NASA collaborates with the German Aerospace Center (DLR) on one of our more interesting observatories. SOFIA, the Stratospheric Observatory for Infrared Astronomy, is a Boeing 747 aircraft that flies an infrared telescope with a 2.7 m diameter mirror. Located on the port side of the fuselage near the tail, the telescope houses a number of instruments for infrared astronomy at wavelengths from 1-655 micrometers (μm). One of these is FORCAST (Faint Object Infrared Camera for the SOFIA Telescope), which has now spotted an intriguing phenomenon, one that may be flagging a collision of two exoplanets.
The stars in question form a double system called BD +20 307, some 300 light years from Earth. Note the age of this system, about one billion years, an important consideration in what follows. About ten years ago, observations from the Spitzer instrument as well as ground observatories produced evidence of warm debris here, whereas from age alone, we would have expected warm circumstellar dust to have disappeared, just as it has in our own system.
What SOFIA brings to the table is a new set of measurements that shows the infrared brightness from the debris at BD +20 307 has increased by more than 10 percent in a time period of 10 years. We don’t usually find this kind of rapid fluctuation when studying what ought to be the gradual evolution of a planetary system, especially not in one as mature as this. There should be little dust here to begin with, much less warm dust, and while there are other possible mechanisms in play (see below), the rapid pace implies a collision.
“This is a rare opportunity to study catastrophic collisions occurring late in a planetary system’s history,” said Alycia Weinberger, staff scientist at the Carnegie Institution for Science’s Department of Terrestrial Magnetism in Washington, and lead investigator on the project. “The SOFIA observations show changes in the dusty disk on a timescale of only a few years.”
Image: Artist’s concept illustrating a catastrophic collision between two rocky exoplanets in the planetary system BD +20 307, turning both into dusty debris. Ten years ago, scientists speculated that the warm dust in this system was a result of a planet-to-planet collision. Now, SOFIA found even more warm dust, further supporting that two rocky exoplanets collided. This helps build a more complete picture of our own Solar System’s history. Such a collision could be similar to the type of catastrophic event that ultimately created our Moon. Credit: NASA/SOFIA/Lynette Cook.
From the paper:
We investigated several mechanisms that could cause the observed changes in the disk flux, including making the dust grains hotter, either through an increase in stellar luminosity or moving the dust grains closer to the stars, or increasing the number of dust grains in the system. If the origin of the copious amount of warm dust orbiting BD +20 307 is an extreme collision between planetary-sized bodies, then this system may help unlock clues into planetary systems around binary stars, along with providing a glimpse into catastrophic collisions occurring late in a planetary system’s history.
It’s also true that gaining a stronger understanding of dusty debris disks should give us insights into how binary systems evolve, useful as we investigate such interesting places as the Alpha Centauri triple system. If we are indeed looking at the result of a major collision at BD +20 307, further work should illuminate the kind of catastrophes we find evidence for in the Solar System, from our Moon’s formation (likely through an impact with a Mars-sized object) to the huge axial tilt of Uranus, which is probably the result of multiple impacts. The authors argue that new SOFIA observations at a wider wavelength range out to 20 µm will allow us to draw more definitive conclusions.
The paper is Thompson et al., “Studying the Evolution of Warm Dust Encircling BD +20 307 Using SOFIA,” Astrophysical Journal Vol. 875, No. 1 (12 April 2019). Abstract / preprint.
One thing this finding points out is the need to drop the ‘cleared it’s orbit’ part of the IAU’s definition of “planet”, since then, by definition, it would be impossible for two planets to collide.
That’s nonsense. The definition of planet doesn’t have to hold for billions of years, it’s a temporal property of bodies. If ‘cleared the orbit’ should be dropped, ‘orbiting a star’ or ‘being roud-shaped by it’s own gravity’ should be dropped too, because we can’t know whether a planet will be expelled from its system eventually or even destroyed by its host star becoming a red giant. There would be no planets in the Universe!
In common usage, the word planet does include bodies that haven’t yet cleared their orbits, and, as the suspected case under discussion indicates, such objects can even collide. Planets can be ejected too. I’d just like to see the technical definition of ‘planet’ changed to include such objects also. After all, even professional astronomers use the term for objects that don’t fit into the IAU’s definition all the time.
The IAU definition does not distinguish between stable and chaotic/unstable systems. But then, is it possible to have a general definition of planet which includes orbital dynamics, in chaotic systems, where all types of chaos can occur and future configuration is intrinsically uncertain?
