≡ Menu

WISE: First Ultra-cool Brown Dwarf

“To a man with a hammer, everything looks like a nail,” said Mark Twain, one take on which is that the way we see problems shapes how we see solutions. That fact can be either confining or liberating depending on how open we are to examining our preconceptions, but in the case of Amy Mainzer (JPL), it leads to a natural way to describe a failed star. Mainzer, who is deputy project scientist on the Wide-field Infrared Survey Explorer mission (WISE), is an amateur jewelry-maker. For her, it’s easy to look at the image below and see gems. “The brown dwarfs,” says Mainzer, “jump out at you like big, fat, green emeralds.”

And that emerald below, dead center in the image, is hard to miss.

Image: The green dot in the middle of this image might look like an emerald amidst glittering diamonds, but it is actually a dim star belonging to a class called brown dwarfs. This particular object, named “WISEPC J045853.90+643451.9” after its location in the sky, is the first ultra-cool brown dwarf discovered by NASA’s Wide-field Infrared Survey Explorer, or WISE. WISE is scanning the skies in infrared light, picking up the signatures of all sort of cosmic gems, including brown dwarfs. Credit: NASA/JPL-Caltech/UCLA.

The brown dwarf in question is somewhere between 18 and 30 light years away in the constellation Camelopardalis (the giraffe), and is one of the coolest such objects known, with a temperature of roughly 600 Kelvin (326 degrees Celsius). This is one of those brown dwarfs that burn at temperatures close to a hot oven here on Earth, cool enough that it takes WISE’s infrared view from space to pick it up. In the image, we’re looking at three of the four WISE infrared channels, color-coded so that blue shows the shortest infrared wavelengths and red the longest. The methane in the brown dwarf atmosphere absorbs the blue-coded light and the faint object gives off little of the red, leaving green as the dominant color.

The best news about the new brown dwarf is that it turned up a mere 57 days into the survey mission, meaning that WISE is on track to find many more. Given how hard they are to spot, the possibility of an ultracool brown dwarf being in the Sun’s neighborhood, and perhaps closer than the Alpha Centauri stars, cannot be ruled out. In any case, mission planners think WISE will find hundreds of the objects within a few parsecs of the Sun. And who knows, we may yet find a perturbing body whose presence accounts for anomalous orbits like that of Sedna.

Remember this: Back in June, we learned that the Spitzer Space Telescope had found fourteen of the coldest brown dwarfs then known, so faint that visible light telescopes could not find them. The Spitzer study focused on a region in the constellation Boötes, whereas WISE will be looking at the entire sky. Peter Eisenhardt, a WISE project scientist at JPL, puts it this way:

“WISE is looking everywhere, so the coolest brown dwarfs are going to pop up all around us. We might even find a cool brown dwarf that is closer to us than Proxima Centauri, the closest known star. WISE is going to transform our view of the solar neighborhood. We’ll be studying these new neighbors in minute detail — they may contain the nearest planetary system to our own.”

We’ve only been able to confirm the existence of brown dwarfs for fifteen years, and we certainly aren’t through classifying them. T dwarfs are defined as being less than about 1500 Kelvin (1226 Celsius), but the colder class of Y dwarfs, found in stellar models but not yet confirmed, makes for prime hunting for WISE, which should be able to detect them. Are there more brown dwarfs within 25 light years of the Sun than normal stars? WISE should be able to tell us, and in doing so may tune up our target list for future deep space missions into the Oort Cloud and beyond.


Comments on this entry are closed.

  • Istvan November 10, 2010, 12:51

    I am eagerly awaiting more of the WISE survey data. The possibility of finding a nearby stepping-stone target, easier to reach than Proxima Centauri, is in my opinion easily the most tantalizing prospect in all of space sciences today.

  • Alex Tolley November 10, 2010, 13:03

    Shouldn’t models of stellar births indicate the likely prevalence of brown dwarfs?

  • kzb November 10, 2010, 14:29

    Do we know what proportion of the sky has been examined in detail so far? Also I thought WISE data was being kept secret until the WISE team had first crack at important discoveries and publications?

  • andy November 10, 2010, 15:21

    One question is whether there is whether any new spectral features will become apparent to determine a new spectral type Y. One of the things that seems to come out from studying L and T dwarfs is that effects like gravity become more important for cooler objects: it is not necessarily the case that a cooler object will have a later spectral type.

  • Michael November 10, 2010, 15:42

    A brown dwarf at 4.5 Gyr’s old is going to be quite cool and hard to spot for sure

  • Phil November 10, 2010, 21:05

    This is just the start of a fascinating period! Just think of the potential for finding very odd and exotic planetary systems around some of these (hopefully) nearby brown dwarves.

