Thousands of extrasolar planets have been discovered, but none is a planet in its infancy. Observations have finally been made of a young planet growing in its birthplace — opening the way to many more such discoveries. See Letter p.342
Finding young planets in their birthplaces is extremely challenging, because actively forming planetary systems are far away and obscured by dust. On page 342 of this issue, Sallum et al.1 report the use of a new technique to detect an emission signal from a growing planet. This discovery has far-reaching implications for our understanding of the planet-formation process and of the properties of young planets.
When a star is born, a flat rotating disk of gas and dust forms around it, known as the circumstellar disk. This disk continuously transports dust and gas inward to feed the young star for millions of years, a process known as disk accretion. Planets are thought to form from the leftover material from this disk. Earth, for example, was born in the circumstellar disk surrounding the young Sun 4.6 billion years ago. But little is known about how microscopic dust particles can grow 14 orders of magnitude bigger to become a giant planet within the relatively short lifetime of the disk. Finding young planets in circumstellar disks should provide important clues about when, where and how young planets are born.
But finding planets is difficult because they are small and dim. The presence of most known planets has therefore been inferred from observations of the stars around which they revolve. For instance, the Kepler satellite2 has discovered more than 1,000 planets by measuring the tiny dimming of stellar light that occurs when a planet passes in front of its star.
Such methods cannot be used to find young planets around young stars, because such stars are highly active and the light they emit is variable. Most attempts to find young planets use 10-metre-diameter optical telescopes to directly image planets in circumstellar disks. Disks with large cavities are particularly targeted3, because such cavities are thought to be opened up by giant planets in orbit around the central star.
In 2012, a protoplanet candidate 1,000 times fainter than its host star was discovered4 in a system called LkCa 15 (Fig. 1). The central star of this system is similar to our Sun, but is only 2 million years old. The Sun-like star has a circumstellar disk with a cavity 50 astronomical units in radius (1 AU is the distance from Earth to the Sun).
The protoplanet candidate, called LkCa 15 b, resides inside the cavity, 16 AU away from the central star. But the nature of LkCa 15 b is unclear — it is redder than would be expected for a young planet. Several young planet candidates5,6 have since been discovered in other circumstellar disks, and they are all quite red. This has led to the hypothesis that the detected red objects are young planets with circumplanetary disks. When a planet is born, a rotating circumplanetary disk of gas and dust forms around it, similar to the circumstellar disks around young stars. As the accreting disk feeds the nascent planet, it releases energy and becomes bright. The emission from such a disk should be redder than the planet itself7.
In their study, Sallum et al. searched for a signature of young planets: Hα photons, which are emitted from hydrogen atoms only when a circumplanetary disk accretes onto a planet. If a young planet has strong magnetic fields, the fields form a large magnetosphere around the planet, which can truncate the circumplanetary disk7,8. The material in the disk therefore has to follow the planet's magnetic fields to accrete onto the planet. During this accretion process, the magnetosphere can be as hot as 10,000 kelvin, which is what causes hydrogen atoms to emit Hα photons9.
Although the emission of Hα photons has been widely observed when circumstellar disks accrete onto young stars, Sallum and colleagues are the first to directly image accreting circumplanetary disks around young planets using Hα photons. To do this, they used a filter that allows only Hα photons to reach their telescope10. The authors report that LkCa 15 b is an Hα emitter, providing strong evidence that it is a young planet with a circumplanetary disk still accreting onto the planet. Furthermore, they found two other objects inside the cavity of the LkCa 15 system, although these do not seem to be Hα emitters. By combining information from observations made over several time periods with data from the initial discovery in 2012, Sallum et al. determined the orbits of two of the young planet candidates.
The researchers' discovery provides stringent constraints on planet-formation theories. For example, such theories now have to explain how a giant planet can form 15–16 AU from its star within 2 million years, and still be growing after this time. Another implication of the findings is that a young planet's magnetic fields need to be at least 20 times stronger than Jupiter's current magnetic fields to truncate the accreting circumplanetary disk. This in turn implies that the internal motion of young giant planets is much greater than that of the giant planets in the Solar System, and provides an indirect probe of the internal structure of such planets.
Both the red colour and the Hα emission from LkCa 15 b can be explained by the presence of an accreting circumplanetary disk, but some caveats should be kept in mind. Our knowledge of Hα emission from accretion disks builds on data from disks around young stars that are hundreds of times more massive than planets. Sallum and co-workers therefore extrapolate the known relationship between Hα flux and disk-accretion rate to a completely new size scale. Measurements of other accretion tracers would be desirable, such as 'continuum' emissions at ultraviolet and optical wavelengths11. The nature of the two sources that do not emit Hα photons also remains unclear. Follow-up observations, especially at mid-infrared and submillimetre wavelengths, are needed to clear up these issues.
Nevertheless, the authors have demonstrated a powerful technique to find young planets in circumstellar disks, one that will discover many such planets in the future. This would potentially allow the distribution and occurrence of young planets to be determined with a comparable accuracy to that for the mature exoplanets discovered by the Kepler satellite. Such an understanding of the young planet population will shed light on the decades-old problem of planet formation, and reveal how young planetary systems can evolve into older ones such as our Solar System, billions of years after they were born.Footnote 1
Sallum, S. et al. Nature 527, 342–344 (2015).
Borucki, W. et al. Science 327, 977–980 (2010).
Espaillat, C. et al. Protostars and Planets VI 497 (Univ. Arizona Press, 2014).
Kraus, A. L. & Ireland, M. J. Astrophys. J. 745, 5 (2012).
Quanz, S. P. et al. Astrophys. J. 766, L1 (2013).
Reggiani, M. et al. Astrophys. J. 792, L23 (2014).
Zhu, Z. Astrophys. J. 799, 16 (2015).
Lovelace, R. V. E., Covey, K. R. & Lloyd, J. P. Astron. J. 141, 51 (2011).
Hartmann, L., Hewett, R. & Calvet, N. Astrophys. J. 426, 669 (1994).
Close, L. M. et al. Proc. SPIE 9148, 91481M (2014).
Calvet, N. & Gullbring, E. Astrophys. J. 509, 802 (1998).