Astronomy & Celestial Mechanics
the sky, computed
Concept

Error propagation — where it appears

How the uncertainty in each ingredient carries into a derived quantity, adding in quadrature so that a chain of corrections can only make it larger. It is why a distance ladder's uncertainty is dominated by its weakest rung, and why adding a rung that is individually excellent can still make the final answer worse.

Named by 7 essays across 5 fields — each of them below, with the objects they name alongside it.

Ceres from five directions and no distance. Ceres seen five times over 41 days, from an Earth on a circular orbit, reduced in the plane. Each sighting gives a direction and no range, so the object is somewhere on its sight line; the five lines here span 1.37° of geocentric arc altogether, and Earth's own motion supplies the only baseline there is — 0.403 AU of its 0.691 AU of travel lies across the sight lines. A planar orbit is four numbers, so five angles over-determine it and one orbit comes out: a = 2.7658 AU, e = 0.0785. That is the answer and not an input — the sightings were generated at a = 2.7658 AU and e = 0.0785, and the solve, which sees only the directions and the dates, returns them to 3e-12. The two shaded sectors are what closes the determination: between the first and middle sightings the radius vector sweeps 0.2993 AU² in 21.0 days and between the middle and last 0.2853 AU² in 20.0 days, a ratio of 1.04918 against a time ratio of 1.04918. Slide all three crossings out along their sight lines together and that equality fails at once, so it fixes the distance by itself, with no propagation anywhere in the argument — and it gives a = 2.7658 AU over again. The two dashed curves are candidates that thread the same three sight lines at 85% and 108% of the recovered distance: a = 1.86 AU at e = 0.35, sweeping its areas in a ratio 4.5% wrong; and a = 5.61 AU at e = 0.49, sweeping its areas in a ratio 2.9% wrong. A few per cent in the distance is an orbit of another kind, which is the same fact the conditioning panel measures: one arcsecond of angle error moves a by 0.60% on this arc.

Five directions and no distance among them

An image of a moving point records an angle and throws the range away, so an orbit has to be assembled out of angles alone. How many angles are needed is not a detail of the method — it is the whole of what a determination is.

orbits · Orbit determination A distance of 52.0 parsecs with nothing underneath it. Two ways to a distance for the same pair. The orbital parallax needs no iteration and no assumption: a double-lined spectroscopic orbit gives the relative orbit's linear size as (K₁+K₂)P√(1−e²)/2π sin i = 0.2268 AU, an astrometric orbit gives its angular size as 4.36 milliarcseconds, and the ratio is 52.0 parsecs — a length divided by an angle, with no rung of the distance ladder below it and no property of the stars assumed. The curves show the dynamical parallax, the version available when only one spectrum can be measured: guess the mass sum, take the linear size from the harmonic law, divide by the angular size, convert the apparent magnitude to an absolute one and read a new mass sum off a mass–luminosity relation. Three starting guesses spanning a factor of 10 in mass converge to the same distance in 8 passes and agree to 0.001 per cent. It converges because the distance depends on the assumed mass only as its cube root — the measured exponent here is 0.3333 — so a factor of two in the mass is 26 per cent in the distance, and one pass removes most of that. What it converges to is not the orbital parallax: the iteration settles at 54.2 pc against 52.0, 4.2 per cent away, because the fixed point is set by the mass–luminosity relation and the apparent magnitude rather than by anything measured about this orbit. The same insensitivity that makes it converge is why it is never better than the relation it leans on.

Two orbits of one pair, and a distance falls out

Measure the same binary spectroscopically and astrometrically and the orbit comes back twice — once as a length in kilometres and once as an angle on the sky. The ratio is a distance that owes nothing to parallax, nothing to a standard candle, and nothing to any assumption about the stars.

stars · Binary stars A gap where planets should be. The number of planets per star per interval of log radius, for orbital periods under a hundred days, corrected for detection efficiency. There are two peaks — super-Earths near 1.3 R⊕ and sub-Neptunes near 2.4 R⊕ — and a deficit between them at 1.89 R⊕, where the occurrence falls to 33% of the peak. The gap is not a gap in what can be detected: detection efficiency rises smoothly through it, so a smooth underlying distribution could not produce a dip there.

