Every colour on this site is — worked out from what a planet is made of and what its star shines on it. Nobody has ever measured the visible colour of a planet around another star. The carries an instrument that could, for a few dozen planets at most. This page is the list of those planets, and a deliberately empty space beside each one, waiting.
A planet next to its star is like a firefly next to a lighthouse. Roman's blocks the starlight, but it only works in a ring around the star — roughly 150 to 450 thousandths of an arcsecond out. A planet has to be far enough from its star, and its star close enough to us, to fall in that ring. Almost none are.
Each planet below shows the colour we predict Roman's three filters would report. Beside it is a second, empty slot for the colour Roman actually . Those slots are empty because the measurements do not exist yet. The day one lands, it goes in the slot — and this site stops being entirely a prediction.
Roman's coronagraph is a , not a survey: it is there to prove the technique works. Which planets it observes, and when, is not a published schedule. This board shows who is eligible, drawn from the published study named at the bottom — not a promise that any particular planet will be looked at.
30 August 2026, 07:20 EDT
NASA set this date in June 2026, pulling the launch roughly eight months ahead of its earlier commitment of no later than May 2027.
The coronagraph is a technology demonstration, switched on after commissioning rather than on day one. NASA has not published fixed dates for individual target observations, so this board counts down to the launch — the one date that is announced — and nothing more.
No exoplanet has ever had its visible colour measured. Every swatch below is a prediction — which is exactly what makes the empty column worth watching.
Each planet's marker sits at the widest it can ever appear from its star, seen from Earth, measured in thousandths of an arcsecond. We work that out from the planet's orbit and its distance — both on its own page. The lit band is the ring Roman's coronagraph works in, 150 to 450.
Landing in the band is necessary, not sufficient. Orbits are tilted, and a planet spends much of its year closer in than its widest point, so being eligible on paper is not the same as being catchable on the night Roman looks. Working out those odds properly is what the study below did; this bar is the intuition, not the calculation.
The nearest of the lot, and the widest-swinging: at its furthest it opens out past the coronagraph's outer edge.
On an extraordinarily stretched orbit, which swings it right across the coronagraph's working range.
A young giant still embedded in its birth disc, far enough out that it glows with its own heat as much as it reflects.
The Archive's composite table calls this one HD 39091 b — the same planet under its catalogue name.
Pollux — the brightest star with a known planet, and one of the brightest stars in the sky.
The hardest case on the list: even at its widest it stays well inside the coronagraph's inner edge.
Not carried in the Exoplanet Archive's composite-parameters table for our data release — the signal has been contested since — so we have no parameters to model a colour from.
Likewise absent from the Archive's composite-parameters table for our data release, so there is nothing here to colour.
The eligible planets are Table 4 of Catalogue of exoplanets accessible in reflected starlight to the Nancy Grace Roman Space Telescope, Carrión-González et al. (2021), Astronomy & Astrophysics 651, A7 — planets around stars brighter than magnitude 7 with better than a one-in-four chance of being observable.
“At our reference wavelength λ = 575 nm, the number of Roman-accessible exoplanets in the optimistic, intermediate, and pessimistic CGI scenarios is 26, 10, and 3, respectively.”
The list below is that paper's Table 4 — planets whose host stars are brighter than V = 7 and whose chance of being observable exceeds 25% under optimistic assumptions. The paper does not enumerate which ten make up its intermediate scenario, so this board does not guess: it marks only the three it names as accessible under every scenario. Of those, 23 are in our catalogue and have a modelled colour here; 2 are not, and say why on their own row.
Launch date and vehicle: NASA. Roman's coronagraph works in an annulus — a 'dark hole' — from about 0.15 to 0.45 arcseconds around the star. Closer in, the starlight is not suppressed; further out, the instrument stops looking. Its tech-demo requirement is raw contrast better than 1 part in 10 million inside that annulus.
How we get the colours → Sources & credits → ◧ Browse all planets →