≡ Menu

Measuring a Brown Dwarf’s Winds

The brown dwarf 2MASS J10475385+2124234 is about the size of Jupiter, but maybe 40 times more massive. 33.2 light years from Earth, this object is in that category between planet and star, not massive enough to launch the same kind of nuclear reactions that power the Sun, but considerably more massive than any planet. Combining two tools — the Very Large Array (VLA) and NASA’s Spitzer Space Telescope — scientists have now measured the wind speed here.

Katelyn Allers (Bucknell University), who led the research team, realized that the combination of radio observations (VLA) and infrared (Spitzer) would make this kind of measurement possible, and expressed surprise that no one else had thought to do it before. After all, we already knew that the rotation period of Jupiter found through radio measurements differs from the period found at visible and infrared wavelengths. That disparity is key to the new measurement.

For the difference is the result of two separate phenomena. Radio emissions from Jupiter are caused by electrons interacting with the planet’s huge magnetic field, located deep in the interior. The infrared emission is measured at the top of the atmosphere. Because the atmosphere is rotating more quickly than the interior, we have a difference in velocity due to atmospheric winds. The same kind of measurement is now put to use at a brown dwarf.

“Since the magnetic field originates deep in the planet, or in this case brown dwarf, the radio data allows us to determine the interior period of rotation,” Allers says. “When you have an interior rotation rate and an atmospheric rotation rate, you can compare them to see how fast the wind is blowing.”

Image: Artist’s conception of a brown dwarf and its magnetic field. The magnetic field, rooted deep in its interior, rotates at a different rate than the top of the atmosphere. The difference allowed astronomers to determine the object’s wind speed. Credit: Bill Saxton, NRAO/AUI/NSF.

Allers and company observed 2MASS J10475385+2124234 with Spitzer in 2017 and 2018, noting the interesting fact that its infrared brightness varies regularly, indicating a long-lived feature of some sort in the upper atmosphere. Does this brown dwarf have something analogous to Jupiter’s Great Red Spot? That we can’t know at this point, but a set of VLA observations from 2018 measured the rotation period of the brown dwarf’s interior, so we do know about the winds.

The Jupiter analogy holds, for the brown dwarf’s atmosphere does indeed rotate faster than its interior, allowing the calculation of a wind speed of about 2300 kilometers per hour. By comparison, Jupiter’s wind speed clocks in at around 370 kilometers per hour. According to Allers, the difference is in agreement with theory and simulations of brown dwarf wind speeds.

This initial measurement may alert us to a technique that can measure winds not only on other brown dwarfs but on some giant exoplanets, although as co-author Peter K.G. Williams notes, the magnetic fields of such worlds are weaker than those of brown dwarfs, so the radio measurements would be conducted at lower frequencies than those used in the 2MASS J10475385+2124234 work. Williams (Center for Astrophysics | Harvard & Smithsonian) led the radio astronomical observations at the VLA.

“For the first time ever, we measured the speed of the winds of a brown dwarf — too big to be a planet, too small to be a star,” Williams adds. “The results rule out a few unusual models and prove that this new technique works and can be applied to more objects… This new technique opens the way to better understanding the behavior of atmospheres that are unlike anything found in our solar system.”

Image: Brown dwarf, left, and Jupiter, right. Artist’s conception of brown dwarf illustrates magnetic field and atmosphere’s top, which were observed at different wavelengths to determine wind speeds. Credit: Bill Saxton, NRAO/AUI/NSF.

The paper is Allers et al., “A Measurement of the Wind Speed on a Brown Dwarf,” Science Vol. 368, Issue 6487 (10 April 2020). Abstract.


Comments on this entry are closed.

  • Ivar Brouwer April 9, 2020, 16:35

    I wonder how analogues Jupiter can be to Brown Dwarfs in these regards: Brown Dwarfs pack a lot more mass in the same volume, so I would assume it’s top atmosphere is a lot more dense and maybe more active

  • Geoffrey Hillend April 9, 2020, 17:21

    Awesome. A wind speed of 2300 kilometers per hour. And of course like Jupiter, the magnetic field in a brown dwarf is created by Coriolis deflection of the convection currents within the pressure ionized, liquid metallic hydrogen in the interior. The hydrogen is so dense it is pressure ionized, so the electrons are more free, so the hydrogen conducts electricity like metal where electrons are more free. The rotation moves the charged particles in a circle in the interior to make a magnetic field.

  • Geoffrey Hillend April 9, 2020, 17:29

    The radio emission frequency is higher with a brown dwarf because energy is inversely proportion to wavelength, and a higher frequency represents shorter wavelength and higher energy due to the fact that the magnetic field of a brown dwarf is stronger than Jupiter’s magnetic field, so it can accelerate electrons faster than Jupiter. The radio waves are created by Synchrotron radiation where charged particles emit electromagnetic radiation when accelerated rapidly in a magnetic field.

