Astronomy · Measuring distance and motion

Proper motion versus annual parallax

Decompose an original sky-direction record into a local angular drift and an annual observer-driven term, while preserving the reference epoch, units and limits of a joint fit.

Continue in the approved chapter order. Necessary vocabulary, quantities and assumptions are explained in the lesson.

Back to the chapter

After reading, you can

Distinguish secular proper motion from annual parallax.
Calculate combined positions without requiring a yearly return to the same apparent point.
Explain the frame, epoch and covariance needed to interpret fitted angular motions.

The observer moves and the source’s direction evolves

Annual parallax describes an apparent directional contribution caused by the observer’s changing orbital position. Proper motion describes the source’s secular angular change in a stated reference frame. Over a limited interval it is commonly represented by a linear angular rate, distinct from the repeating parallax contribution.

Neither term should be confused with the daily sweep of the sky associated with Earth’s rotation. Observed apparent coordinates need appropriate observer and reference-system corrections before a catalogue motion model is fitted or compared across epochs.

A proper motion is an angular rate, such as milliarcseconds per year. It is not a physical speed in kilometres per second. Distance is required to convert a local angular drift into a transverse speed, as the next lesson explains.

A source can have both proper motion and annual parallax at once. Their superposition can produce an apparent track that never closes after one year. A repeated bend in a sky track is therefore not automatically evidence for an orbiting companion.

Position includes a reference epoch and two angular components

Use two local orthogonal angular axes, called u and v, in a small tangent-plane region of the sky. A reference epoch is the instant at which the position parameters are defined. Units, axis orientation and the reference frame must accompany those positions.

A first-order illustration combines a constant position, linear proper motion and a parallax parameter multiplied by known time-dependent parallax factors. The factors depend on the observer’s trajectory and source direction. They are not arbitrary additional proper motions.

For real right ascension and declination, the angular geometry also matters. A right-ascension coordinate rate is commonly reported with a cosine-of-declination factor when it represents the corresponding orthogonal tangential component. Components must use compatible definitions before they are combined.

The original example chooses local u and v axes directly, avoiding that coordinate ambiguity. It supplies a noiseless illustration over two Julian years, where one Julian year is 365.25 days of 86,400 SI seconds each. It is not an actual astrometric catalogue or scanning pattern.

Add the contributions at each epoch

Let t be time since the reference epoch, expressed numerically in Julian years. Stipulate u(t) = 120t + 20 cos(2πt) mas and v(t) = −40t + 10 sin(2πt) mas. The proper-motion components are +120 and −40 mas per Julian year.

The annual terms have semiamplitudes 20 and 10 mas. They can be represented by a 20-mas parallax parameter times factors cos(2πt) and 0.5 sin(2πt). An ideal circular observer orbit and a suitable sky direction can produce this ratio up to phase and orientation; no real mission trajectory is claimed.

At t = 0 the linear contribution is (0, 0) mas and the annual contribution is (20, 0) mas, so the combined position is (20, 0) mas. At t = 0.5, the drift gives (60, −20) and parallax gives (−20, 0), yielding (40, −20) mas.

After one full year the annual contribution returns to (20, 0) mas, but the drift has advanced by (120, −40) mas. The combined position is now (140, −40), rather than the initial point. At t = 2 it is (260, −80) mas.

Original local tangent-plane model. Positions are mas; time is Julian years from the declared reference epoch. Tiny numerical sine/cosine roundoff is represented as exact zero.
t / yr	Linear (u,v)	Annual (u,v)	Combined (u,v)
0	(0, 0)	(20, 0)	(20, 0)
0.25	(30, −10)	(0, 10)	(30, 0)
0.5	(60, −20)	(−20, 0)	(40, −20)
0.75	(90, −30)	(0, −10)	(90, −40)
1	(120, −40)	(20, 0)	(140, −40)
1.25	(150, −50)	(0, 10)	(150, −40)
1.5	(180, −60)	(−20, 0)	(160, −60)
1.75	(210, −70)	(0, −10)	(210, −80)
2	(240, −80)	(20, 0)	(260, −80)

A repeating signal plus continuing drift

Scroll sideways to see the full diagram. Its explanation is below.

