What color is a red dwarf star?

I read recently that a red dwarf star might actually look white. I realize there's a wide range in red dwarf surface temperature, 2,500 to 3,500 degrees K, compared to 5,778 for our Sun. Metal at 2,500 degrees K is nearly white hot with just traces of yellow. Is that true for 93% hydrogen 7% helium as well? Would the color of red dwarf stars be closer to yellow/white rather than red based on our vision, if we were to be close enough to get a good look?

• "White" is not a color in the sense you want, I think. Better to ask what the peak wavelength is either in terms of energy or photon flux <-- and those are different! Aug 12 '20 at 15:41

The spectrum is of course heavily weighted towards red wavelengths, but your answer is mainly to do with the physiology of the eye, not astrophysics.

This site has folded stellar spectra through the appropriate response and arrived at "orange-ish" for M-dwarfs.

A "white-hot" piece of metal is not really white. Sunlight appears white - a blackbody spectrum at around 5800 K. The composition doesn't play a great role, though the spectrum of heated pure metal is probably closer to a blackbody than is an M-dwarf. I suspect that when people talk about "white hot metal", what is going on is the eye being overwhelmed by a large flux of photons. The intrinsic spectrum is clearly not that of a "white" object (such as sunlight). However, it could be that the emissivity is a bit lower in the red part of the visible spectrum so that when at least some green/blue light is produced (recall that the Wien tail is an exponential fall on the short wavelength side) that the eye's color sensory cells may be triggered sufficiently to fool it that the light is white. Either way that's physiology, not astrophysics.

Postscript: Most people can make out the orange-ish colour of M-type stars like Betelgeuse and Antares. These are (super)giants, but their spectra are not so different from M-dwarfs.

• What is your definition of white to consider a white-hot metal not really white? Aug 11 '20 at 13:38
• Emmisivity of liquid metals (and the gargantuous difference between them) is a fascinating study. Iron is "redder" than most. I once had the pleasure of witnessing aluminium / aluminium carbide at a blazing 2100 deg (C) - that was white enough to leave a nice shadow over my sight the next week. But iron doesn't glow nearly as "hot" as it is. So, bottom line: Not all metals are white hot when white hot, but some are! Aug 11 '20 at 14:01
• @StianYttervik you might be right. If there is a big hole in emissivity in the red part of the spectrum then it might be that the overall spectrum is flat enough that the little blue light that is produced ( a blackbody at 2400K produces a negligible fraction of blue/green light and no thermal processes can produce more light than a blackbody at any wavelength).matches the red light. Such objects would then have to be much less bright than a blackbody at the same temperature. Otherwise, all that you are describing is some physiology of the eye when confronted by high flux levels of light. Aug 11 '20 at 14:21
• @PeterCordes my point is that even direct sunlight at the top of atmosphere is "not really white" if we compare it to daylight that includes scattered sky light giving something like 6500 K that's taken as the sRGB white point. OTOH, if we consider the solar spectrum at the ToA as the definition of white, then the sRGB white point is not really white, being blue instead. White is a pretty ambiguous word when we talk about emissive colors. Aug 11 '20 at 19:53
• @Ruslan sunlight at zenith is very close to "white", as shown in the link I supplied. However, we do definitely know that blackbodies at temperatures below 3500K are certainly not white. I'm not arguing about eye physiology and perception and you can actually see the orange colour of nearby M-type stars (e.g. Betelgeuse, Aldebaran). These are giants, but M-dwarf spectra are not very different. Aug 11 '20 at 21:37

The apparent colour of a blackbody radiator in a CIE colour diagram is given by the Planckian locus. A pure blackbody spectrum for a given temperature, run through the standardized response of the human eye, will correspond to a point along the Planckian locus.

Of course, neither stars nor metal are perfect blackbody radiators so the colour may be a bit different when using a real spectrum, but I think it should give you an answer that's pretty close.

The following diagram (from the wikipedia page on Planckian locus) shows the Planckian locus as the curve going from 1000 K to infinity. As you can see, at around 5800 K, the temperature of our sun, it's pretty close to being white. At 2500 K, it looks to me like it would be an orange-yellow colour, with 3500 K giving a bright yellow, almost white colour.

The $$D_{xx}$$ points represent various daylight standards. As you can see, they're decently close to the Planckian locus, though I don't know whether a dwarf star would differ more or less from the Planckian locus than our sun.

The straight tick marks along the curve aren't really relevant to your question, but are for calculating the correlated colour temperature. Essentially, if you have some light source like an LED, you can find it's (u, v) coordinate and whichever line it lies along is the approximate colour temperature of that light source that you put on the box for selling it in a store.