    Problem is, do we have the equipment to do high precision RV and/or transit work on such faint and red objects? Is this going to be a job for James Webb (assuming it launches and works!), or might perhaps this be within the capability of some of the planned 10+metre terrestrial super-telescopes?


  • James M Essig November 10, 2010, 21:30

    Hi Folks;

    Such a cool brown dwarf could make an excellent IR heater for a large orbiting infrastructure, perhaps such as a Dyson Sphere.

    A brown dwarf with a temperature of between 100 K and 300 K located between here and the Proxima Centauri System may have gone undetected, but would be a great stepping stone to Apha Centauri.

    Istvan, I hear you on that. The completing of the WISE survey almost can’t come soon enough.

  • Michael Simmons November 10, 2010, 22:57

    I thought WISE had finished its full sky scan and that images of the entire sky are sitting on some ones hard disk somewhere.
    I would think a computer program looking for green dots would find all of them very fast…
    So the number of highly likely candiates (but without distance information and other follow up information) should already be known.

  • Adam November 11, 2010, 1:49

    Got to wonder if Ohmic heating could warm any close orbitting planets around a cool brown dwarf. Would be a strange world only dimly lit in the visible. Wonder if strange near-infrared photosynthesis could sustain an ecology? Reminds me of William Hope Hodgson’s “In the Night Land” which is set on a dark Earth after the Sun has cooled below visible light-levels. Based on archaic astrophysics, but Kelvin-Helmholtz contraction is still relevant to any brown dwarf planets.

  • Wedge November 11, 2010, 5:24

    @Phil, unfortunately performing exoplanetary analysis on these brown dwarfs is likely to be problematic. The most useful modern techniques (radial velocity & transit) are highly reliant on bright enough stars to get a clear signal, the dimmer the star the less signal there is. Sadly, many brown dwarfs, especially these, are extremely faint (many orders of magnitude less bright than the stars being studied for exoplanets) so there’s very little hope of using these techniques to study their possibly planetary systems.

    The best hope might be indirect studies of systems comprised of Sun-like stars with distant brown dwarf companions. However, there’s a bit of a catch-22 there as well. If the brown dwarf is too close to the Sun-like star it’ll likely not have its own planetary system, due to gravitational disruptions. However, the planets in a distantly separated binary system would have a very much smaller affect on the brighter star and so would be that much harder to detect via radial velocity methods. They would also be that much more unlikely to be detected via transit methods, since it would require a lengthy observation period combined with a lot of luck. The best luck would be catching several transits of the bright star in a row as a planet orbited the smaller dwarf, but there’s a pretty low chance of that happening.

  • djlactin November 11, 2010, 12:24

    If the brown dwarf is too close to the Sun-like star it’ll likely not have its own planetary system, due to gravitational disruptions.

    @ Wedge. Not sure about this: Jupiter is smaller than a BD and has its own “planetary system”…

  • kzb November 11, 2010, 14:17

    Michael Simmons, I was under the same impression. I’m confused as to why news of this single discovery has come out now. I hope this is not the closest BD to be discovered in the whole survey, because if it is, they are nowhere near as common as thought.
    Alex Tolley, most current publications on the IMF have the mass function turning down below spectral class M5. That implies that BD’s are not as common as M-dwarfs , and the probability is there is NOT one (or more) any nearer than Proxima.

    However, I was reading one paper recently which hinted there was the possiblity of a “pile up” at the bottom of the spectral distribution, of old cool BD’s. All BD’s eventually end up in this pile-up region as they age and cool. This acts to subtract BD numbers from the apparent IMF. Hope I’ve interpreted this correctly :)

  • James M Essig November 12, 2010, 9:48

    Hi Adam;

    It would indeed by strange for a bioshere world to be lit by dim visible light alone.

    It would be cool to stop by such a world where a large brown or perhaps brown-deep red orb would provide daylight. I would imagine that the angular diameter of a 1,000 K sun would need to be about {{[(5,800 K)/(1,000 K)] EXP 4} EXP (1/2)} (0.5 degrees) ~ 16 degrees to provide a temperate climate. A 750 K star would need to be about {{[(5,800 K)/(750 K)] EXP 4} EXP (1/2)} (0.5 degrees) ~ 30 degrees.

    The number of whimsically beautiful sun rises much be huge in our Galaxy alone.