Original noiseless local tangent-plane illustration over two Julian years since a declared reference epoch. The orthogonal local sky axes u and v are angular coordinates in milliarcseconds, rather than physical lengths. The model is u(t) = 120t + 20 cos(2πt) mas and v(t) = −40t + 10 sin(2πt) mas, with t expressed numerically in Julian years. One Julian year is 365.25 days of 86,400 SI seconds each. Solid blue combined curves, dashed orange linear drift and dotted green annual terms appear in separate time panels, each with one labelled mas axis. Both panels share time from 0 to 2 and the same pixels per mas; their drawn heights differ with the labelled angular ranges. Linear rates are +120 and −40 mas per Julian year. Annual semiamplitudes are 20 and 10 mas, which are angular contributions rather than physical orbit radii. Dense analytic curves are separate from the nine selected quarter-year fallback samples. Symbols mark synthetic model values; filled symbols at t = 0, 1 and 2 have the same annual phase. Combined positions are (20, 0), (140, −40) and (260, −80) mas at those times. Each one-year same-phase increment is (+120, −40) mas, even though the annual term repeats. An ideal circular observer orbit and a suitable source direction can give the stated parallax-factor ratio up to phase and orientation. This is a limited local first-order model, rather than actual Gaia data, a real scan law, proof that every wobble is parallax, an eternal linear path or an exact spatial source orbit. Real epochs, parallax factors, calibration, binarity, perspective effects and covariance require a joint model.

Repeating contributions do not require a repeating total

The figure separates the two coordinate time series and their linear and annual contributions. The annual term repeats in each panel, while the total acquires the secular drift. Dense curves follow the supplied functions; the quarter-year table is a set of selected points, not an exact piecewise-linear law.

Comparing the same annual phase at t = 0, 1 and 2 reveals the common drift increment (+120, −40) mas per year. Comparing arbitrary dates without removing their annual factors would mix that drift with parallax.

The two annual semiamplitudes are angular coordinate scales. They are not the physical axes of the observer’s orbit. Similarly, a projected ellipse in a parallax illustration is not the spatial orbit of the source.

A linear proper-motion description is local in time. Perspective effects, orbital companions, acceleration and reference or calibration changes can require a fuller model. Extrapolating a simple line indefinitely is not warranted by the two-year example.

Fit coupled parameters with their uncertainty

Real measurements sample finite times with uncertainty, often along instrument-dependent directions. Position, proper motion, parallax and calibration can be fitted together. Their estimates can be correlated because the same observations constrain several terms.

Retain the covariance information when propagating the fit to another epoch or inferring a physical quantity. Treating each reported uncertainty as independent can misrepresent the resulting uncertainty. More digits in a catalogue do not supply missing model conditions.

Before comparing two records, check their origin, reference orientation, epoch, component convention and whether they are apparent directions or fitted catalogue parameters. Agreement or disagreement has meaning only after the relevant transformation and calibration questions are addressed.

The next lesson combines a suitable distance and proper motion with signed radial velocity. That combination supplies spatial speed components in a declared frame, rather than reading a full physical motion directly from one angular track.

Worked examples

Add drift and annual displacement

What combined position does the model give at t = 0.5 Julian years?

Linear drift is (120 × 0.5, −40 × 0.5) = (60, −20) mas.
Annual terms are (20 cos π, 10 sin π) = (−20, 0) mas.
Adding gives (40, −20) mas.

The combined local angular position is (40, −20) mas in the stated axes and reference-epoch model.

Compare corresponding phases

Why do combined positions at t = 0 and t = 1 differ despite the repeated annual term?

The annual contribution is (20, 0) mas at both epochs.
The linear term changes from (0, 0) to (120, −40) mas.
The combined displacement is therefore (+120, −40) mas.

Parallax repeats while secular drift continues. Their sum need not return to the initial apparent position.

Try the reasoning

Does proper motion in mas per year by itself give kilometres per second?
Show the answer and reasoning
No.
Distance is required to convert angular drift to transverse speed.
Must the total apparent track close after one year whenever parallax is present?
Show the answer and reasoning
No.
Proper motion and other contributions can change the total while the annual term repeats.
Can two listed positions be compared fully without their epochs and coordinate conventions?
Show the answer and reasoning
No.
Motion propagation and reference definitions are required for a meaningful comparison.
Are fitted parallax and proper-motion uncertainties automatically independent?
Show the answer and reasoning
No.
Shared observations and calibration can correlate the estimated parameters.

What to carry forward

Annual parallax and proper motion describe different contributions to changing direction.
Their sum must be evaluated with a specified epoch, frame, axes and angular units.
Joint fitting, covariance and model limits govern how a measured track supports later physical inferences.

References

Teaching resources and research supporting this lesson’s explanations and examples.