta_\odot=\arcsin(\sin\varepsilon\sin\lambda_\odot)$ with a maximum of exactly $\varepsilon$ — so the $\pm23.44^\circ$ seasonal swing is a direct readout of the tilt.

Step 1. Write the ecliptic direction as a unit vector $(\cos\beta\cos\lambda,\ \cos\beta\sin\lambda,\ \sin\beta)$ , with $x$ toward the vernal equinox and $z$ toward the ecliptic pole. Why: turning the two angles into Cartesian components lets a rotation act by simple matrix multiplication instead of spherical bookkeeping.

Step 2. Rotate about the shared $x$ -axis by $\varepsilon$ to reach equatorial components: $x_{\text{eq}}=\cos\beta\cos\lambda$ , $y_{\text{eq}}=\cos\beta\sin\lambda\cos\varepsilon-\sin\beta\sin\varepsilon$ , $z_{\text{eq}}=\cos\beta\sin\lambda\sin\varepsilon+\sin\beta\cos\varepsilon$ . Why: the equinox line is common to both frames, so the only difference between them is the tilt of the fundamental plane, i.e. a rotation about $x$ through the obliquity.

Step 3. Read off the equatorial angles: $\sin\delta=z_{\text{eq}}=\sin\beta\cos\varepsilon+\cos\beta\sin\varepsilon\sin\lambda$ (equation 6) and $\tan\alpha=y_{\text{eq}}/x_{\text{eq}}=(\sin\lambda\cos\varepsilon-\tan\beta\sin\varepsilon)/\cos\lambda$ (equation 7). Why: declination is the elevation of the equatorial unit vector and right ascension its azimuth, so both drop straight out of the rotated components.

Step 4. Specialize to the Sun, which lies on the ecliptic, $\beta=0$ . Then $\sin\delta_\odot=\sin\varepsilon\sin\lambda_\odot$ , maximal at $\lambda_\odot=90^\circ$ where $\delta_\odot=\varepsilon$ . Why: the seasons, the tropics, and the $\pm23.44^\circ$ solar declination range are all the obliquity seen through this one substitution. *Two numerical checks confirm the formulas: at $\lambda_\odot=60^\circ$ , $\delta_\odot=\arcsin(\sin23.44^\circ\sin60^\circ)=20.15^\circ$ and $\alpha=\arctan(\tan60^\circ\cos23.44^\circ)=57.82^\circ=3^\text{h}51^\text{m}$ ; at $\lambda_\odot=90^\circ$ , $\delta_\odot=23.44^\circ$ exactly.*

Limiting cases

Special positions make the transformations transparent and are the fastest sanity checks on any conversion code:

Case	Result	Meaning
Observer at north pole ( $\phi=90^\circ$ )	$a=\delta$ , azimuth tracks $H$	The pole sits at the zenith; altitude equals declination and every star circles at constant height — nothing rises or sets.
Observer at equator ( $\phi=0^\circ$ )	$a_{\max}=90^\circ-\lvert\delta\rvert$	The poles lie on the horizon; every star rises straight up, culminates, and sets vertically, up half the time.
Object on the ecliptic ( $\beta=0$ )	$\sin\delta=\sin\varepsilon\sin\lambda$	Declination oscillates between $\pm\varepsilon$ as $\lambda$ runs $0$ – $360^\circ$ — the Sun's annual declination wave.
Equinoxes ( $\lambda=0^\circ,180^\circ$ )	$\alpha=\lambda,\ \delta=0$	On the celestial equator the ecliptic and equatorial longitudes coincide and declination vanishes.

