Exoplanet Detection Lab

δ = (Rₚ/Rₛ)² · radial velocity · fitting noisy data

δ = (Rp/Rs)² · radial velocity · fitting noisy data

You cannot see the planet.
You can see what it does to its star.

A planet around another star is a billion times fainter than the star and lost in its glare, so nearly every one known was found from the star’s light alone. If its orbit is lined up with us, the planet crosses the star’s disk once an orbit and dims it, by about one percent for a Jupiter and less than one part in ten thousand for an Earth: the size of the dip gives the planet’s size. And the star itself is pulled around in a small orbit of its own, so its light shifts back and forth in color by a few meters per second: the size of that wobble gives the planet’s mass. Real data are noisy. The trick is to fold many orbits together, fit a model, and know how far to trust what comes out.

Live: a hot Jupiter around a Sun-like star. From above (left), star and planet circle their center of mass, the star’s wobble drawn exaggerated; seen from Earth (right), the planet crosses the star’s disk. Below, the light curve dips once an orbit and the star’s radial velocity swings through ±134 m/s, both from the same orbit.

The transit gives the sizeA planet crossing its star blocks the fraction δ = (Rp/Rs)² of its light; the time between dips is the period, and Kepler’s third law turns the period into the orbit’s size.
The wobble gives the massStar and planet orbit their center of mass. The star’s Doppler shift swings by K, which gives the planet’s minimum mass Mp sin i; a transit fixes i, and both together give the density.
Signals hide in noiseFolding N transits on the period lowers the scatter as 1/√N; a fit judged by χ² gives each value and its uncertainty.
Not every dip is a planetEclipsing binaries, grazing orbits and spotted stars mimic planets; and both methods favor big planets close to their stars, so the planets found are not a fair sample.
Stay in touch