Science Release: Hubble observations used to answer key exoplanet questions

heic2206 — Science Release

25 April 2022

Archival observations of 25 hot Jupiters by the NASA/ESA Hubble Space Telescope have been analysed by an international team of astronomers, enabling them to answer five open questions important to our understanding of exoplanet atmospheres. Amongst other findings, the team found that the presence of metal oxides and hydrides in the hottest exoplanet atmospheres was clearly correlated with the atmospheres’ being thermally inverted.

The field of exoplanet science has long since shifted its focus from just detection onto characterisation [1], although characterisation remains extremely challenging. Thus far, the majority of research into characterisation has been directed towards modelling, or studies focusing on one or a few exoplanets. This new work, led by researchers based at University College London (UCL), used the largest amount of archival data ever examined in a single exoplanet atmosphere survey to analyse the atmospheres of 25 exoplanets. The majority of the data came from observations taken with the NASA/ESA Hubble Space Telescope. The lead author, Quentin Changeat, explains: “Hubble enabled the in-depth characterisation of 25 exoplanets, and the amount of information we learnt about their chemistry and formation — thanks to a decade of intense observing campaigns — is incredible.”

The science team sought to find answers to five open questions about exoplanet atmospheres — an ambitious goal that they succeeded in reaching.  Their questions probed what H– [2] and certain metals [3] can tell us about the chemistry and circulation of exoplanet atmospheres, and about planet formation. They chose to investigate a wide range of hot Jupiters [4], with the intention of identifying trends within their sample population that might provide insight into exoplanet atmospheres more generally. The study’s co-leader, Billy Edwards of UCL and the Commissariat à l’énergie atomique et aux énergies alternatives (CEA) said: “Our paper marks a turning point for the field: we are now moving from the characterisation of individual exoplanet atmospheres to the characterisation of atmospheric populations.”

In order to investigate their sample of 25 exoplanets, the team reanalysed an enormous amount of archival data [5], consisting of 600 hours of Hubble observations, which they complemented with more than 400 hours of observations from the Spitzer Space Telescope. Their data contained eclipses for all 25 exoplanets, and transits for 17 of them. An eclipse occurs when an exoplanet passes behind its star as seen from Earth, and a transit occurs when a planet passes in front of its star. Eclipse and transit data can both provide crucial information about an exoplanet’s atmosphere. 

The large-scale survey yielded results, with the team able to identify some clear trends and correlations between the exoplanets’ atmospheric constitutions and observed behaviour.  Some of their key findings related to the presence or absence of thermal inversions [6] in the atmospheres of their exoplanet sample. They found that almost all the exoplanets with a thermally inverted atmosphere were extremely hot, with temperatures over 2000 Kelvins. Importantly, this is sufficiently hot that the metallic species TiO (titanium oxide), VO (vanadium oxide) and FeH (iron hydride) are stable in an atmosphere. Of the exoplanets displaying thermal inversions, almost all of them were found to have H–, TiO, VO or FeH in their atmospheres. 

It is always challenging to draw inferences from such results, because correlation does not necessarily equal causation. However, the team were able to propose a compelling argument for why the presence of H–, TiO, VO or FeH might lead to a thermal inversion — namely that all these metallic species are very efficient absorbers of stellar light. It might be that exoplanet atmospheres hot enough to sustain these species tend to be thermally inverted because they then absorb so much stellar light that their upper atmospheres heat up even more. Conversely, the team also found that colder hot Jupiters (with temperatures less than 2000 Kelvins, and thus without H–, TiO, VO or FeH in their atmospheres) almost never had thermally inverted atmospheres. 

A significant aspect of this research was that the team were able to use a large sample of exoplanets and an extremely large amount of data to determine trends, which can be used to predict behaviour in other exoplanets. This is extremely useful, because it provides insight into how planets may form, and also because it allows other astronomers to more effectively plan future observations. Conversely, if a paper studies a single exoplanet in great detail, whilst that is valuable it is much harder to extrapolate trends from. An improved understanding of exoplanet populations could also bring us closer to solving open mysteries about our own Solar System. As Changeat says: “Many issues such as the origins of the water on Earth, the formation of the Moon, and the different evolutionary histories of Earth and Mars, are still unsolved despite our ability to obtain in-situ measurements. Large exoplanet population studies, such as the one we present here, aim at understanding those general processes.”


[1] Exoplanet characterisation involves investigating the physical properties (such as radius) and chemical properties (such as atmospheric composition) of an exoplanet. It is crucial both for better understanding planet formation and evolution, and for determining whether complex processes — such as the evolution and maintenance of life — could be possible on an exoplanet.

[2] H– is a negative hydrogen ion that has been formed by the dissociation of a molecule such as H2 (hydrogen) or H2O (water). These molecules dissociate at very high temperatures (over 2500 Kelvins, or 2227 °C). 

[3] In astronomy, a ‘metal’ is defined as any element with more protons in its nucleus than hydrogen or helium (which have one and two protons respectively). Thus, ‘metallicity’ is a measure of how many elements or molecules are present that are not hydrogen or helium. 

[4] Hot Jupiters are an informal class of exoplanets with short-period orbits (orbiting their parent star in roughly ten days or less), and large, inflated gassy atmospheres. They are of particular interest because i) they are relatively easy to detect and ii) there is no hot Jupiter within our Solar System, so we have to look to exoplanets to study them. 

[5] Archival data are data that the team did not collect specifically for this research, but were collected by another team(s) for a different initial purpose and are now publicly available. The reanalysis and repurposing of archival data can be an extremely cost and environmentally effective way of getting new results. 

[6] A thermal inversion is a natural phenomenon where a planet’s or exoplanet’s atmosphere does not steadily cool with increased altitude, but instead reverses from cooling to heating at a higher altitude. Thermal inversions are thought to occur because of the presence of certain metallic species in the atmosphere. For example, the Earth’s atmosphere has a thermal inversion that is due to the presence of ozone (O3). 

More information

The Hubble Space Telescope is a project of international cooperation between ESA and NASA.

These results have been published in the Astrophysical Journal Supplement Series.

The international team of astronomers in this study consists of: Q. Changeat (University College London, United Kingdom [UCL]), B. Edwards (UCL and Commissariat à l’énergie atomique et aux énergies alternatives [CEA], Université Paris-Saclay, Université de Paris, France), A. F. Al-Refaie (UCL), A. Tsiaras (UCL and Osservatorio Astrofisico di Arcetri, Firenze, Italy), J. W. Skinner (Queen Mary University of London, United Kingdom), J. Y-K. Cho (Center for Computational Astrophysics, Flatiron Institute, New York, USA), K. H. Yip (UCL), L. Anisman (UCL), M. Ikoma  (National Astronomical Observatory of Japan, Tokyo, Japan and The Graduate University for Advanced Studies [SOKENDAI], Tokyo, Japan), M. F. Bieger (University of Exeter, United Kingdom), O. Venot (Université de Paris and Université Paris Est Creteil, France), S. Shibata (University of Zurich, Switzerland), I. P. Waldmann (UCL), G. Tinetti (UCL).

Image credit: ESA/Hubble, N. Bartmann



Quentin Changeat
University College London
United Kingdom

Bethany Downer
ESA/Hubble Chief Science Communications Officer