News & Views | Published:


Migration of giants

Nature volume 537, pages 496497 (22 September 2016) | Download Citation

The origin of hot Jupiters, large gaseous planets in close orbits around stars, is unknown. Observations suggest that such planets are abundant in stellar clusters, and can result from encounters with other celestial bodies.

Hot Jupiters are a rare and peculiar class of exoplanet that have masses comparable to that of Jupiter and short orbital periods of typically four to five days1. The first hot Jupiter was discovered2 in 1995, and, since then, astronomers have debated the physical processes that are responsible for their production. Because hot Jupiters are the most extreme outcome of planet formation, discovering their origin will considerably refine our understanding of how planets are assembled and how their orbits evolve. Writing in Astronomy & Astrophysics, Brucalassi et al.3 find that hot Jupiters are five to ten times more abundant in stellar clusters than elsewhere in the Milky Way, and the authors provide evidence for a scenario that explains the production of such planets.

Hot Jupiters are extreme in every way. Imagine objects that have radii ranging between half and twice that of Jupiter, and atmospheric temperatures of between about 1,000 and 3,000 kelvin. These temperatures imply that many hot Jupiters emit more infrared radiation than some stars4.

The orbital proximity of hot Jupiters to their host stars also means that they experience intense gravitational (tidal) forces, which elongate the planets into a rugby-ball shape5. These forces synchronize the planets' rotation and orbital speeds in a similar way to the forces exerted on the Moon, such that one side of the planet (the dayside) is always facing the star. Powerful winds can reach supersonic velocities within the atmospheres of hot Jupiters6, transporting heat from the dayside to the nightside. Finally, several of these planets have been observed to lose mass under the intense radiation that they receive from their host stars7.

Three properties distinguish a hot-Jupiter population from all other exoplanets. First, there is an excess of objects that have orbital periods shorter than about 10 days and masses similar to that of Jupiter1. Second, the objects rarely have companion planets on nearby orbits8. Finally, nearly one-third of hot Jupiters have orbital paths that are inclined with respect to their star's equator9, and several planets in the population rotate in the opposite direction to the star.

Several hypotheses have been put forward to explain the existence of hot Jupiters. For example, they could have formed close to their host stars, although this idea is disputed by most of the scientific literature. A more likely explanation is that they formed far from their stars and then migrated inward.

Two main scenarios could explain this migration. First, a hot Jupiter might lose angular momentum to the material of the protoplanetary disk from which it formed10. Second, the planet might exchange angular momentum with another celestial body (a planet, a stellar companion or a passing star), throwing the planet onto a highly eccentric orbit11,12,13 (a process called dynamical migration; Fig. 1). Then, when the planet reaches periastron — the closest point on its orbit around the host star — tidal forces deform it, and friction dissipates its angular momentum as heat. The planet's orbital distance therefore shrinks, and the shape of its orbit becomes increasingly circular. In my opinion, a mixture of these two scenarios is likely to have produced most hot Jupiters; this mixture has led to our uncertainty over the origin of these planets.

Figure 1: Possible scenario for the dynamical migration of a hot Jupiter.
Figure 1

Brucalassi et al.3 show that a class of exoplanet called hot Jupiters can form because of interactions with other celestial bodies. a, Two or more planets form within a protoplanetary disk — the dust and gas surrounding a newborn star — far from the star. b, The disk material accretes onto the star, but also onto the planets, whose masses become similar to that of Jupiter. c, At some point, a dynamical instability occurs, either as an inherent, unstable planetary configuration (shown) or as a result of a passing celestial body such as a star (not shown). d, One of the planets is launched on a highly eccentric orbit, whereas the others either adopt wide orbits or are ejected. e, When the eccentrically orbiting planet passes close to the star, tidal forces cause its orbit to shrink and become more circular. The result is a large planet that has a short orbital period: a hot Jupiter.