Common trapAzimuth has no universal zero — always state the convention

The single most common blunder in horizontal coordinates is quoting an azimuth without saying where it is measured from. Astronomers conventionally count azimuth from north through east ( $0^\circ=$ N, $90^\circ=$ E, $180^\circ=$ S), but navigators, surveyors, and much geographic software also count from north, while older astronomical tables and some ephemeris codes count from south, differing by exactly $180^\circ$ . A sign or a half-turn error here sends a telescope to the wrong side of the sky. The deeper point is that a horizontal coordinate is meaningless on its own: $(a,A)$ is a statement about one observer at one instant, so it must always be tagged with a location and a time (through the sidereal time and hence the hour angle). Equatorial coordinates $(\alpha,\delta)$ carry no such baggage — they are what you archive; horizontal coordinates are what you compute on the spot.

Worked examples

1 · Transit altitude and airmass of Sirius (confidence builder)

Sirius has declination $\delta=-16.72^\circ$ . From a site at latitude $\phi=+40^\circ$ , how high does it climb at best, and how much atmosphere does its light cross there?

Step 1. Sirius is highest at meridian transit, $H=0$ , so use the rule from Derivation B: $a_{\text{transit}}=90^\circ-|\phi-\delta|=90^\circ-|40^\circ-(-16.72^\circ)|=90^\circ-56.72^\circ=\mathbf{33.3^\circ}$ . Lying well south of the equator, it never rises far for a mid-northern observer.

Step 2. The zenith distance at transit is $z=90^\circ-a=56.72^\circ$ . The plane-parallel airmass is $X\approx\sec z=1/\cos56.72^\circ=1/0.549=\mathbf{1.82}$ .

Step 3. Interpretation: even at culmination Sirius' light crosses $1.8$ times the straight-up air column, so it is dimmed and reddened more than a near-zenith star and twinkles hard — the famous colour-flashing of Sirius low in the sky. An observer at $\phi=+70^\circ$ would find it barely clearing the horizon at $3.3^\circ$ . Best altitude 33.3°, airmass 1.82 — a low, heavily-extincted target from 40°N.

2 · The Sun's place on 21 May: ecliptic → equatorial → horizontal

Around 21 May the Sun's ecliptic longitude is $\lambda_\odot\approx60^\circ$ (and $\beta=0$ ). Take $\varepsilon=23.44^\circ$ and an observer at $\phi=+40^\circ$ . Find the Sun's declination, right ascension, and its altitude and azimuth at local apparent noon.

Step 1. Declination from equation (6) with $\beta=0$ : $\sin\delta_\odot=\sin\varepsilon\sin\lambda_\odot=\sin23.44^\circ\sin60^\circ=0.3978\times0.8660=0.3445$ , so $\delta_\odot=\arcsin(0.3445)=\mathbf{20.15^\circ}$ . The Sun is well north of the equator, climbing toward its June-solstice maximum of $\varepsilon=23.44^\circ$ .

Step 2. Right ascension from equation (7) with $\beta=0$ : $\tan\alpha=\tan\lambda_\odot\cos\varepsilon=\tan60^\circ\cos23.44^\circ=1.589$ . Both $\sin\lambda,\cos\lambda>0$ , so $\alpha$ is first-quadrant: $\alpha=\arctan(1.589)=57.82^\circ=\mathbf{3^\text{h}51^\text{m}}$ .

Step 3. Local apparent noon is the Sun's meridian transit, $H=0$ , so apply the transit rule: $a=90^\circ-|\phi-\delta_\odot|=90^\circ-|40^\circ-20.15^\circ|=90^\circ-19.85^\circ=\mathbf{70.15^\circ}$ . Because $\delta_\odot<\phi$ , the Sun culminates south of the zenith, so its azimuth is due south, $A=180^\circ$ .

Step 4. Airmass at noon: $z=19.85^\circ$ , so $X\approx\sec19.85^\circ=1.06$ — near-minimal air, which is why solar irradiance and UV peak around noon. δ⊙ = 20.15°, α = 3ʰ51ᵐ, noon altitude 70.15° due south, airmass 1.06.

Physics of the portrait

The coordinate systems in this article were forged over two millennia of watching the sky turn. The person who first welded measurement to a coordinate frame — cataloguing stars in ecliptic longitude and latitude and reading the obliquity off the solstices — was Hipparchus of Nicaea.

Origins · 1 minHipparchus and how this was discovered

Hipparchus

c. 190 – c. 120 BC · founder of the ecliptic coordinate system and mathematical astronomy

Hipparchus is the first astronomer we know to have combined systematic observation with the mathematics to make sense of it, and the coordinate machinery of this article is largely his legacy. Working from Rhodes in the second century BC, he compiled a catalogue of roughly 850 stars — the first in the Greek world to give each a numerical position — and expressed those positions in ecliptic longitude and latitude, the frame natural to a sky-watcher tracking the Sun, Moon, and planets along the zodiac. This demanded the obliquity of the ecliptic, which he took near $23^\circ51'$ from the Sun's solstitial shadows, and the beginnings of trigonometry: he tabulated chords (the ancestor of the sine) so that problems on the sphere became calculation. His most celebrated discovery came from comparing his ecliptic longitudes with older Babylonian records — the slow drift of the equinox along the ecliptic that we now call the precession of the equinoxes, about $1^\circ$ per human lifetime. That is why coordinates must be tagged with an epoch, and why the vernal-equinox zero-point of both systems is a moving target. Ptolemy's Almagest later preserved his catalogue; medieval Islamic astronomers refined his obliquity and trigonometry into the spherical toolkit behind the astronomical triangle used here.

History

240

BC — Eratosthenes brackets the obliquity of the ecliptic near 23°51′ from solstitial shadow lengths

129

BC — Hipparchus catalogues ~850 stars in ecliptic coordinates and discovers the precession of the equinoxes

150

— Ptolemy's Almagest standardizes ecliptic longitude and latitude and the spherical-astronomy methods

1437

— Ulugh Beg's Samarkand zij remeasures the obliquity as 23°30′17″

1851

— precision alt-azimuth and transit instruments make horizontal coordinates the basis of meridian surveying

Going further

Texts. Carroll & Ostlie, An Introduction to Modern Astrophysics, 2nd ed., §1.3 · Green, Spherical Astronomy, chs. 2–3 · Meeus, Astronomical Algorithms, 2nd ed., chs. 12–13 (the transformation and obliquity formulas used here) · Binney & Tremaine, Galactic Dynamics, 2nd ed., coordinate-systems appendix · Smart, Textbook on Spherical Astronomy. Standard numerical values: $\varepsilon=23^\circ26'21''=23.4393^\circ$ at J2000, $-46.8''$ /century; $\sin\varepsilon=0.3978$ , $\cos\varepsilon=0.9175$ . Key worked-example results: Sirius ( $\delta=-16.72^\circ$ ) transit $33.3^\circ$ , airmass $1.82$ from $\phi=40^\circ$ ; Sun at $\lambda_\odot=60^\circ$ gives $\delta_\odot=20.15^\circ$ , $\alpha=3^\text{h}51^\text{m}$ , noon altitude $70.15^\circ$ ; Canopus ( $\delta=-52.7^\circ$ ) invisible north of $37.3^\circ$ .

Prerequisites. The Celestial Sphere · Equatorial Coordinate System · Precession, Nutation and Epochs. Leads to. Optical Telescopes and Their Optics · Stellar Parallax and the Parsec · The Cosmic Distance Ladder.

Ask the tutor: Give me any star's right ascension and declination and your latitude, and I will walk you through its rise, transit altitude, azimuth, and whether it is circumpolar — one triangle at a time.

Why do we need three coordinate systems instead of one?

Because each is glued to a different physical plane and unbeatable for its own job. Equatorial is anchored to the celestial equator and equinox, which ride with the stars, so it is nearly time-independent and ideal for archiving positions — a catalogue frame. Horizontal is anchored to your local horizon, so it tells a telescope where to point and how much air the light crosses — a pointing frame, at the price of changing every second and differing site to site. Ecliptic is anchored to Earth's orbital plane, so it is simplest for the Sun, Moon, and planets, which all lie near it — an ephemeris frame. Since all three describe one sphere, a few rotations convert freely among them; you pick whichever makes your problem trivial.