    Regarding photo-synthesis using IR, some forms of lettice and other produce come close at having an optimal growing frequency within the red visible spectrum. Popular Science Magazine had an excellent urban sustainability article a few years back that discussed red light growing of certian produce to maximize product yield for a given amount of energy efficient illumination.

  • Ron S November 13, 2010, 0:23

    It would seem that brown dwarfs aren’t actually brown:

    “…a lot of these brown dwarfs actually would appear to be magenta to the eye. “

  • Ronald November 13, 2010, 14:15

    What kzb is saying in response to Alex Tolley about abundance of BDs is quite relevant; my understanding was also that the general trend is that the smaller the stellar mass the more abundant (e.g. red dwarfs being so much more common than sunlike stars, these in turn being much more common than white and blue giants, etc.). Apparently this trend does not hold below a certain minimum (M5?) and there may even be a minimum threshold, a cut-off point, below which there is no stellar formation.

    kzb, do you have a good publication in mind which describes this?

    Furthermore, I honestly do not see the great relevance of a BD stepping stone roughly in between here and Alpha Centauri. I mean, if we can travel 2 ly, we can also travel 4 ly. A whole chain of stepping stones might be a different matter.

    Finally, I think it is a bit premature to consider a BD even at 1 or 2 ly distance a stepping stone: it being at intermediate distance does not mean it is anywhere in the right direction!
    Madrid may be about half as far as Moscow (from where I live in The Netherlands), however, it would be highly unsuitable as a stop-over on the way to Moscow :-)

  • kzb November 15, 2010, 9:15

    Ronald, I’ll try and find the publication I had in mind later today.

    There are quite a few papers on the IMF of young stellar clusters. BD’s in these are still quite hot for their size, so there should not be the observational bias that is inevitable in old stellar populations. These studies seem to agree the mass function turns down below about M5.

    If there is any absolute lower cutoff BTW, I don’t think there is concrete evidence of that from these studies. The mass function extends right down into the planetary mass range. Any apparent cutoff at these low masses could have an observational bias origin.

    However there is this very interesting article from the WISE people, and this has BD’s being potentially very common:


  • jkittle November 15, 2010, 16:03

    The idea of observing old brown dwarf objects near the solar system depend a lot on modeling. One prospect not mentioned is that these warm objects also have a pretty steep gravity well and perhaps slowly accumulate mass from captured bodies or molecular clouds. If one were in the Oort cloud for example, it would at some frequency, capture volatile( read hydrogen) rich objects in distant orbit to our sun. It is probably more likely to find a brown dwarf in proximity to a star, ( because of the gravity field and its likely origin) than in deep space.
    Taken together, a brown dwarf in the Oort could may not cool down but stay warm because of the slow accumulation / resupply of deuterium from the ” comet precursors ” it accumulates. You see comet and asteriod impacts on the sun and on Jupiter all the time, so this is plausible. Thus an Old brown dwarf may still be in the 300 to 500K plus range. Remember we still do not really know why Jupiter emits so much more heat than it receives from the sun, so all of this speculation depends on the way you model brown dwarf behavior. In response to an earlier post, we have to build the BD model to suit the observations, not the other way around.
    Here is the wishful thinking part: a brown dwarf at 5,000 to 10,000 AU. t this MIGHT cause space exploration to reach beyond our present technology but still have hope of success. Given the nature of this blog I think it is OK to post such sci fi – like speculation. It is a teleological argument. ” I want it to be this way! and I am unapologetic about my bias, but recognize it is a bias not based on facts (yet) .

  • Eniac November 15, 2010, 17:48

    As has been pointed out, the “stepping stone” concept is not really a good one, here. With current prospects, we’d be lucky to visit one target and fly by in a few minutes. If we ever have the luxury of a second burn, we would decelerate and stay around. “Stopping over” would cost the equivalent of a return mission, without the returning part.

    Recall that the constraints of the rocket equation are such that each “burn” adds a factor, not a term. If, miraculously, we could accelerate to cruising velocity with a fuel/payload ratio of only 10, a stop at the target would require 100, a return 10,000. A two-stop one-way would also be 10,000. If you could “gas up” at the stop, thing would be different, but that is not as easy to do as it may sound.

    So, what nearby BDs it would give us is more targets, and perhaps some that are easier to reach than the easiest so far, Alpha Centauri. Also, perhaps a more interesting one. But not a “stepping-stone”.

  • jkittle November 15, 2010, 19:18

    I agree a stepping stone does not help if you are considering fuel logistics. However in terms of technology it might be really interesting to send a probe to the hypothetical inner Oort cloud or to some Kuiper belt objects, if they exisit, in a region beyond the “Kuiper Cliff”. sucess then breeds sucess to look and futher and longer missions.

