Artist impression of Proxima Centauri c. Press “HD” on the player for the best image quality (E. Tasker).

The discovery of Proxima Centauri b in 2016 caused a flood excitement. We had found an extrasolar planet around our nearest star, making this the closest possible world outside of our solar system!

But despite its proximity, discovering more about this planet is difficult. Proxima Centauri b was found via the radial velocity technique, which measures the star’s wobble due to the gravity of the orbiting planet. This technique gives a minimum mass, the average distance between the star and planet and the time for one orbit, but no details about conditions on the planet surface.

If the planet had transited its star, we might have tried detecting starlight that passed through the planet’s atmosphere. This technique is known as transit spectroscopy, and reveals the composition of a planet’s atmosphere by detecting what wavelengths of light are absorbed by the molecules in the planet’s air. But searches for a transit proved fruitless, suggesting the planet’s orbit did not pass in front of the star from our viewpoint.

The radial velocity technique measures the motion of the star due to the gravity of the planet. As the star moves away from the Earth, its light becomes stretched and redder. As it moves back towards Earth, the light shifts to bluer wavelengths. The technique gives the planet’s period, distance from the star and its minimum mass. (E. Tasker)

Another option for planet characterization is to capture a direct image of the planet. This is one of the most exciting observational techniques, as it reveals the planet itself, not its influence on the star. Temporal changes in the planet’s light could reveal surface features as the planet rotates, and if enough light is detected to analyze different wavelengths, then the atmospheric composition could be deduced.

But direct imaging requires that the planet’s light can be differentiated from the much brighter star. With our current instruments, Proxima Centauri b orbits too close to its star to be distinguished. This seemed to close the door on finding out more about our nearest neighbors, until the discovery of a second planet in the system was announced early this year.

Also identified via the radial velocity technique, Proxima Centauri c has a minimum mass of 5.8 Earth masses. It sits further out than its sibling, with a chilly orbit that takes 5.2 years. On the sky, the angle between the planet and star is 1.14 arcseconds, where 1 degree is 3600 arcseconds. For comparison, the width of your index finger raised to the sky with your arm outstretch is about 1 degree. Directly imaging a planet that is only separated from its star by 1.14 arcseconds is therefore challenging, but not necessarily unfeasible. Last week, a paper came out that had tried to give it a go.

Led by Dr Raffaele Gratton at INAF – Osservatorio Astronomico di Padova in Italy, the Astronomy & Astrophysics journal paper describes the detection of a possible candidate for Proxima Centauri c in images taken at near infrared wavelengths. The paper is immensely cautious about the discovery, noting that the result was not sufficiently conclusive to rule out a false detection from noise created due to natural variations in the amount of light each pixel receives.

So why it is so hard to be sure that you are seeing a planet?

Left: planet phases. Right: planet orbit orientation, showing the orbit inclination and longitude of the ascending node. (E. Tasker)

One problem is that you do not know exactly where the planet should appear. The radial velocity technique tells you the size of the orbit, but it does not reveal the orbit’s orientation.

Imagine holding a hoop in your hands. This is the planet’s orbit. Tilt that hoop up or down and that it the orbit inclination. In a tilted position, rotate the orbit so the upwards tilt points in a different direction. That angle has the impressive name of the longitude of the ascending node.

Without knowing either of these angles, the path the planet will take across the sky is unknown. So if the light from the planet is not very bright, it is very difficult to be certain you are looking at the right object.

However, Gratton’s team were able to drastically narrow down the search area by using a clever trick. If you draw a circle around the star at the planet’s average distance, the planet will cross the circle during its orbit regardless of the orbit’s orientation. What is more, this interception always occurs when the planet is at a quarter-phase, when a quarter of its surface it being illuminated by the star.

While the distance between the star and planet appears to change due to the projection of the orbit on the sky, it always equals the true distance when the planet is at quarter-phase. This is a difficult to picture, but is shown in the animation below. The planet orbit is rotated around different angles, but it always crosses the circle at the planet’s true distance from the star.

Regardless of orbit inclination, a planet will intercept a circle around its host star with a radius equal to the true separation between the planet and star. This can be used to identify a planet in direct imagining data. (E. Tasker)

The team had five observations taken in 2018 when the planet was at quarter-phase and these revealed a bright point in the direct images on the expected circle. But is this truly the planet or just random noise?

A good sign was that the planet candidate was the brightest point in the whole image, not just around the ring which filled just 6% of the area. The five observations were also rotated to allow for the planet’s expected motion, which would change the relative noise distribution and lower the chances of a stray peak being maximised. The team also rotated the images a full 360 degrees to find the maximum possible bright peak due to random noise. They discovered only one example where a randomly generated point was brighter than the planet candidate. This suggested that the chance of the planet candidate being a stray artefact was only about 0.9%.

But the team remained cautious. The promisingly low percentage assumed that the distribution of noise is the same over all the image, whereas this is unlikely to be true. Moreover, the fluctuations in the peak strength between images was consistent both with a genuine detection and a noise artefact.

Figure 1 from Gratton et al. 2020. Median of the five images collected in 2018. The green circle marks the search area for the planet. The blue line is the orbit assuming an inclination of 120 degrees. The lower yellow circle is the candidate for Proxima Centauri c, the upper circle indicates the speed and direction of the orbit.

If the bright peak was Proxima Centauri c, then its location on the ring revealed the orbit orientation. This allowed the expected position of the planet to be mapped across the sky, and compared to other observation when the planet was not at a quarter-phase. Images taken in 2016 and 2019 did not align exactly with the predicted planet, but they did show a bright peak close to the estimated location. This could mean the orbital properties needed a minor adjustment, that the planet was not detectable at these different times due to poorer atmospheric conditions or that the planet had never been imaged at all.

There was an additional intrigue. If Proxima Centauri c was the observed bright point and its orbit inclination was now known, then the planet’s true mass could be calculated. The radial velocity detection provides a lower limit on the mass as only the star’s wobble that is directed towards Earth is measurable. Without knowing the orbit inclination, there is no way of telling if that it the full extent of the stellar motion or if the tug from the planet is inclined away from our view.

An inclination of 120 degrees would make Proxima Centauri c a 7.2 Earth mass planet. While this is a super Earth, it is not large enough to explain the brightness of the observed peak. If the peak is genuine, then Proxima Centauri c might be surrounded by extensive circumplanetary material, such as a dust cloud or ring system. The size of these rings is estimated at about 5 times the radius of Jupiter. For comparison, Saturn’s outer tenuous E-ring extends to seven Jupiter radii, but the bright A and B rings only go out as far as two Jupiter radii.

With the planet’s light being at the edge of detectability, the authors ultimately declared the find interesting, but that further observations were required to confirm the detection. These may come with the next data release from ESA’s Gaia, which should be able to detect the presence of the planet through an exceedingly accurate measurement of the variations in the star positions in the sky.

If the detection does prove to be Proxima Centauri c, it will be the first planet detected by radial velocity to be imaged. This is an exciting step forward, as worlds further from their star are unlikely to transit and will be discovered by radial velocity. This includes any Earth analogues, where finding out about their nature will depend on the advances made in direct imaging.