Stellar radial velocity: curve, semi-amplitude K and minimum mass
The radial velocity curve of a star with a planet, its semi-amplitude K, the minimum mass m sin i, and how it compares with a spectrograph’s precision.
Radial velocity
the reflex orbit and the curvePlayback covers one whole period in twelve seconds of screen time, that is 0.352 442 d of orbit per second.
- Semi-amplitude
- K = (2π/P)^(1/3) · μp sin i / (μ* + μp)^(2/3) / √(1 − e²) = 59.05 m/s
- Radial velocity
- vr(t) = K [cos(ν + ω) + e cos ω] = 59.05 m/s, at ν = 0°
- Reflex orbit
- a* = a · μp / (μ* + μp) = 3434.2 km
Note
The star and the planet are never drawn in the same frame: the orbit of the planet is 2228.86 times larger than that of the star.
the reflex orbit and the measured curvethe answer, and the position at time tthe quoted precision of the spectrograph
Result
continuous computation- Peak to peak2K
- 118.101 m/s
- Maximumvmax
- 59.0505 m/s
- Minimumvmin
- −59.0505 m/s
- Midpoint of the extremesK e cos ω
- 0 m/s
- Quoted precisionσ
- 13 m/s
- Signal over precision|vr|/σ
- 4.54234
The curve reaches 59.0505 m/s and leaves the band of 13 m/s, which it exceeds by a factor 4.54234.
- In Jupiter massesm sin i
- 0.47
- In Earth massesm sin i
- 149.379
- True mass, in Jupiter massesmp
- 0.47
Mass function: f(m) = 1.19749 × 10¹⁰ m³/s², from the curve alone.
- Reflex orbita*
- 3434.2 km
- Projected by sin ia* sin i
- 3434.2 km
- Relative orbita
- 0.051 189 2 au
- Ratio of the orbitsap/a*
- 2228.86
- Radial velocityvr
- 59.0505 m/s
- True anomalyν
- 0°
- Orbital phaseφ
- 0
At 0 d from periastron, the star is receding from the observer.
Calculation, assumptions and sources
Provisional · Physical model · 6 sources · reviewed on 26 August 2026
OverviewThe "predict" mode starts from a planet and computes the radial velocity curve of its star and its semi-amplitude K.
The "predict" mode starts from a planet and computes the radial velocity curve of its star and its semi-amplitude K. The "weigh" mode starts from a recorded curve and computes the mass function, then the minimum mass m sin i. The figure draws v_r over one period with the instrument’s precision band and, above it, the reflex orbit of the star.
Method & formulasK = (2πG/P)^(1/3) · m_p sin i / (M + m_p)^(2/3) / √(1 − e²)
K is equation 12 of Lovis & Fischer (2010), taken up as equation 2.28 in Perryman (2018). The true anomaly ν comes from Kepler’s equation, solved numerically.
- Radial velocity
- the component of a body’s velocity along the line of sight, counted positive when it recedes.
- Semi-amplitude K
- half the gap between the two extremes of the curve, obtained by fitting the measurements. The peak-to-peak amplitude is 2K.
- m sin i
- the product of the planet’s mass and the sine of its orbital inclination. A radial velocity alone gives only this product.
- Mass function
- f(m) = (m_p sin i)³/(M + m_p)², computed from P, K and e alone. Converting it into a planetary mass requires the stellar mass.
- Argument of periastron (ω)
- the angle, in the orbital plane, between the line of nodes and the periastron of the star. It sets the shape of the curve, not its period or semi-amplitude.
- Reflex orbit
- the small orbit of the star around the barycentre, a scaled copy of the planet’s in the ratio of the masses, about 1 048 for Jupiter and the Sun.
- Systemic velocity (γ)
- the bulk velocity of the system relative to us, a constant added to the whole curve.
AssumptionsThe signal is the ideal Keplerian signal of a single planet: stellar jitter, photon noise, the sampling of observations and the superposition of several planets are not modelled.
- The signal is the ideal Keplerian signal of a single planet: stellar jitter, photon noise, the sampling of observations and the superposition of several planets are not modelled.
- The band drawn is a published instrumental precision, not a detection threshold. Deciding that a planet is detected requires a number of measurements, an error distribution and a statistical test, none of them computed.
- The mass of the star is an assumed input: a ten per cent error on it gives roughly seven per cent on the minimum mass m sin i.
- The systemic velocity is fixed at zero: an additive constant, it changes neither the shape nor the amplitude of the curve, nor the deduced masses.
- In "weigh" mode, the fixed point takes i = 90 degrees in the total mass term, following the convention of the literature. The result is the minimum mass.
Validity domain3 validity limits; the calculation refuses values outside 3 of them
5 ≤ incl ≤ 90model · refusal
The inclination runs from 5 to 90 degrees. At 90 degrees, edge-on, the minimum mass equals the true mass; below 5 degrees, sin i falls under 0.087 and the same curve would require a planet eleven times heavier.
Instead · For a nearly face-on orbit, astrometry replaces radial velocity.