    Now if we were to colonize the small icy worlds in some distant future, then our decendents might slowly expand across the universe. This makes little sense with our earth- centric point of view, where most people would consider that only a planet near a warm sun is interesting to colonize. However if we populate Mars, then their grandchildren may not be worried about the cold and the dark and instead look at the rich resources of the small worlds, which might become quasi independent city states like the Greeks of old. In this case pool of objects around a brown dwarf might be very inviting!

  • Ronald November 16, 2010, 6:04

    jkittle: “However if we populate Mars, then their grandchildren may not be worried about the cold and the dark and instead look at the rich resources of the small worlds”.

    Strongly doubt that: if we ever really colonize Mars, one of the first things we will do is try to warm it up as part of terraforming. And introduced life on Mars would also largely depend on the sun for light and warmth.
    This is no analogue for the little icy worlds of the Kuiper Belt and Oort Cloud.
    As I have argued several times before, small icy worlds are extremely risky and even outright unsuitable long-term habitats: just too small and too few resources. Even worse than asteroids.

    As for the stepping stone idea, I think enough has been said about that by now and we can put that to rest, except for one probably valid concept: Alpha Centauri itself may become a stepping stone on our way to even more distant targets and the rest of the MW galaxy.

  • kzb November 16, 2010, 8:37

    The stepping stone idea might be of use if it turns out the only practical way of manned interstellar travel is “slow”, ie multi-generational ships or travellers in suspended animation. Also requres the space density of BDs (or loan planets come to that) to be higher than currently thought.

  • kzb November 16, 2010, 8:47

    Ronald, sorry about the delay, but here as promised is the paper of interest:

    “The ultracool-field dwarf luminosity-function and space density from the Canada-France Brown Dwarf Survey”


    I’ve got a mathematical question regarding this paper (and others on the stellar mass function) , which I am sure some of you clever people will be able to clarify straight away:

    In the paper linked above, the BD mass function is said to be consistent with a “flat” function. However, the function in question is

    dN/dM is proportional to M^(-a)

    They say a “flat” function is when a= zero, and that is consistent with their observations of BD numbers.

    But any number raised to the power zero is 1. So how is the function “flat” ?

  • Ronald November 16, 2010, 10:19

    kzb: thanks!

    I did some research myself, on M dwarf abundance in relation to spectral subclass, using RECONS, NASA NStars, Hipparcos/Yale catalog data, and base data from Nordstrom, Holmberg 2009.

    Indeed, M dwarf abundance drops off sharply at M5 (RECONS) or even starting at M4 (NStars, Hipparcos/Yale).

    Remarkably, according to RECONS, M dwarf abundance does not drop off relative to mass, but keeps increasing even toward the lowest mass classes (0.06 – 0.20 * solar mass).

    In contrast, according to Nordstrom & Holmberg, M dwarf abundance drops off below about 0.35 * solar mass and very sharply below about 0.2 * solar mass.

  • kzb November 16, 2010, 14:30

    RECONS -I thought this site was dead, but I now notice it has been updated slightly in August this year. No new objects have been added since 2009 though.

    RECONS within 10pc gives 332 stars (M and above) in 249 systems, giving average of 0.08 stars/pc^3 and 0.056 systems/pc^3.

    I have some notes, I think I got it from NStars, about the within-5pc population, and that gives 0.12 stars/pc^3 and 0.095 systems/pc^3 (M and above). This 5pc sample is more nearly complete than the 10pc sample, hence the higher apparent density. The average system separation distance is then 2.2pc or 7.1 light-year.

    If you sum the BD numbers in the final table and associated text of that paper I linked above, you will get BD density about an order of mag below the 5pc star density.

    However I could be over-interpreting the data. For one thing, it is not actually very clear what volume of space is being sampled in that paper. It certainly is not the whole sky.

  • Ron S November 17, 2010, 15:48


    I had a very quick look at that paper to see what it said about α. I found this immediately preceding the equation on page 10 you asked about:

    “…where α is the slope of the power-law mass function…”

    In that context if the slope of a curve is zero, it is flat, and therefore M^0 is a “flat” function. The answer to your question may be as simple as that.

