This article was originally published in February 2017, and was republished on October 8, 2019, at 3:50 pm on the occasion of Didier Queloz and Michael Mayor being awarded one half of the 2019 Nobel Prize for physics.
The exoplanet called 51 Pegasi b was discovered in 1995 to international acclaim. Though the first exoplanet had been discovered three years prior orbiting a pulsar, 51 Peg b was the first found to be orbiting a main sequence star. Main sequence is astronomy speak for a star’s being positioned on the Hertzsprung-Russell diagram, which categorises stars according to their brightness and colour. If a star is said to be on the main sequence, then it’s simply a conventional star. Indeed, 51 Peg b orbits 51 Pegasi, a star much like our Sun.
At the time, Michael Mayor and Didier Queloz detected the planet by studying how it was making its host star wobble – using changes in the measurement of the star’s radiation as a proxy. Their measurements indicated that 51 Peg b was a ‘hot Jupiter’, a giant planet orbiting much closer to its star than astronomers at the time thought was possible. Encouraged by Mayor’s and Queloz’s efforts, however, planetary astronomy soon pivoted to looking for more planetary hot Jupiters, also known as pegasean planets. Most of them have been identified using the same radial velocity method, also called Doppler spectroscopy, that the duo used. According to scientists at NASA’s Jet Propulsion Lab, “Astronomers now believe that large planets may form far from their stars and ‘migrate’ closer to the stars over millions of years.”
Over two decades after the discovery, a group of Dutch and American astronomers have made superior observations of the same planet, getting a better fix on its mass, orbital properties and atmospheric contents. Specifically, they have found that 51 Peg b’s atmosphere is 0.01% water – a sizeable proportion – and that it orbits its star at 133 km/s and weighs 0.476-times as much as Jupiter (give or take 7%). Their results were uploaded to the arXiv preprint server on January 25.
More than the specific measurements themselves, the significance of the astronomers’ accomplishment lies in their methods – as it did 21 years ago. When a planet orbits a star, the planet’s gravity will cause the star to start wobbling, as if it was moving in a tiny orbit around a slightly displaced centre of motion. When observed from Earth, the star will be discernible by a corresponding back-and-forth motion, understood in terms of the star’s radial velocity, the velocity at which it undertakes this motion.
When radiation is emitted by this star and observed from Earth, a ‘back’ motion will cause the radiation to become shifted ever so slightly to the redder end of the spectrum; a ‘forth’ motion will offset it towards the bluer end. These red- and blue-shifts, a consequence of the Doppler effect, are measured by a spectrometer to then determine how much the star is wobbling.
As it happens, the back-and-forth motion is also applicable to the planet as it moves in its orbit around the star. However, this is much harder to study because the star’s emissions are three to four orders of magnitude stronger and could overwhelm the planet’s own radiation when observed from telescopes on Earth.
Nonetheless, advancements since 1995 have allowed astronomers to build extremely sensitive telescopes – even better than the ELODIE spectrograph in southeast France that Mayor and Queloz used and which could pick up on a radial velocity of 7 m/s. The CRIRES instrument, located in north Chile and used by the Dutch-American team, is sensitive to velocities of 5-10 m/s. An upgraded version, called CRIRES+, aspires to a range of 2-3 m/s once it is ready in 2018. At the same time, CRIRES boasts of a resolution more than twice that of ELODIE’s.
Interestingly, Ignas Snellen, one of the authors of the new study on 51 Peg b, was among the first astronomers to demonstrate how the radial velocity method could be used to study the properties of planets, in 2010.
Getting, cleaning and mining the data
In the new study, Snellen and his colleagues studied 51 Pegasi for four hours in the infrared spectrum of radiation, as the planet ‘b’ swung around in its rapid four-day orbit. They gathered data about radiation emanating from the star as well as radiation bouncing off the planet from the star. Once they were done, they had to clean the data in two steps. The first was to remove all the information that corresponded to Earth’s atmosphere itself. The second, more difficult step was to find out which radiation in their data was coming off the planet and which from the star. The astronomers were aided in this exercises by two things, one a natural phenomenon and the other a tool.
The natural phenomenon is that planets usually possess those gases that strongly absorb radiation of a particular wavelength emitted a star. So when a spectroscope observes the radiation emitted by a star after it has bounced off such a planet, it will notice that a part of the radiation is missing. As a result, knowing what kind of radiation a star emits and which parts of it are absorbed by different substances can be used to determine what makes up the planet’s atmosphere.
Next, the tool was called cross-correlation. Essentially, Snellen and co. came up with a list of plausible compositions of 51 Peg b’s atmosphere and what their radiation would look like to a spectroscope. Then, they scanned the radiation signature of each composition against their CRIRES data to find the closest match. When this process was repeated for different combinations of water, gases and other materials, optimising at each step to accommodate other positive comparisons, the team arrived at the most likely and eventual scenario: 0.01% water, and no carbon dioxide or methane that CRIRES could detect.
The presence of water does not immediately suggest the presence of life, however. That will require further study, especially of 51 Peg b’s composition, surface characteristics, what other gases are present in the atmosphere and, finally, of the presence of biomarkers.
Their paper concluded thus: “After 21 years, the detailed nature of 51 Peg b is beginning to reveal itself, yet it remains an intriguing and extreme solar system.” And the radial velocity method – which allows us to study the surface-level properties of an object 473,000 billion km away – is only just getting started, its relevance controlled mostly by the ability of human engineering to build more sensitive instruments. Recently, the method was used to predict that the three exoplanets in the Wolf 1061 system 13.8 lightyears away were likely uninhabitable; to establish Proxima Centauri’s orbit is ‘bound’ to Alpha and Beta Centauri’s; and to understand how often brown-dwarf stars form. It seems radial velocities are set to feature in the answers to some of the most fascinating questions astronomers have to ask about the universe.