  • wdk April 9, 2020, 21:11

    This report ties nicely into some solar system discussions we have had locally about Jupiter and Saturn’s atmospheres and winds – and the notion that Jupiter resembles a star – and yet is not with the intermediate distinction of brown dwarf in between.

    The implications of being a brown dwarf vs. an M2 main sequence star is that the object’s mass is between 13 and 80 Jupiter masses, allowing a deuterium fusion reaction to ignite in the interior ( and possibly others).
    So consequently there is considerably more heat in the core than that delivered by the heat of formation from falling into the gravity well.
    But I only remember a figure of 300,000 years of deuterium burning.
    Probably a figure of merit. The theory probably gives burn times across the mass range with different amounts of energy contained. Still, if there is more energy contained, they you can imagine more rapdily moving belts and zones transporting energy in the convective outer regions.

    But then there are a couple of questions that arise contemplating the picture so far. Our main sequence sun does not have belts and zones, but something more like cells, similar to grains boiling in a pan. Brown dwarfs evidently resemble Jupiters more than G stars. Now what about M dwarfs? And then there are still discussions about whether Jupiter has a significant iron core. What’s the case with a brown dwarf? Would you want to have an iron core sitting in the center just when you wanted to start nuclear fusion? Seems like an inhibitor.

    • Ashley Baldwin April 11, 2020, 7:44

      Brown dwarfs are ‘true’ failed stars. Formed in the same way – from collapse of an interstellar gas cloud . For Brown dwarfs , insufficient mass to reach the temperature and pressure required for prolonged hydrogen fusion by overcoming electron degeneracy pressure as the cloud contracts. ‘Top down‘.

      By way of comparison, Gas giants ( as seems to have been confirmed by Juno’s provisional findings ) form from the ‘bottom up’. Seeded by the rapid pebble accretion of a large rocky core ( circa 10 Earth mass ?) which then rapidly attracts a massive additional helium and hydrogen atmospheric envelope. Out of the primordial circumstellar disk – before it is dispersed by the nascent parent star’s stellar winds.

      There is presumably some overlap in the masses of the two types of body .

      What surprises me here most is the variations in the infrared emissions . This brown dwarf is late T spectrum . According to conventional doctrine it should as such have a largely clear atmosphere free from large clouds or storms.

      I guess the rotation periods of the inner and outer layers were calculated by the Doppler method – changes in the radial velocity of the dwarf as seen from Earth – in the two different wavelengths measured ( radio versus NIR) . With the wind speed then produced by simply subtracting one from the other.

      I wonder if they measured the cadence of the infrared variations ? If they were indeed due to a large exo Red Spot, then this could offer an alternative measure of outer atmospheric circulation/ wind speed.

      Ive not seen the full paper . Are these things discussed Paul?

      • Paul Gilster April 11, 2020, 20:35

        Ashley, I didn’t have time today, but let me look back through the paper and get back to you tomorrow.

      • Paul Gilster April 13, 2020, 11:55

        Ashley, I can’t find a comment on cadence in the paper, but let me send you some other info back channel.

        • Ashley Baldwin April 14, 2020, 7:46


          Question: When is an Brown dwarf not a brown dwarf?

          Answer: When its a high radio emitter and you need to measure its surface wind speeds in which case it’s considered a planet !

          The technique employed in this work essentially closely mirrors what has been employed for wind speeds on Jupiter .

          The periodicity of the radio emissions ( as measured by the VLA) of the Brown Dwarf in question is used as a proxy for the rotation of its magnetosphere . The assumption is that this will be the same as seen with Jupiter – with the magnetosphere located several thousand Kms below the visible surface and behaving like a rigid body . ( be that situated in liquid hydrogen or whatever ) Presumably everything exterior to that is considered as atmosphere.

          The variation in the infrared photometry seen with Spitzer are assumed to be due to some large ( cloud or storm) surface phenomena moving within atmospheric jet streams . Not actually resolved, as with Jupiter, but the photo metric fluctuations are strong enough for a periodicity to be calculated. From this the speed ( and direction) of the movement can be calculated. This is faster than the rotational speed of the magnetosphere . By simple subtraction the surface “wind” speeds can be calculated. In terms of inclination , the brown dwarf is considered to be viewed side on in the plane of its equator, as this has been shown to be the case with most other strong radio emitting Brown dwarfs whose inclination has been measured .

          The premise here is that despite forming in a completely different manner and with considerably greater mass ( circa 40Mjup) , brown dwarfs , even massive ones, are a assumed to follow gas giants in their behaviour and magnetic structure.

          There was another brown dwarf surface rotational rate calculated by spectrophometry derived from HST oobservational published on just Astro-ph today.

          The authors in the study above propose their method as a way of calculating the wind speeds on exo-Jupiters. Though the magnetic fields and thus radio emissions of these would be much smaller. A gas giant like Jupiter has a magnetic field strength ( and resultant radio emissivity ) of one to two orders of magnitude less than brown dwarfs.

          RV spectroscopic measurement of exo Jupiter wind speeds is limited to precision photometry of transiting and tidally locked Hot Jupiters.

