Research

The leading explanation is high-eccentricity migration. A giant planet that formed a few times further from its star than Earth is from the Sun is knocked onto an elongated orbit, and tides raised at each close approach to the star then shrink the orbit and make it circular. Each way of knocking a planet inwards tilts its orbit by a characteristic amount, and many hot Jupiters, especially around hot, fast-spinning stars, orbit far out of line with their star's equator. Young stars, whose planets have only just formed, show these processes closest to their start.

Fast-spinning stars offer a direct way to measure the tilt. Their rotation makes them bulge at the equator, which is then cooler and dimmer than the poles: an effect called gravity darkening. A planet crossing such a star blocks a different amount of light at each point along its path, so the shape of the transit records the true angle between the orbit and the star's equator, not just its projection on the sky. The signal is subtle, and the result depends on how carefully the light curve is modelled.

With light curves from CHEOPS and Spitzer, I confirmed that the hot Jupiter MASCARA-1 b has a near-polar orbit, and showed that observing the transit in two colours makes the measurement far more reliable. I am now analysing all the existing multicolour light curves of planets crossing fast-spinning stars in one consistent way, to look for patterns in their tilts and orbital shapes that point to how they migrated. JWST's light curves are about ten times more precise, so the same method now works for smaller planets.

I was part of the team that found the six planets in 2021. They range from just larger than Earth to almost three times its size, and five of them form a chain of Laplace resonances, like the one linking three of Jupiter's moons. I lead a JWST programme that observed three of the planets with the NIRSpec spectrograph. Water, carbon monoxide and carbon dioxide freeze out at different distances from a young star, so their proportions in each atmosphere show whether the planet formed where it is now or moved there.

With the same instrument, Chloe Fisher and I looked for an atmosphere on TOI-1685 b, a hot rocky planet. We found no clear sign of one: its spectrum fits a bare rock or an atmosphere of heavy molecules.

SPECULOOS is a network of robotic one-metre telescopes, led by the University of Liège with Cambridge, Birmingham, MIT, Bern and ETH Zurich as partners. It succeeds TRAPPIST, whose smaller telescopes found the seven Earth-sized planets of TRAPPIST-1.

I develop the pipeline that turns each night's images from the SPECULOOS telescopes into the light curves used to find planets. The data are kept in an archive at Cambridge. The pipeline has made possible two SPECULOOS planet discoveries and the confirmation of dozens of planets first found by TESS.

CHEOPS, which ESA launched in 2019, watches bright, nearby stars already known to have planets. I am on its science team. I led its largest observing programme, which refined the sizes of small planets in systems with several planets, and I have led the analysis of CHEOPS data in papers on new multi-planet systems and on the atmospheres of hot Jupiters. I also coordinate the consortium's proposals for JWST time to follow up its most interesting planets.

At Queen's University Belfast I designed and led a campaign to observe these eclipses at wavelengths that Hubble and Spitzer did not routinely cover. I won telescope time as principal investigator on the VLT, Gemini-North and other telescopes, coordinated a team of QUB astrophysicists to carry out the observations, analysed the data with software I wrote, and worked with theorists to interpret the results.

I measured an upper limit of 181 parts per million on the depth of KELT-9b's eclipse in the near-ultraviolet U band, the first such measurement from the ground. It constrained the planet's thermal emission and how its temperature changes with height. Eclipses of WASP-12b in the i band hinted that a hot Jupiter's thermal emission can change over time. The campaign also included multicolour observations of WASP-103b and red-optical emission spectroscopy of KELT-16b and WASP-121b.