Try the measurement with one finger
Hold a finger in front of a doorframe across the room. Close one eye, then the other. Your finger jumps sideways against the frame, although you have held it still. Each eye views it from a different place. Move your finger farther away and the jump shrinks.
Astronomers use the same geometry for stars. The separation between human eyes is much too small: even the nearest stars are immensely farther away than anything in a room. Earth provides a much longer baseline. Observe a star again after six months and the observer has moved to the other side of the Sun. A nearby star shifts slightly against the more distant sky. That change caused by the viewing position is parallax.
The angles are tiny. A milliarcsecond is one thousandth of an arcsecond, and an arcsecond is one three-thousand-six-hundredth of a degree. Stars tens of parsecs away wobble by only tens of milliarcseconds. Nobody watches them jump by eye; the measurement comes from repeated, precise positions.
Reading: Lindegren et al. · Bailer-Jones
A year draws a small ellipse
Earth moves around an approximately circular path. Seen from the direction of a star, that circle generally looks like an ellipse. For a star near the plane of Earth’s orbit, called the ecliptic plane, the ellipse becomes thin. In the plane itself it becomes a line. Looking towards an ecliptic pole presents the orbit almost face on, giving a circle.
The definition matters here. A star’s parallax is the semimajor axis of that little apparent ellipse. Positions seen from opposite sides of Earth’s orbit can differ by twice the parallax. Mistaking the whole excursion for the parallax makes the inferred distance too small by a factor of two.
The parsec is a distance unit built around this geometry. A star with a parallax of one arcsecond lies one parsec away. Halve the parallax and the distance doubles. The physical ruler is the size of Earth’s orbit, so the measurement works independently of how brightly the star shines.
The star has its own journey
A star usually fails to return to precisely its old position after one year. The Sun and other stars travel through the Galaxy. Their relative motion steadily changes the star’s direction. This continuing angular drift is its proper motion.
Add proper motion to parallax and the result is a loop that moves across the star chart. The repeated annual part follows the observer’s orbit. The accumulating part follows the relative motion of the star and Sun. Observations spanning several years distinguish these two patterns.
A large proper motion need not mean an unusually high physical speed. The same sideways speed sweeps out a larger angle when the star is nearby. Distance lets us convert the measured angular drift into a transverse speed.
Reading: Lindegren et al.
Depth changes what a star chart means
Gaia puts positions, parallaxes and proper motions into one catalogue. Two nearly overlapping stars can separate widely in a three-dimensional map. Conversely, a cluster can cover a broad area of sky yet contain stars at similar distances, moving in similar directions. The extra measurements make that distinction possible.
Real positions have errors. A distant star’s wobble may be smaller than the noise, and a fit can sometimes return a negative parallax. A negative measured parallax describes a noisy angular estimate; physical distance remains positive. Estimating distance then requires the measurement uncertainty together with assumptions about the distribution of stars.
Think of a constellation as a set of long sightlines beginning at Earth. Its familiar outline joins directions on the sky. Parallax lets us travel along those sightlines and ask where the stars actually sit.
Reading: Bailer-Jones · Gaia DR3 documentation release 1.3 — 14.2.2 Astrometry