By and large, the Solar system is dominated by the giant planets (Jupiter, Saturn) that lie just beyond the ice/snow line, which is the region of the system where ice can survive the Sun’s heat. We can call planets found in this region in extrasolar systems “Jupiter-like planets” if they have a mass equal or larger than Jupiter’s. An external observer would likely discover Jupiter-like planets much before detecting smaller inner rocky planets like Earth. Therefore, a basic step to answer the above questions is to establish how common Jupiter-like planets are around Sun-like stars, as it is already known that giant planets are rare among stars of lower mass.
In the last twenty years surveys where planets are detected using the variation of radial velocities (the first and among the most common tools used to discover planets outside the solar system) have found that Jupiter-like planets are relatively rare around Sun-like stars: only about 6% of these stars appear to host Jupiter-like planets. This indicates that our Solar System is relatively unusual. However, Jupiter-like planets are difficult to discover through radial velocities, because of the long periods and low amplitudes of the radial velocity variations. Typically, these surveys target stars with ages of several billion years, whose sites of formation are essentially unknown. They do not help much to understand the circumstances where our Solar System formed.
A new observation might provide important hints. Using direct imaging, Dino Mesa and colleagues (A&A, 672, A93) discovered a third star hosting a Jupiter-like planet in the β Pic Moving Group (BPMG), AF Lep, which adds to β Pic itself and 51 Eri. The Jupiter-like planets discovered so far in the BPMG are actually four because β Pic is known to host two Jupiter-like planets. The BPMG is a sparse group of 20-million-year-old stars, likely born in a single episode of star formation. The BPMG has only about 30 members with a mass higher than 0.8 solar masses, that is Sun-like stars. Due to its young age and nearness to the Sun, the BPMG is the most favorable star ensemble for planet detection by means of direct high contrast imaging. Even so, discovery of Jupiter-like planets with this technique is still difficult, and only Jupiter-like planets with a mass greater than three times that of Jupiter (i.e., roughly a third of all) can be detected, and that can only occur under favorable conditions. Discovery of three Jupiter-like planet hosting stars through high contrast imaging in this small group of stars is therefore quite unexpected.
In a paper recently published in Nature Communications https://doi.org/10.1038/s41467-023-41665-0 we tried to provide a reference framework for these surprising discoveries. First, we prepared a list of all known members of the BPMG and focused on systems for which enough data are available. We then searched the literature for planets so massive to render the orbit of Jupiter-like planets unstable, thus making their existence impossible. We found that there are 17 systems with sufficient data, which can potentially host Jupiter-like planets with stable orbits. We then considered possible biases that may affect the detection of Jupiter-like planets. Given their very small apparent separation from their Sun-like star, Jupiter-like planets can only be discovered during a portion of their orbits, because otherwise they are projected too close to the star and cannot be seen because of the overwhelming stellar light. We constructed a model to correct for this bias, concluding that the number of detection of Jupiter-like planets should be multiplied by a factor of 2.5 to account for this confound.
As an alternative, we may infer the presence of Jupiter-Like planets from variations in the apparent motion of the star on the sky. Combined with upper limits obtained with imaging and radial velocities this datum may indicate the presence of a Jupiter-like planet with a mass like that obtained with high contrast imaging. This result is independent of the planet position along its orbit. By this method we were able to infer that, in addition to the three systems where planets were detected with high contrast imaging, it is likely that massive Jupiter-like planets are present in four additional systems.
These two estimates of the number of Jupiter-like planets based on high contrast imaging converge. Both indicate that 7 out of 17 stars in the BPMG could have hosted Jupiter-like planets. If we then consider that only a third of the Jupiter-like planets are so massive, our conclusion is that nearly all the Sun-like stars in the BPMG that can host a Jupiter-like planet do so. This tells a very different story about how planets form than the result obtained from radial velocity surveys: the frequency of Jupiter-like planets is low overall, but it is large in the BPMG.
There are several possible explanations for the lack of convergence between radial velocity and imaging methods. The presence of giant planets is known to depend on the metal content and mass of their Sun-like stars, but this is insufficient to explain the large discrepancy observed across methods. Stars may lose planets due to interactions with other passing objects, whose perturbations may cause long-term instabilities in the systems, especially when several planets are present. A final possibility is that Jupiter-like planets form preferably in low density environments, such as the BPMG, while most old stars surveyed through radial velocities formed in larger and higher density environments.
Why may planets form more frequently in this low-density environment? The favorite mechanism for planet formation from the proto-planetary disk is the core accretion scenario. This requires a few million years to form the planets. Perhaps proto-planetary disk may survive that long only in low dense environments because in denser ones there are nearby massive stars whose perturbation may destroy the disk before they could generate Jupiter-like planets. We would then expect that the Solar System might have formed in a low-density environment. Future observations, and especially those provided by the next release of the Gaia satellite data, will provide a clearer picture.