Brucalassi and colleagues observe a sample of stars in the 'open' cluster M67, which consists of about 1,000 stars within a radius of 3 parsecs (10 light years). With respect to most open clusters, M67 is quite old (about 4 billion years, slightly younger than our Sun). Whereas only about 0.5–1% of Sun-like stars host a hot Jupiter1,14,15,16, the authors report that an astonishing 5.6% of stars in their sample are accompanied by such extreme planets. This result indicates that hot Jupiters are produced in greater quantities within a cluster environment (where stars frequently encounter each other) than within the general stellar field of our Milky Way. In addition, all of the hot Jupiters that the authors studied have eccentric orbits.

Both the abundance of hot Jupiters in M67 and the planets' eccentric orbits are consistent with the scenario of dynamical migration initiated by a passing star. In this case, when a star passes a planetary system, the star's gravitational pull will slightly alter the orbital paths of the planets around their host star. From an a priori stable configuration, even a single stellar encounter can sow the seed of discord, leading to a major instability within the system a few billion years later17. Occasionally, this instability can throw one of the planets onto an eccentric and inclined orbital path. The authors observe hot Jupiters in the final stage of dynamical migration, before the planets' orbits have become completely circular.

Brucalassi and collaborators' results are consistent with earlier predictions17,18 that were made by studying the impact of stellar fly-bys on planets; these are relatively frequent events for planetary systems within open clusters such as M67. This is a rare example in exoplanetary research of observations verifying predictions — usually, observational evidence drives theoretical developments. The authors' results also show that, by investigating hot Jupiters in an open cluster, we can obtain useful information about the dynamical migration process. This information will allow astronomers to verify whether a similar scenario is responsible for hot-Jupiter formation around stars elsewhere in the Milky Way, and will enhance our understanding of tidal forces.

As we learn more about planets in stellar clusters — for instance, by using the Kepler–K2 satellite — it will be possible to determine the inclination of planetary orbits, thanks to a spectroscopic phenomenon called the Rossiter–McLaughlin effect19. We will also be able to study planetary atmospheres and measure their chemical make-up. Planets that form far from stars are expected to contain more water than those that form close by20. Therefore, water-rich gas giants on close-in, eccentric and inclined orbits within an open cluster would leave little room for alternative formation scenarios besides dynamical migration. After more than two decades, we might finally solve the mystery of hot Jupiters.



  1. 1.

    et al. Astron. Astrophys. 587, A64 (2016).

  2. 2.

    & Nature 378, 355–359 (1995).

  3. 3.

    et al. Astron. Astrophys. 592, L1 (2016).

  4. 4.

    Mon. Not. R. Astron. Soc. 439, L61–L64 (2014).

  5. 5.

    , & Astron. Astrophys. 528, A41 (2011).

  6. 6.

    Astrophys. J. 761, L1 (2012).

  7. 7.

    et al. Nature 422, 143–146 (2003).

  8. 8.

    , & Astrophys. J. 825, 98 (2016).

  9. 9.

    & Annu. Rev. Astron. Astrophys. 53, 409–477 (2015).

  10. 10.

    et al. in Protostars and Planets VI (eds Beuther, H. et al.) 667–689 (Univ. Arizona Press, 2014).

  11. 11.

    , & Astrophys. J. 678, 498–508 (2008).

  12. 12.

    & Astrophys. J. 735, 109 (2011).

  13. 13.

    Astrophys. J. 799, 27 (2015).

  14. 14.

    et al. Publ. Astron. Soc. Pacif. 120, 531–554 (2008).

  15. 15.

    et al. Preprint at (2011).

  16. 16.

    et al. Science 330, 653–655 (2010).

  17. 17.

    , & Mon. Not. R. Astron. Soc. 411, 859–877 (2011).

  18. 18.

    , & Astrophys. J. 816, 59 (2016).

  19. 19.

    et al. Astron. Astrophys. 359, L13–L17 (2000).

  20. 20.

    , , & Astrophys. J. 743, 191 (2011).

Download references

Author information


  1. Amaury Triaud is at the Institute of Astronomy, University of Cambridge, Cambridge CB3 0HA, UK.

    • Amaury Triaud


  1. Search for Amaury Triaud in:

Corresponding author

Correspondence to Amaury Triaud.

About this article

Publication history




By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Newsletter Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing