The hard part is the glare, not the distance
By now you have spent whole rungs learning to read starlight: its color tells temperature, its Doppler shift tells motion toward or away, its brightness combined with distance tells true power output. A planet is the next thing to find with those same tools — but it hides a problem that distance alone does not capture. The naive worry is that other stars are far away, and they are: the nearest is over four light-years off. Yet the real obstacle is closer to home. A planet sits right next to a star that outshines it by a factor of millions to billions, and from light-years away the two blur into a single point. Seeing the planet is like spotting a firefly hovering beside a lighthouse, from the next city over.
That single fact organizes the whole field. Because the planet is lost in the glare, almost every method is *indirect*: instead of imaging the planet, we watch what the planet does to the star, or to the light passing by. Each trick exploits a different small effect, and — this is the point worth holding onto — each one measures a different physical quantity. None of them hands you a planet's full identity card. One gives a lower limit on mass, another gives a radius, another a true mass, another just a fleeting detection. Learning the methods is really learning which numbers you can trust from which technique.
The wobble: radial velocity
We say a planet orbits its star, but that is a half-truth you already have the tools to correct. From the gravity rung you know two bodies actually both orbit their shared [[center-of-mass|center of mass]] — the balance point of the pair. The star is far heavier, so its loop is tiny, but it is not zero: as the planet swings around, the star traces a small mirror-image circle, lurching slightly toward us and then away, over and over with the orbital period. We cannot see that lurch as motion across the sky, but we can hear it in the light. The [[radial-velocity-method|radial-velocity method]] measures the star's to-and-fro speed by watching its spectral lines blue-shift and red-shift in a rhythmic cycle — the same Doppler trick you met for stars, now turned into a planet detector.
The size of the wobble is breathtakingly small. Jupiter tugs the Sun around at only about 12 meters per second — a brisk jog — and Earth pulls it at roughly 9 centimeters per second, slower than a tortoise. Modern spectrographs can measure a star's speed to better than a meter per second, which is how this method works at all. What it returns is the period of the orbit (how long one cycle takes) and, from the speed of the wobble, an estimate of the planet's mass. But there is an honest catch: we only sense the part of the star's motion aimed along our line of sight. If the orbit is tilted, we see a smaller wobble than the full one, so radial velocity gives a *minimum* mass — a 'mass times an unknown tilt factor' — not the true value, unless another method pins down the tilt.
The dip: transits
Now suppose the orbit is tilted *just so*, edge-on enough that the planet passes directly in front of the star from our point of view. Once per orbit it blocks a sliver of the starlight, and the star dims by a tiny, repeating amount — a [[transit-method|transit]]. This is pure brightness measurement: no spectrum needed, just watch the star and wait for the regular little eclipse. The depth of the dip is wonderfully clean to interpret. A planet covers a fraction of the star's disk equal to the ratio of their areas, so the fractional drop in brightness is essentially (planet radius / star radius) squared. Measure the dip, know the star's size, and you read off the planet's [[stellar-radius|radius]] directly.
The numbers keep this honest. Jupiter crossing the Sun would dim it by about 1 percent; Earth would dim it by only about 0.008 percent — eighty parts per million, like noticing one streetlight in a stadium flicker out. That is why the workhorse instrument was a space telescope, NASA's [[kepler-mission|Kepler]], staring without blinking at one patch of sky above the blur of the atmosphere, and it is why transits found planets by the thousand. The catch mirrors radial velocity's: we only see the lucky few systems aligned nearly edge-on, so most planets never transit from our angle, and a transit alone gives radius but says nothing about mass.
Seeing it, lensing it, weighing the wobble in position
Three more methods round out the toolkit, each reaching where the first two cannot. [[direct-imaging|Direct imaging]] does the hard thing head-on: actually photograph the planet by blotting out the star's light with a mask called a coronagraph, the way you raise a hand to block the Sun and finally see a bird beside it. It works only for the easiest targets — young, hot, giant planets glowing in the infrared and orbiting far from their star, where the glare is least overwhelming — and it leans on adaptive optics or a space telescope to beat the blur. But when it works it is uniquely powerful: you get the planet's own light, which means a spectrum of the planet itself, and you can watch it move along its orbit over years.