If Solar system is sufficiently perturbed, it can enter a state where all eight planets still continue to orbit the Sun for some time, but only Jupiter can be sure that it will remain under IAU definition when things settle down…
What about planets not orbiting stars?
Good points, but in my admittedly irrelevant opinion, even an ejected planet would still be a planet. The origin of the word was ‘wanderer’ after all.
Not impossible. Consider two co-planar planets, both with moderate eccentricity(i.e. 25-30%), one whose periastron is ~1 AU, the other whose apastron is ~1 AU. Origianally thought to be ~1 GYr, a revised age puts this binary system at roughly the age of our solar system(MORE ON THIS LATER!). Initially this system’s planets would have resonant orbits, most likely 5 to 2, so that one planet’s apastron passage would NEVER be NEAR the same time as the other’s periastron passage, rendering a collision IMPOSSIBLE! Howrver, should a relatively recent(astronomically speaking)close fly-by of a star(or very close fly-by of a brown dwarf)DISRUPT this resonance EVEN SLIGHTLY, disaster is inevitable. NOW: Getting back to the age of this system! Since there are two stars of nearly solar mass and nearly equal luminosity, the flux would be ~twice the luminosity of our solar system at 1 AU. If the outer planet were between 30-35% Earth mass, it would be able to hold on to only a small fraction of the bulk of Earth’s atmosphere, preventing a runaway greenhouse turning it into an exovenus, but a large enough fraction to prevent it from becoming Mars-like. Should this have occurred, the planet could have been marginally habitable, although Andrew Lepage may argue strongly against this. Despite the most likely extreme unlikelihood of this scenario being true, we should conduct an extensive search for technosignatures in this system!
Just to be clear, I wasn’t suggesting that these planets couldn’t have collided, I’m just making a gripe against the IAU’s too limiting definition of planet.
Four years ago I happened across what seemed like a decent stab at a new definition of a planet: https://arxiv.org/pdf/1507.06300.pdf (Jean-Luc Margot) But this isn’t my field, and I haven’t paid attention to how this was received or revised.
Really, are we still rules-lawyering the IAU definition in 2019?
*sigh*
There is still some rock in icy Kuiper belt objects so their collision might make a lot of dust and ice fragments which might be responsible for the observed variation in brightness.
I am awed to know that we can observe (potentially) such events. But a bit scared too.
It would be very interesting to observe this system with ALMA.
One thing to note here is that the orbital period of the BD+20°307 binary is 3.4 days. The Kepler mission results indicated a lack of transiting circumbinary planets around short-period binaries, with the current record holder being the 7.4-day period of Kepler-47. So whatever’s going on at BD+20°307, it may well be an entirely different class of system to the previously-known circumbinary planets.
I see no estimate of the total mass of the two colliding objects. Assuming we get better at imaging these events allowing us to make such estimates then…
If we can find collision debris data of other star systems could
we not make a guess at the average size of rocky planets in solar
systems. Kepler data cannot tells that as it skews data points
towards very large terrestrial at the HZ and beyond or very close in smaller (down to mars size I think I read)rocky worlds (very hot worlds).
Thanks for sharing that link Mike. They did come up with a well reasoned alternative definition for planet, although siding against calling ejected objects planets. Here’s what the group recommended:
“A planet is a celestial body that (a) is in orbit around one or more stars or stellar remnants, (b) has sufficient mass to clear [or dynamically dominate] the neighbourhood aroundits orbit, i.e., ? ? 1, (c) has a mass below 13 Jupiter masses, a nominal value close to the limiting mass for thermonuclear fusion of deuterium.”
The formula for the term ? didn’t copy well, see page 6 of their paper if interested.
Their proposal eliminates the roundness requirement because any object large enough to to clear its orbit will naturally be massive enough to also be round. This doesn’t address systems that are unstable.
“This doesn’t address systems that are unstable.”
What do you mean by this? Inherent in the differential equations of Newtonian gravitation (and GR of course) no multi-planet system is provably stable, including our own solar system. That is, they are chaotic. Run the most high precision models of our solar system and going beyond a several hundred million years it is entirely possible that Earth or other small planets can be ejected. Past and current stability is no guarantee of future stability. This doesn’t even encompass perturbations by an passing interstellar body in the unknowable future. Yet I’m pretty sure that any sensible definition would not decline to classify Earth as a planet.
In any case I am happy to ignore both sensible and insensible definitions. From this you can discern that I’m not a lawyer.