The majority of exoplanet detections over the past two decades have come from large transit surveys (such as Kepler and TESS), looking for the periodic dip in the brightness of a star which implies the existence of a planet passing between the star and us on Earth. But what happens once you’ve detected your planet and you want to learn more about it?
One key parameter we try to find out about exoplanets is their mass. By combining mass measurements with the radius of the planet (which we find from the transit observations) we can find a bulk density, and from this we can try to estimate the composition of planets: are they rocky and Earth-like, ice worlds, gas giants, or something else entirely?
The most common way of finding planet masses (and in fact the original method of exoplanet detection before we had transit surveys) is the radial velocity (RV) method. In fact the RV method is responsible for the very first detection of an exoplanet around a main sequence star in 1995, as well as many of the other early detections of planets — today the majority of planets are detected through dedicated transit surveys such as Kepler or TESS.
The RV method is where we make use of the effect of gravity to indirectly infer the presence of a planet. As a planet is orbiting a star, what is actually happening is that both the planet and star are orbiting a common centre of mass, this causes the star to ‘wobble’ a little. When the star ‘wobbles’ towards us here on Earth, the waves of light from the star are squashed up due to the Doppler effect, meaning they are shifted slightly to the blue end of the spectrum. When the star ‘wobbles’ away from us, the light is stretched out slightly, shifting it towards the red end of the spectrum. If we observe the spectrum of the star we can see whether the lines within the spectrum have been red- or blue-shifted. And if we do this at multiple times during the orbit, we can measure the forward and backward motion of the star (the velocity in the radial direction, hence radial velocity) and infer the presence of a planet (see graph). The higher the mass of the planet relative to the star, the more the star will ‘wobble’, so provided we know the mass of the star pretty well we can find the planet mass. Typically, these velocities range from several tens of centimetre per second for small exoplanets up to hundreds of meter per second for the most massive giant planets. The RV signal of the Earth-Sun system is only 9 centimetre per second, which is too small for us to detect in an exoplanet system with current capabilities.
In order to make these very precise spectral measurements over a long period of time we need highly-stabilised instruments known as spectrographs. Over the years, improving instrument capabilities have allowed astronomers to measure increasingly lower mass planets — for example the ESPRESSO instrument at the VLT has been shown the potential to reach precision levels that are as low as 10 centimetre per second. However, the problem now is that we are limited by the underlying ‘noise’ level from the star itself. Stars are known to exhibit different sources of activity, including star spots, flares, and more. All of these contribute to the RV signal and can cover up the signal from exoplanets in the system.
Exoplanet researchers haven’t yet found a perfect way to deal with these signals from stellar activity, but there are ongoing projects (for example here) to investigate different techniques which will help to find small mass planets in the future. So far, the RV method has discovered hundreds of exoplanets and is finding precise masses for an increasing number of small planets. This method also allows scientists to investigate the architecture of exoplanetary systems to try and understand the formation and evolution of these distant worlds. One day we may be able to use the RV method to find the mass of a truly Earth-like planet.