One dot per world
Every confirmed exoplanet in the NASA Exoplanet Archive, plotted at its host star’s real position on the sky. No artistic license: right ascension runs left across the map the way astronomers draw it, declination runs up, and each of the 6,324 dots is a planet we have actually measured orbiting one of 4,733 stars.
Drag to pan. Scroll or pinch to zoom. Click any dot to pull its record — discovery method, year, distance, radius, mass, temperature, orbital period. Recolor the whole census with the buttons, and tap a legend chip to light up one population against the rest.
LOADING CENSUS / 6,324 RECORDS
Data: NASA Exoplanet Archive (Caltech/IPAC, operated for NASA's Exoplanet Exploration Program), Planetary Systems table, default solutions. Retrieved 2026-07-26.
The first thing everyone notices is the dense amber block in the upper middle of the map. That is not a structure in the galaxy. It is a camera.
The map of our instruments
That block is the Kepler field — a single patch of Cygnus, about the size of your outstretched hand, that the Kepler space telescope stared at without blinking from 2009 to 2013. Roughly two out of every five planetary systems humanity knows sit inside it. We didn’t find planets there because Cygnus is special. We found them there because that’s where we pointed.
Switch the map to METHOD and tap through the chips. Each one is a different instrument, and each instrument finds a different galaxy:
- Transit (4,667 planets). The dip-in-starlight method. It needs an edge-on orbit and many repeated crossings, so it harvests close-in planets by the thousand — and it clusters wherever a survey telescope stared longest.
- Radial velocity (1,195). The wobble method, reading a star’s spectrum as planets tug it around. It needs bright, quiet, nearby stars, so its planets scatter evenly across the whole sky — the only population that looks the way you’d naively expect.
- Microlensing (281). A one-time gravitational coincidence: a star with planets drifts exactly in front of a more distant star and briefly magnifies it. It only works where the background is dense with stars — watch the chip light up a stripe hugging the galactic plane, pointed at the galactic center. These are among the most distant planets we know, and we will never observe any of them again.
- Imaging (98). Actual photographs. Only possible for young, self-luminous giants orbiting far from nearby stars — a rare and totally unrepresentative population.
- Timing & other (83). The oddballs, including the very first: the pulsar planets of 1992, discovered because they orbit a dead star whose radio metronome they disturb.
Five instruments, five contradictory galaxies. The atlas is their union, not a survey of what’s out there.
A time-lapse of a gold rush
Switch to YEAR and press play. Two planets in 1992. The lonely dot of 51 Pegasi b in 1995. A slow drip through the 2000s as radial velocity teams grind out a few dozen worlds a year — then 2014 detonates with 869 planets, and 2016 with 1,504, as statisticians validated Kepler’s backlog in bulk. Discovery stopped being an event and became a pipeline.
We found more planets in 2016 than in the first twenty years combined.
The rhythm since then — a few hundred a year, spread more evenly across the sky — is TESS, Kepler’s all-sky successor, plus ground surveys quietly industrializing the process.
What the atlas cannot show
Honest maps admit their gaps, and this census is mostly gap:
- F—01: 25% of these planets have no measured radius, and 51% no measured mass. In SIZE mode the “no radius” population renders deliberately gray — absence of data, drawn as absence of color.
- F—02: 73% have no equilibrium temperature. In TEMPERATURE mode the map runs hot red and cold blue but goes dark near Earth-like 255 K — partly because temperate worlds are genuinely hard to find, partly because we simply haven’t measured most of them.
- F—03: The median known exoplanet is about 1,200 light-years away, and the farthest 27,000 — in a galactic disk 100,000 light-years across holding an estimated hundred billion planets. In DISTANCE mode, watch the census fade with distance: we have catalogued a bubble, not a galaxy.
Statistically, every survey says planets outnumber stars. The 6,324 dots above are not the galaxy’s planets. They are the shadow our instruments cast on it — and the shadow is still growing by hundreds of worlds a year.
The next experiment
The obvious follow-up is depth: this map flattens the third coordinate the archive actually provides. The next build takes the same census into a navigable 3D neighborhood — with the distance axis honestly log-compressed, and labeled as such — so the Kepler field stops being a patch and becomes a cone.