0 ≤ ecc ≤ 0.95model · refusal
The eccentricity runs from 0 to 0.95, a display bound: beyond it, the curve becomes a spike too narrow for the 181 points of the figure.
Instead · At zero eccentricity, the curve is a sinusoid; this case is included.
0.01 ≤ per ≤ 100000practical · refusal
The period runs from 0.01 to 100 000 days. These are interface bounds; the real lower bound depends on the radius of the star, which is not entered.
Instead · For an orbit tighter than the stellar radius, tidal destruction (the Roche limit) applies.
The minimum mass m sin iA spectrograph measures only the projection of the star’s motion on the line of sight, a fraction sin i of the motion.
A spectrograph measures only the projection of the star’s motion on the line of sight, a fraction sin i of the motion. The curve contains only the product m sin i, a lower bound on the true mass since sin i is at most 1. A planet of 0.47 Jupiter masses may also be a brown dwarf of several tens of Jupiter masses seen nearly face-on. A transit, which forces an inclination near 90 degrees, or astrometry lifts the degeneracy.
What the figure showsAn animated spectral line would show nothing: the Doppler shift of a wobble of a few tens of metres per second is about two hundredths of a pixel on a detector.
An animated spectral line would show nothing: the Doppler shift of a wobble of a few tens of metres per second is about two hundredths of a pixel on a detector. Spectrographs measure it by correlation over thousands of lines.
The e cos ω termThe radial velocity is the time derivative of the star’s coordinate along the line of sight.
The radial velocity is the time derivative of the star’s coordinate along the line of sight. That coordinate is periodic, and the integral of its derivative over one period is zero. Without the constant term e cos ω, the time average of the curve would not be zero. The midpoint of the two extremes of the curve equals K e cos ω; it coincides with the average only for a circular orbit or when periastron is perpendicular to the line of sight. At zero eccentricity, ω loses its meaning (a circle has no periastron) and only shifts the phase of the curve.
A calculation without GThe published formulas contain G because they express masses in kilograms.
The published formulas contain G because they express masses in kilograms. Written with nominal gravitational parameters (solar and Jupiter masses, fixed by the International Astronomical Union in 2015), the semi-amplitude, the mass function and both semi-major axes no longer contain G. Its relative uncertainty, two parts in a hundred thousand, does not enter the calculation.
Example: 51 Pegasi b, the 1995 discoveryThe form is prefilled with the elements published by Michel Mayor and Didier Queloz (Nature, 1995) for the first planet discovered around a solar-type star: P = 4.2293 days, K = 59 m/s, a circular orbit, a star of one solar mass.
The form is prefilled with the elements published by Michel Mayor and Didier Queloz (Nature, 1995) for the first planet discovered around a solar-type star: P = 4.2293 days, K = 59 m/s, a circular orbit, a star of one solar mass. The "weigh" mode draws from them m sin i = 0.470 Jupiter masses, the value of the article; the "predict" mode, with that mass and an edge-on orbit, returns K = 59.05 m/s. The reflex orbit of the star has a semi-major axis of 3 434 km, and that of the planet is 2 229 times larger. The example precision, 13 m/s, is that of the ELODIE spectrograph used for the discovery; the signal exceeds it by a factor of four and a half. The period of 4.2308 days sometimes attributed to that article is the value refined by later follow-ups.
SourcesLovis, C. & Fischer, D. · Perryman, M. · Mayor, M. & Queloz, D. + 3 more
Lovis, C. & Fischer, D., Radial Velocity Techniques for Exoplanets, 2010.
Perryman, M., The Exoplanet Handbook, 2nd edition, 2018.
Mayor, M. & Queloz, D., A Jupiter-mass companion to a solar-type star, 1995.
International Astronomical Union, Resolution B3 on recommended nominal conversion constants, 2015.
Baranne, A. et al., ELODIE: A spectrograph for accurate radial velocity measurements, 1996.
Mayor, M. et al., Setting New Standards with HARPS, 2003.
Scientific validationProvisional, reviewed on 26 August 2026
Provisional · Sources and assumptions are provided; the documentary review is incomplete.
Physical model. The result follows the significant figures of the least precise input.
Revision log
Creation of the tool: radial velocity curve and semi-amplitude K of a star with one planet (Lovis & Fischer 2010 eq. 12, Perryman 2018 eq. 2.28), and an inverse mode giving the mass function, then the minimum mass m sin i.
For P = 4.2293 d, K = 59 m/s, e = 0 and one solar mass, m sin i = 0.470 Jupiter masses, the value of Mayor & Queloz 1995; the same minimum mass returns K = 59.050 m/s. Jupiter on the Sun gives 12.474 m/s at e = 0.0489 (12.459 m/s at zero eccentricity), the Earth on the Sun 0.089 458 m/s. The time average of v_r is zero to within 1.3 × 10⁻¹³ m/s over 56 (e, ω) pairs, for K = 59 m/s. The envelope of the sweep is K(1 + e) and not K(1 + e)/√(1 − e²), a bound 3.2 times too wide at e = 0.95; G cancels in every formula written with gravitational parameters.