Posted by Astronomers spent most of the twentieth century assuming other stars had planets and being unable to prove it. The instruments were too crude, the signals too small, and the few announced detections kept dissolving under inspection. Then the evidence started arriving, and it arrived fast. In about three decades the tally of confirmed exoplanets went from zero to several thousand, and the interesting question stopped being whether they exist at all.
The field has since reinvented itself. Counting exoplanets was the first era; describing them is the second. A telescope parked well beyond the Moon now pulls molecular fingerprints out of starlight that has passed through the air of a world a hundred light-years away.
The Wobble at Haute-Provence
The first planets found outside the Solar System were not the ones anybody wanted. In 1992 Aleksander Wolszczan and Dale Frail reported worlds orbiting PSR B1257+12, a pulsar: the spinning corpse of an exploded star, sterilized by radiation, about as far from a second Earth as a place gets. The detection was solid. It simply was not homely.
The one that changed everything came three years later. In October 1995 Michel Mayor and Didier Queloz, using the ELODIE spectrograph at the Observatoire de Haute-Provence, announced a planet around 51 Pegasi, a star broadly like the Sun. It carried roughly half the mass of Jupiter and finished an orbit in a little over four days, far closer in than Mercury, at temperatures that would melt lead. Nothing in the standard picture of planet formation allowed for it. Gas giants were supposed to form far out where it is cold, and stay put.
The models lost. Theorists rebuilt them around orbital migration, and the two discoverers collected a Nobel Prize in 2019.
Watching for a Shadow
Most of the catalogue comes from a simpler trick. If a planet’s orbit happens to be edge-on from here, it crosses in front of its star once per orbit and blocks a sliver of the light. Kepler, launched in 2009, exploited this with almost absurd single-mindedness: it stared at one patch of sky toward Cygnus and Lyra for four years, watching roughly 150,000 stars for dips that repeat on schedule. A Jupiter crossing a Sun-like star dims it by about one percent. An Earth-sized planet dims it by less than a hundredth of a percent.
It worked. Kepler-186f, announced in 2014, was the first Earth-sized planet found in the habitable zone of its star, a small red dwarf. Kepler-452b followed in 2015 around a star more like the Sun, though its size sits awkwardly between rock and gas and it has never been well characterized.
The geometry is unforgiving, though. Only a modest fraction of systems are tilted the right way, so the method samples a biased slice of the galaxy and says nothing about the rest. Kepler ran out of fuel in 2018. TESS had taken over with the opposite strategy: sweeping nearly the whole sky in sectors, favoring bright nearby stars that other instruments can follow up. TOI-700 d came out of that survey in 2020.
A transit gives you a radius and a period. It does not give you a mass, and without mass you cannot tell a rock from a puffball.
The Star Gives Itself Away
For mass you go back to the wobble. Star and planet both orbit their shared center of gravity, so the star traces a small circle and its spectral lines shift blue, then red, then blue again. Jupiter shoves the Sun around at roughly twelve meters per second. Earth manages about nine centimeters per second, slower than a snail, measured across trillions of kilometers. HARPS in Chile and ESPRESSO on the Very Large Telescope were built to chase exactly that.
The catch is that the technique yields a minimum mass, because the tilt of the orbit is usually unknown. Combine it with a transit and you get radius and mass together, and from those a density: the difference between an iron ball, a water world, and something mostly hydrogen.
Proxima b came out of this approach. In 2016 the Pale Red Dot campaign pulled an eleven-day signal from Proxima Centauri, the nearest star to the Sun at just over four light-years, consistent with a planet slightly heavier than Earth in the temperate zone. It does not transit from our vantage point, so the atmospheric tricks below do not apply to it. We know it is there. We know almost nothing else, and that gap is the honest state of most of the catalogue.
Two further methods fill in the corners. Direct imaging blocks the starlight with a coronagraph and photographs the planet itself, which works only for young, hot, massive worlds far from their stars; the giants of HR 8799 were captured that way from 2008. Gravitational microlensing catches the brightening when one star drifts precisely in front of another and focuses its light. The alignments never repeat.