The fourth method borrows a result from general relativity that you glimpsed on the gravity rung: mass bends light. In [[gravitational-microlensing|gravitational microlensing]], a foreground star drifts almost exactly in front of a far more distant background star, and its gravity acts like a lens, briefly brightening the background star. If the foreground star has a planet, the planet adds its own tiny, sharp blip on top of that brightening — a spike lasting hours to days. It is the strangest method of all: the signal is a one-time alignment that never repeats, so you cannot go back and confirm it. Its great gift is reach — it is sensitive to planets far from their stars, even free-floating planets bound to no star at all, and to systems thousands of light-years away. Its price is that you learn the planet's mass and rough orbit, but get no second look.
The fifth method is the wobble seen the *other* way. Radial velocity catches the star's to-and-fro along our line of sight; [[astrometry-method|astrometry]] instead measures the star's tiny shimmy *across* the sky — its position drawing a minuscule loop year after year. This is the same precise position-measuring craft as the parallax you used to find stellar distances, now pushed to even finer angles. Astrometry has a clean advantage: because it tracks the full two-dimensional loop, it can recover the true mass and the full orbit, with no tilt ambiguity. The reason it has found far fewer planets so far is brutal precision: the required angular shifts are mind-bendingly small, and only recently have missions like Gaia begun to deliver them at scale.
What each method actually measures
Lay the five side by side and the division of labor snaps into focus. The single most useful habit in this field is to read any exoplanet headline and immediately ask which method found it, because that tells you which number is real and which is inferred. The table below is worth keeping in your head — not the exact sensitivities, which keep improving, but the column that matters most: what each technique directly measures.
FINDING EXOPLANETS — WHAT EACH METHOD MEASURES
method watches... directly gives biased toward
------------------- -------------------------- -------------------- ----------------------
radial velocity star's to-and-fro speed minimum mass, period massive + close-in
transit star dimming as planet radius, period big planets, edge-on
crosses in front orbits
direct imaging the planet's own light orbit + a spectrum young, hot, wide giants
microlensing brief lensing brightening mass, rough orbit distant; far-out + free-
(one-time only) floating planets
astrometry star's loop ACROSS the sky true mass, full orbit massive + wide orbits
transit + radial velocity TOGETHER -> true mass AND radius -> density (rock? gas? ice?)
(sensitivities keep improving; the 'directly gives' column is the durable lesson)Notice the last column, the selection effect, because it shapes everything we think we know. Each method is biased toward the planets it finds *easily*: radial velocity and astrometry toward heavy planets, transits toward big ones in edge-on orbits, imaging toward young giants far out, microlensing toward distant and wide systems. So the early catalogs were full of hot Jupiters — giant planets whipping around their stars in days — not because such worlds are common, but because they make the loudest wobble and the deepest, most frequent dip. Small, far-out, Earth-like planets are the hardest of all to detect and the most undercounted. A census is only as honest as its awareness of what it cannot see.
Putting a method to work
To make this concrete, walk through how a real transit discovery actually gets nailed down. It is never one measurement and a press release; it is a chain of checks, each ruling out an impostor, because plenty of things dim a star without being a planet.
- Watch and wait. Monitor the star's brightness continuously and spot a small dip. One dip proves nothing — it could be a cosmic-ray hit, a passing cloud of instrument noise, or a stellar flicker.
- Demand a rhythm. See the same dip recur with a fixed period and identical shape. A real planet is a clock; it transits on schedule, every orbit, with the same depth and duration.
- Rule out the great impostor. The classic fake is an eclipsing pair of stars, where one star partly blocks the other and mimics a planet's dip. Check the dip's shape and depth, and look for a faint second, shallower dip — telltale signs of two stars rather than a planet.
- Confirm with a second method. Point a spectrograph at the star and look for the radial-velocity wobble at the same period. A matching wobble both confirms a planet and, since the transit already fixed the tilt, delivers the true mass to pair with the radius.
That chain is the field in miniature, and it carries the lessons of this guide forward into the rest of the rung. Each method is a partial view; confidence comes from agreement between independent methods; and every number is shadowed by the bias of how it was found. With those habits in hand, you are ready for what the worlds themselves turned out to be — the puffed-up hot Jupiters, the in-between super-Earths and mini-Neptunes that have no analog in our solar system — and, further along, the hardest and most human question of all: which of these worlds sits in the habitable zone where liquid water might endure, and how we could ever tell if anything lives there.