  • Alex Tolley April 10, 2020, 10:59

    2300 kph wind speed suggests it might either be supersonic like those on Neptune, or subsonic if the gases in the BD atmosphere are heavier (e.g. methane) or much denser.

    Was this predicted or is this observation showing something unexpected?

    • Paul Gilster April 10, 2020, 11:50

      My understanding is that this work confirms existing models, so no huge surprises here.

  • Geoffrey Hillend April 10, 2020, 20:54

    wdk, actually, the brown dwarf is more like a star than Jupiter which does not have the strong, convective layer. The brown dwarf has the boiling, convective interior because it is much hotter than Jupiter in it’s core due to the greater mass and thermal induced gravitational compression of the core. The surface of this brown dwarf is over 2800K. It has a coronal plasma which is three million degrees. It is as bright as the Sun in x ray light. https://en.wikipedia.org/wiki/Brown_dwarf Our Sun has a core surrounded by a radiation layer surrounded by a convective layer up to it’s surface. Small stars like M dwarfs don’t have a radiation layer. They only have a convective layer. Inglis, 2015, Astophysics is Easy!

    The radio frequencies of a brown dwarf are explained by the strong, tangled magnetic field lines. I would like to know more about how the magnetic fields in a brown dwarf are made. On the Sun the magnetic field is made from it’s differential rotation. The equator takes 25 days to make one revolution, but the pole takes 34 days. The differential rotation causes the magnetic field lines to become twisted like a rope. They wind up and snap and come through the surface in a loop magnetic field. Electrons are accelerated in the field loop and hit the surface or photosphere where they electrons collide with atoms there to slow them down and make x-rays from bremsstrahlung or braking radiation.

    There must be some braking radiation to make the x rays from a brown dwarf’s surface. A brown dwarf also is assumed to have a differential rotation, so maybe the tangled magnetic fields break the same way. If not the x rays would still have to be made by the breaking of fast particles accelerated in the brown dwarfs magnetic field near it’s surface.

    • wdk April 17, 2020, 19:00

      Hello, G.H.
      Sorry for the delay in response. Work overtook me. But I appreciate the research into the questions posed earlier. Most of it sorts things out or clarifies things for me.

      Admittedly, mass and internal thermal processes early on make brown dwarfs closer to suns than Jupiter is. And your distinctions on layers of heat transport also helps. But still, the illustration of bands and zones intrigues me. For Jupiter it is a convective surface process. And somewhere lower down there is a degenerative or metallic hydrogen phase that changes the heat transport. And if memory hasn’t failed me, I believe that some red dwarfs had that feature as well. But the sun’s surface, convective as it is, has cell structure rather than bands. With the coiling magnetic field model, is that directly connected to west to east wind surface winds on jovian planets and brown dwarfs? Or to put it another way, are cell structures seen in the sun’s surface features like
      “trees” and the wind is more like the forest of them, but moving?

      Best regards.

  • anon86 April 10, 2020, 22:26

    I saw some posts on this site about exomoons. Has any work been done on planets orbiting brown dwarfs in a binary system with a true star?

  • Robert Flores April 11, 2020, 11:42

    I think really odd, that this brown dwarf is not fusing H2 at it’s core.
    Many decades ago, it was though that about 15 or so Jupiter masses would be enough to start fusion. Is there a new standard for minimum mass required to start fusion in a protostar core.?

    • anon86 April 11, 2020, 14:25

      I thought the H2 burns up pretty fast.

      • Ashley Baldwin April 11, 2020, 17:51

        H2 is molecular hydrogen . At the sort of temperatures – and pressures, required for even the most rudimentary nuclear fusion ( circa 1 million degrees C) all elements have lost their electrons and consist of ionised nucleii. The nucleii of the Deuterium isotope of hydrogen can be fused in the centre of brown dwarfs that mass above about 13 Mjupiter – circa 0.013 M sun. ( the uncertainty of this lower mass is due to variations arising from the starting metallicity of the brown dwarf ) This “fuel” is used up in about 10-30 million years. For the largest brown dwarfs , with mass above about 60 Mjupiter , lithium too can be fused – a source that can last for as long as 100 million years. Nothing compared to the trillion year life span of smaller hydrogen fusing M dwarfs , but ironically significantly longer than the most massive main sequence stars.

        Again, dependent on metallicity, full hydrogen fusion commences in stars with a mass of above 75-80 Mjupiter at temperatures approaching fifteen million C.

  • Geoffrey Hillend April 12, 2020, 16:03

    The core temperature of Jupiter is only 17,000 K. at a pressure of 70 MB or 70 million bars. P. 197, The New Solar System, Kelly. 1 bar is one Earth atmosphere, the pressure at sea level of 14.7 pounds per square inch. Jupiter and brown dwarfs are mostly hydrogen and helium, so they resemble stars chemically. Brown dwarfs will always have their helium three, the product of deuterium fusion, but their infra red light source is not bright enough to warm any planets.