A Zoo Nobody Ordered
The most useful thing exoplanets have done is embarrass our sample size of one. Hot Jupiters were the opening surprise. Then came the realization that the most common class of planet in the galaxy, the super-Earths and mini-Neptunes sitting between Earth and Neptune in size, has no representative in our own system at all. There is even a curious scarcity around 1.8 times Earth’s radius, apparently carved out by atmospheric escape.
Some planets orbit two stars at once, like Kepler-16b. Some sit so close in that their day sides are oceans of molten rock.
And then there is TRAPPIST-1, laid out in full in 2017: seven roughly Earth-sized planets around an ultracool dwarf some forty light-years away, packed so tightly that all seven orbits would fit inside Mercury’s, locked into a resonant chain that keeps them tugging at each other in rhythm. Several sit in the temperate zone. Worth being blunt about that phrase, because headlines rarely are. Earth-sized describes a radius. It implies no ocean, no breathable air, no soil.
What JWST Is Actually Doing Up There
Launched on Christmas Day 2021, JWST does something the survey telescopes never could. When a planet transits, a thin ring of starlight grazes its atmosphere on the way past, and the gases there swallow specific infrared wavelengths. Subtract one spectrum from the other and the leftovers name the molecules. The effect is measured in parts per million.
The early results were emphatic. In 2022 the telescope made the first unambiguous detection of carbon dioxide in another planet’s atmosphere, at the hot giant WASP-39b, and found sulfur dioxide there too, a molecule the star’s ultraviolet light manufactures on the spot. Photochemistry, watched from seven hundred light-years away.
K2-18b is the messier and more instructive story. It is a sub-Neptune roughly 124 light-years off, and in 2023 JWST found methane and carbon dioxide there, plus a faint hint of dimethyl sulfide, a compound that on Earth is produced almost entirely by marine microbes. The headlines wrote themselves. The science declined to cooperate: independent teams reanalyzing the same data found the sulfur signal was not statistically solid, a stronger claim in 2025 drew immediate pushback, and researchers still disagree about whether the planet is a warm ocean world or a hydrogen-rich body with no surface. Nobody knows yet. The argument is the process working.
It has also delivered results nobody enjoys reporting. Thermal measurements of the innermost TRAPPIST-1 planets point to little or no substantial atmosphere, which fits the old worry that flares from red dwarfs strip their close-in worlds bare. Red dwarfs are the most abundant stars in the galaxy, and most temperate planets we can study orbit them. If those worlds are routinely stripped, the list of promising addresses gets shorter.
An Earth twin around a Sun-like star stays out of reach for now. The signal is too faint for JWST, and it will take the next generation: the 39-meter ELT rising on a mountain in Chile, and a space observatory that for the moment exists as a design study rather than a launch date.
Where the Search Points Now
When Frank Drake wrote his equation in 1961, the fraction of stars with planets and the number of habitable worlds per system were pure guesswork. Exoplanets turned both into measured quantities, or at least bounded ones. That does not settle the equation, since the term for how long a technological civilization survives is still anybody’s guess.
It changed the tactics too. Project Phoenix spent 1995 to 2004 listening to roughly eight hundred nearby Sun-like stars picked on general principles. Modern campaigns work from the Kepler and TESS catalogues and point at systems where planets are already known. Breakthrough Listen has run since 2015 on Green Bank and other dishes, and its most publicized candidate, BLC1, was a narrowband signal from the direction of Proxima Centauri that survived months of scrutiny before being traced back to human equipment on Earth. The Wow! signal of August 1977 never repeated.
The hardware keeps shifting under the field. Arecibo collapsed in 2020, the Allen Telescope Array has been refurbished, piggyback searches now ride along on the VLA, and the Square Kilometre Array is going up across Australia and South Africa. Biosignatures and technosignatures are converging on the same target lists, which is the whole point.
If you want to follow that convergence as it happens, with its disputed spectra, its withdrawn signals and the next world that briefly looks like home, SETIworld tracks it week by week. Come and argue with us about what the data actually says.
