When it comes to finding new worlds, NASA's Kepler spacecraft hogs the headlines, having racked up thousands of exoplanet discoveries since its launch in 2009. But before Kepler, the workhorses of exoplanet identification were ground-based instruments that measure tiny stellar wobbles caused by the gravity of an orbiting planet. They are now undergoing a quiet renaissance. The new generation of these devices may be precise enough to find a true Earth twin: a planet with the same mass as ours, orbiting a sunlike star once a year. That's something Kepler—sensitive to planet size, but not mass—can't do.
On 9 December 2017, the Extreme Precision Spectrometer (EXPRES) took its first view of the sky on the Discovery Channel Telescope in Arizona. And in October 2017, the Echelle Spectrograph for Rocky Exoplanet and Stable Spectroscopic Observations (ESPRESSO) began operating at the European Southern Observatory's (ESO's) Very Large Telescope in Chile. Nearly two dozen other instruments are either under construction or have recently begun service. "It's now clear that exoplanets are a major part of astronomy," says astronomer Jason Wright of Pennsylvania State University in State College. "So every major observatory needs a high-resolution spectrograph."
Such spectrographs spread starlight out into its array of spectral colors, which contains dark lines at wavelengths where gases in the star's atmosphere absorb light. Astronomers then look for tiny oscillating Doppler shifts in these lines over time, caused by planetary tugging.
The technique works best for a massive planet orbiting close to its star, because its gravitational tug will be stronger. As a result, early discoveries—such as the first exoplanet to be found, in 1995—were usually "hot Jupiters," giant planets in tight orbits. Then came the High Accuracy Radial velocity Planet Searcher (HARPS) at ESO's La Silla Observatory in Chile in 2003, a second-generation spectrograph that enabled astronomers to find smaller planets in wider orbits. (It has spotted about 130 to date.)
In recent years, however, a new "transit" technique, led by Kepler, began to dominate. Kepler stared at 145,000 stars in one part of the sky and looked for dips in stellar brightness when a planet passed in front. Although Kepler has been prolific, identifying thousands of exoplanets and gauging their sizes, the planets and their stars are mostly too far away for ground-based spectrographs to determine their masses. Over the coming year, the launch of Kepler's successor, the Transiting Exoplanet Survey Satellite (TESS), along with Europe's Characterising Exoplanets Satellite (CHEOPS), will change all that. They will scour the sky for transits of nearby bright stars—perfect for ground-based follow-up. "To understand them we have to know their mass, that's absolutely fundamental," says René Doyon of the University of Montreal in Canada.
The new spectrographs should be able to weigh many of the planets found by TESS and CHEOPS. Astronomers boosted the instruments' precision in part by isolating them from mechanical and thermal noise—mounting them inside vacuum vessels that are separated from the telescope and piping in light through optical fibers. They also improved the reference spectrum against which the slow shifts in the spectral lines are gauged.
Instruments such as HARPS rely on reference spectra produced by a thorium-argon lamps. But these produce a jumble of spectral lines of varying brightness. The new wave of spectrographs, including ESPRESSO and EXPRES, instead use laser frequency combs, which duplicate a single spectral line from a laser to form an ultraprecise reference grid, with lines of equal brightness spaced at regular intervals.
With these improvements, says lead scientist Francesco Pepe of the University of Geneva in Switzerland, ESPRESSO aims to measure stellar movements as slow as 10 centimeters per second (roughly the speed of a giant tortoise). That would be a factor of 10 better than HARPS, and is exactly the motion expected to be caused by an Earth twin in an Earth-size orbit around a sunlike star.
At this sort of precision, pulsing gases in the star's atmosphere can swamp the signal of any wobble. With some of the new spectrographs, researchers hope to disentangle the wobble from the noise by comparing spectral shifts across a range of wavelengths. But Pepe expects the noise to limit ESPRESSO to planets that are about three or four times as heavy as Earth. Only if the star is much less massive and more susceptible to a planet's tug might ESPRESSO see Earth-size planets.
Other spectrographs aim for increased sensitivity to small planets by focusing on some of the smallest stars: red dwarfs. Red dwarfs emit most of their light in the infrared. Yet Earth's atmosphere allows only a narrow range of near-infrared wavelengths to reach the ground, posing a challenge for instrument builders. "There is a big push to develop instruments," says Doyon, who leads a team building the Near-InfraRed Planet Searcher to work alongside HARPS on ESO's 3.6-meter telescope in Chile.
The new instruments could even enhance transit studies. Besides searching for the spectral wobbles, they might see additional spectral lines caused by starlight passing through the atmosphere of a transiting planet. Transit spectroscopy with existing instruments has already revealed gases such as water and carbon dioxide in exoplanet atmospheres, and astronomers are now using it to look for signs of life. If the new spectrographs can characterize planets as well as find them, the renaissance could become a revolution.