HD 51608 c’s modelled colour is
azure #44acff:
bright water clouds scatter most of the starlight straight back.
Full spectrum: the colour computed from the complete 380–780 nm reflected-light model.
Roman 3-band: the same planet as the 's would see it — half-lit (, the only geometry a coronagraph can catch; fully lit would hide behind the star) and through just four .
Both are modelled, not photographs; flipping between them shows how much of a planet's colour survives Roman's coarse filter set.
Classic: surface colour and brightness taken straight from the physics model (cloud bands are schematic).
Stylised: that same colour restyled into warm/cool bands and turbulence for looks: prettier, not more accurate.
The palette below is identical either way; only the picture changes.
Modelled: the planet as our physics computes it should look: a smooth globe in its derived colour. Not a photograph.
Telescope: the actual processed image astronomers took of this planet: a faint point of light beside its (blocked-out) star. It is almost always and ; the real dot humanity has received, not a picture of the surface.
Only a handful of planets have ever been photographed directly; that is why this knob appears on some pages and not others. When more than one telescope has imaged the same planet (e.g. JWST and, later, Roman), a small tab picks between them.
Light source: same planet, same atmosphere — a different lamp. A planet's colour is its reflectivity times the light falling on it, so part of every colour here belongs to the star, not the planet.
Its own star: lit by HD 51608
(5358 K, modelled as a ).
The Sun: the identical planet re-lit by our Sun (5772 K) —
around the Sun this world would come out
#3ba6ff.
The shift is 1.0 (0 = identical, above ~2 is visible to the eye). Barely any change: this colour belongs to the planet itself, and its host is close enough to Sun-like that swapping lamps does nothing — that is the point.
Both are modelled, not photographed. The Roman 3-band channel has no Sun-lit version, so this knob acts on the full-spectrum view only.
Phase works like Moon phases: how much of the planet's lit side faces us. 0° is "full" (fully lit), 90° half lit, 180° "new" (backlit and dark). Both the brightness and the colour change as a planet wanes — and a coronagraph like Roman's can never see 0°, because a fully-lit planet sits right behind its star.
This planet's phase colours come from phase-resolved model spectra () — both the dimming and the colour shift are modelled.
#00335c
#0065b5
#0f95ff
#69bcff
#c2e4ff
#00325c
#0063b5
#0f92ff
#69bbff
#c2e3ff
#283429
#4f6651
#78967a
#aabdac
#dde4dd
#d9c9bc
the light this planet reflects
its real star:
At magnitude 8.2 this star is below the naked-eye limit (≈ 6.5 under a dark sky) — binoculars territory. It still sits at these coordinates, in Carina.
▸ Find HD 51608 in your sky tonight — every host star above your horizon right nowA planet's colour is mostly a consequence of where it is. Drag it inward and it heats up: methane breaks apart, cloud decks boil away, sodium starts eating the yellow out of its star's light. Drag it outward and everything freezes and condenses back. This slider re-runs the colour model at each distance and shows you the result.
Temperature is scaled from this planet's own as distance to the power of −½, holding the star fixed — so at its real orbit the slider reproduces the colour at the top of this page exactly, and moves away from there.
This is not a measurement. It is one model's opinion about a planet that is not where we put it.
The marks under the slider are the reference grid — independently published model spectra at 0.8, 2, 5 and 10 AU, the set the Roman Coronagraph community uses. Click one to jump there and compare. Where the two disagree, that gap is the honest size of the uncertainty — the grid models cloud condensation chemistry that our smooth model only approximates. The grid is computed for a Sun-like star, so its distances are used as published rather than rescaled for this host.
#44acff
where it really is
Its orbit is (e = 0.14), so it does not stay at one distance: it runs from 0.328 AU at its closest to 0.434 AU at its furthest, and its modelled colour changes as it goes. It moves fastest when closest, so it spends only near that scorching inner point.
This is not a second model — it is the slider above, driven by the clock instead of by your hand. The planet's distance around one orbit comes from solving Kepler's equation, in equal steps of time, which is why the loop dwells out in the cold and flashes through rather than spending half the year at each end.
The caveat that matters: every colour here assumes the atmosphere is instantly in balance with the starlight falling on it. Real atmospheres lag — clouds and chemistry take far longer to catch up than a fast periastron pass allows. Treat this as where the model points, not as a forecast of what a telescope would see on a given night.
Nobody has measured this planet's clouds or its . The pipeline has to assume both to produce a colour at all, and it says so on every page. Here you can turn those assumptions and watch what they were worth.
Each swatch changes exactly one assumption and re-runs the model; everything else, including the planet's real temperature and its star, is held fixed. The number under each is , the standard measure of how far apart two colours look: below about 2 the eye cannot tell, above about 10 they are plainly different colours.
A planet whose swatches barely move is one where the assumptions do not matter much. One where they swing wildly is a planet whose colour you should trust less — and that is worth being able to see rather than being asked to take on trust.
#44acff
| Base hex | #44acff |
| Lit by its own star | |
| Out of |
| 401 measured | |
| Radius | 3.87 measured |
| Mass | 14.3 measured |
| 0.381 measured | |
| Distance from Earth | 114.3 measured |
| Discovery | · 2019 |
△ TOO CLOSE — at 4.30× the starlight Earth receives, any surface water here would have boiled off long ago. The habitable zone lies further out, past 0.602 AU.