  • kzb November 18, 2010, 9:01

    Thanks Ron S, but mathematically M^-0 is 1. Therefore the gradient is 1 (in this context it would be -1 I think). The gradient is not zero, which would be a”flat” relationship. Being -1, it means directly inversely proportional: half the mass doubles the frequency. It was precisely the apparent mis-match between words and equation that made me ask. If you look at their last Table, you will see the number frequency increasing as the spectral type is decreased (perhaps the answer lies in the fact that temperature is not linear with mass?)

    Here’s an interesting paper on planetary mass objects, said to be 1-14% of the IMF in the stellar cluster in question:


  • kzb November 18, 2010, 9:26

    Sorry wrong paper. This is the one with the population fgures of PMOs, < 20M(J) AT LEAST 5% of the stellar fraction:


  • Ron S November 18, 2010, 15:38


    Let me try again since my previous post didn’t include a few key steps, and that omission on my part added to the confusion.

    When the authors use the word “flat” they are explicitly referring to the power-law mass function, which is defined as PSI(M)=dN/dM. They further note that PSI(M) is proportional to M^α. Now we can go through this step by step:

    – When α=0, M^α=1 (as you already noted).
    – Since PSI(M) is proportional to M^α, when α=0 we have PSI(M) proportional to 1.
    – This means that when α=0, PSI(M)=dN/dM= k, where k is some constant, since proportional means that the ration of PSI(M) and M^α is a constant. Therefore, PSI(M)/1=PSI(M)=k.

    PSI(M)=k is most definitely a flat function. They are *not* saying that M^α is a flat function. I did not read further to see what they are claiming (if anything) about k.

  • jkittle November 18, 2010, 16:03

    I understand the concern with resource limitations on smaller world, but consider two factors:
    1) delta v to travel ( if a bit slowly ) between world in similar orbits is very small thus resources can be shared or pooled ( once you get there) thus a population might grow in place.
    2) we actually know very little about these worlds except most are covered in volatiles. The (much) higher-than-expected density of Eris would indicate the much of the composition modeling to date is based on relatively sparse data.
    If you look at the spectroscopic signatures of Near earth objects from the recent Spizter study, there is wide variation , object to object. These objects were previously thought to comprise a class but it looks like their origins and compositions are all over the map . Similarly, I am beginning to wonder if the outer solar systems objects are diverse enough to contain objects with high metal content, or high carbon content , as well as the iceballs (even these have carbon, nitrogen hydrogen and oxygen in various compounds. )
    technology follows the resources available, not the other way around. Your mars colonists and their putative interstellar voyaging offspring may not feel the need to make the place look like some soon-forgotten earth home. Walking barefoot on a front lawn of green grass under a naked blue sky is just a habit, not a necessity for life . Look how many people have adapted to a totally urban environment. If you live in a mall, who cares where the air comes in from as long as it is breathable.
    Mostly , the great void between objects that must increase as you get futhter out is a real interesting concern, and the population density of objects in each class of sizes is still debateable. not much out there for sure, but maybe concentrations in the gravity pools around BD object ( the subject of this thread!)
    Question: if you allow for fusion power, and sophisticated synthetic chemistry and synthetic biology, how big a population might be sustainable, for say 1,000 years on a 10oKm Kuiper belt object. Care to speculate?

    My speculation is >50,000 people , given the surface size ( 30,000 square km) and ability to delve deep into surface due to the low gravity . There is a HUGE volume of water and other volatiles!
    Still Dreaming!

  • kzb November 22, 2010, 8:44

    OK thanks for the maths lesson Ron S :)

    The sum of BD space densities in that paper is about one-seventh the space density of stars (class M and above) in the within-5pc sample, admittedly with a wide uncertainty band.

    However, this density would seem to fit with other papers, which give only a minority (typically 15-35%) of the mass function in the BD area. The most common objects are still M-class stars above the hydrogen burning limit, not brown dwarfs.

    So I now wonder about the WISE mission statement, that it “would probably find the nearest star to the sun”. That would seem to be stretching the truth going by these papers. Should be “possibly” not “probably”.

  • Ronald November 22, 2010, 10:41

    @kzb: maybe an open door, but this also clearly implies something else, namely that BDs cannot account for any significant part of dark matter in the universe, as some have suggested.

  • kzb November 23, 2010, 14:48

    Ronald: about the DM question, the microlensing searches pretty well closed the door on BD’s being the missing mass some time ago. However, I used to think there was the possibility of an “Oops !” moment, (and perhaps I still do-just).

    However, going back to the lead article, the fact that WISE have released this detection as “news” does not fill me with confidence. This BD is 18-30 light years away. It is beginning to look like the sky did not exactly light up with millions of BD’s when they turned on the WISE camera. Otherwise they’d have told us ?