Why does the azimuth convention cause so many errors?

Because there are two widely used zero-points that differ by exactly a half-turn, and nothing in the number tells you which is meant. Most modern astronomy measures azimuth from north through east ( $0^\circ=$ N, $90^\circ=$ E, $180^\circ=$ S). But a long tradition — and some ephemeris software and older tables — measures from south, running $180^\circ$ out of step. Feed a south-referenced azimuth into a north-referenced pointing model and the telescope swings to the wrong side of the sky. The fix is discipline: state the convention explicitly, and in code use the two-argument form (both $\sin A$ and $\cos A$ , an `atan2`) so the quadrant is never ambiguous.

Why do the planets and the Moon stay so close to the ecliptic?

Because they orbit in planes only slightly tilted to Earth's own, and the ecliptic is Earth's orbital plane projected onto the sky. The Solar System formed from a single flattened disk, so the major planets share nearly the same orbital plane; seen from Earth they never wander far from the ecliptic, staying within small ecliptic latitudes $\beta$ (the Moon within about $5^\circ$ , most planets within a few degrees). That is why the zodiac — the band of constellations straddling the ecliptic — is where the "wanderers" are found, and why ecliptic coordinates make Solar-System prediction so simple: the interesting bodies all live near $\beta=0$ , so one coordinate is almost redundant.

Does the obliquity of the ecliptic really change over time?

Yes, slowly and measurably. Torques from the Moon and planets on Earth's equatorial bulge make the equator–ecliptic tilt drift, and the mean obliquity decreases by about $46.8''$ per century: $\varepsilon\approx23^\circ26'21''-46.8''\,T$ with $T$ in Julian centuries from J2000. Over millennia $\varepsilon$ oscillates between roughly $22.1^\circ$ and $24.5^\circ$ on a $\sim41{,}000$ -year cycle — one of the Milankovitch cycles that pace the ice ages, since the obliquity sets the strength of the seasons. On top of this secular change sits nutation, a small periodic nodding (chiefly $\sim9''$ over $18.6$ years) driven by the Moon. Precise coordinate work therefore always specifies an epoch.

Why does the airmass formula X ≈ sec z break down near the horizon?

Because $\sec z$ comes from pretending the atmosphere is a flat, infinite slab, and that fails once the sightline is nearly tangent to the curved Earth. In the plane-parallel model the path grows as $1/\cos z$ and diverges at $z=90^\circ$ — but the real atmosphere is a thin curved shell, so a horizon-grazing ray crosses only about $38$ airmasses, not infinitely many. The secant law is excellent for $z\lesssim60^\circ$ and adequate to about $70^\circ$ ; below that observers switch to fitted formulas such as Kasten & Young's that include curvature. Two effects pile on near the horizon: refraction lifts an object's apparent altitude by up to half a degree, and extinction reddens and dims it severely — which is why serious photometry avoids low-altitude targets.

What exactly is the "astronomical triangle," and why does it keep appearing?

It is the spherical triangle whose three vertices are the north celestial pole, the observer's zenith, and the star (the PZS or navigational triangle). Its sides are the co-latitude $90^\circ-\phi$ , the co-declination $90^\circ-\delta$ , and the zenith distance $z=90^\circ-a$ ; its angle at the pole is the hour angle $H$ and its angle at the zenith is the azimuth $A$ . Every conversion between horizontal and equatorial coordinates — altitude, azimuth, rise/set times, transit altitude, circumpolarity — is one application of the spherical laws of cosines and sines to this single triangle. That is why it is the shared workhorse of celestial navigation and telescope pointing: master its six parts and you can turn any $(\alpha,\delta)$ into a "where do I look" for any observer at any instant.