Science goals
What it is for
Three questions shaped the mission. Roman answers them statistically, by measuring enormous numbers of objects the same way.
Dark energy and the shape of cosmic history
Roughly seventy per cent of the universe is something we can only detect by its effect on expansion, and cannot otherwise explain.
Galaxy shapes
Weak gravitational lensing
Mass bends light, so every distant galaxy is very slightly distorted by the matter in front of it. The distortion is far too small to see in any single galaxy — it is recovered statistically from the shapes of hundreds of millions of them, mapping how matter clumped together over time.
Galaxy clustering
Baryon acoustic oscillations
Sound waves in the hot early universe froze into a preferred separation between galaxies. That fixed length is a ruler: measure how large it appears at different distances and you have measured the expansion directly. Roman's grism supplies the redshifts that turn positions on the sky into a three-dimensional map.
Standard candles
Type Ia supernovae
A particular kind of exploding white dwarf reaches a predictable peak brightness, so its apparent brightness gives its distance. This is the method that discovered acceleration in 1998; Roman extends it to fainter, more distant supernovae with a single well-calibrated instrument rather than a patchwork of telescopes.
The missing half of the planet census
Every method we have for finding planets is blind to cold worlds on wide orbits — the part of a solar system that looks most like our own.
Gravitational magnification
Microlensing
Sensitive to planets from roughly the mass of Mars upward, at separations of about one to ten astronomical units — and to planets bound to no star at all, which cannot be found any other way. Requires imaging a crowded field every twelve minutes for weeks on end.
Photometric dips
Transits, as a by-product
The same relentless cadence over the Galactic bulge produces light curves for enormous numbers of stars, so transiting planets fall out of the data alongside the microlensing events.
Coronagraphy
Direct imaging
The Coronagraph Instrument blocks a star's light to reveal what orbits it. On Roman it flies as a technology demonstration — proving on orbit the starlight suppression a future mission would need to image an Earth-like planet directly.
Where each method can see
Detection methods by orbital separation and planet mass, both logarithmic. The gap in the middle is the point — and it is where most of our own solar system sits.
Microlensing — Roman. Needs no light from the planet at all. Reaches cold, low-mass worlds at the separations every other method misses.
A general-purpose infrared survey of the sky
A survey this wide is not only an answer to the questions it was designed for. It is a dataset other people will use for questions nobody has asked yet.
Deep near-infrared imaging
The early universe
Light from the most distant galaxies is stretched out of the visible entirely. Surveying at 1–2 microns over thousands of square degrees is how you find the rare, bright, early objects that a narrow deep field would never happen to contain.
Resolved stellar populations
The Milky Way
Infrared cuts through the dust that hides the Galactic plane at visible wavelengths, and Roman's field is wide enough to map structure across the Galaxy while still resolving individual stars.
Repeated visits
The time domain
Anything that moves or varies shows up when the same field is imaged again — supernovae, variable stars, asteroids and distant solar system bodies drifting between epochs.
Status
Nothing here has happened yet. Every figure is a published forecast. NASA expects the first science images by early 2027.
Get an email when Roman's first science images are released
NASA expects them by early 2027. One email then, one when survey data starts flowing. Nothing else.