The planets that were not seen

An occurrence rate is a count divided by a probability, and the probability can be a five-hundredth. Everything difficult about saying how common planets are lives in that denominator.

exoplanets · Occurrence rates The best-fitting eccentricity of a circular orbit. What a fitted eccentricity comes out at when the orbit's true eccentricity is 0 and each component of the eccentricity vector carries an error of 0.03. The distribution is not centred on the truth and cannot be: an eccentricity is the length of the vector (e cos ϖ, e sin ϖ), lengths are not negative, and a quantity bounded below by zero whose components scatter symmetrically has a distribution pushed away from the bound. For a circular orbit the most likely fitted value is exactly one error bar, 0.0300 here, and the mean is 0.0376 — 1.2533 error bars, which is √(π/2) and comes from geometry rather than from any property of the data. The practical consequence is a catalogue of small eccentricities that are all measurements of their own error bars, and the fix is not a better fit but a different question: an upper limit rather than a value.

An eccentricity that cannot be zero

Fit an orbit to noisy data and the eccentricity that comes back is never zero, not even when the orbit is a perfect circle. The reason has nothing to do with the data and everything to do with the fact that a length cannot be negative.

orbits · Orbit determination A step size below which a smaller step is worse. The error left in a long integration against the step size, for methods of three different orders, with both contributions drawn. The falling lines are truncation error, whose slope on these axes is exactly the order of the method. The rising line is round-off, identical for all three because it is a property of the arithmetic and not of the algorithm: every operation loses a few bits, the losses are independent, and they accumulate as the square root of the number of steps — which is why its slope is −1/2 and why it rises as the step shrinks. Each method's total has a minimum, at a step of 1.0e-6, 5.0e-6, 9.7e-4 for orders 1, 2, 4. Below that minimum every halving of the step costs time and makes the answer worse. That is the practical reason a solar-system integration is not run at an arbitrarily fine step, and it is a reason with nothing to do with computer time.

An error that grows like a random walk

A long integration accumulates two errors with opposite habits. One falls when the step is made smaller and grows in proportion to the time; the other grows when the step is made smaller and accumulates as a square root. Which of the two dominates decides whether a billion-year integration means anything.

gravitation · Numerical integration 4 per cent in one observable is 48 per cent in an age. How an error in the calibration of the large frequency separation propagates into the quantities derived from it. The two scaling relations are exact in their exponents, so a fractional error in the separation appears as twice that in the radius, four times in the mass, and — because a main-sequence lifetime falls as roughly the two-and-a-half power of the mass — ten times in an age. At the 4 per cent level, which is about what the theoretical corrections to the relation amount to for a red giant, that is 15 per cent in mass and 48 per cent in age. Nothing about the seismology is uncertain at that level; the frequencies are measured to parts in a thousand. What is uncertain is the constant of proportionality, and it is uncertain because it was calibrated on one star.

Two scaling relations calibrated on one star

Asteroseismology gives a star's mass and radius from two numbers read off its oscillation spectrum. The two relations are exact in their exponents and approximate in their constants, and the constants were fixed by requiring that the Sun come out right — so an error of a few per cent in one observable is tens of per cent in a mass and nearly a factor in an age.

stars · Asteroseismology A right angle short by 0.147°, and a ratio of 389 hanging on it. The ratio of the Sun's distance to the Moon's implied by the angle between them at the moment the Moon is exactly half lit, on a logarithmic scale, for angles from 80° to 89.95°. At that moment the angle at the Moon between the directions to the Sun and the Earth is a right angle, so the ratio is the secant of the observed angle — a construction with no distance in it, the same right triangle that gives an inferior planet's orbit from its greatest elongation. Aristarchus measured 87°, which gives 19.1. The mean distances give 389, which corresponds to 89.853°. The curve's steepness is the whole story: across a tenth of a degree centred on each marked angle the ratio changes by 3.4 per cent at 87°, 10.5 per cent at 89° and 103 per cent at the true angle. The method was exact and it asked for a right angle measured to a hundredth of a degree, at an instant the eye can judge only to within hours, on a terminator that is never quite straight.

A right angle short by a seventh of a degree

When the Moon is exactly half lit, the angle at the Moon between the Sun and the Earth is a right angle, so the angle seen from the Earth gives the Sun's distance in units of the Moon's. Aristarchus measured 87° and concluded the Sun was nineteen times further away. The construction was exact; the angle he needed was 89.85°, and at that angle a tenth of a degree is the whole answer.

sky · Apparent motion

Named alongside it

The objects these essays reach for when they reach for this one.

EccentricityAngular semi-major axisAnomalistic monthAstrometric orbitAstrometryAstronomical unit (AU)Benchmark starBrouwers lawCompensated summationCompletenessCovarianceDichotomy

All concepts