Background

We know over 5000Link opens in a new window extrasolar planets, with the overwhelming majority orbiting solar-like stars. Thinking about the long-term evolution of these planets, some questions come to mind:

• What is the future of these planetary systems as their host stars evolve off the main sequence?
• What will happen to our solar system once the Sun dies?
• And what information can we learn from those evolved planetary systems that is otherwise inaccessible?

Within this ERC-funded project, we will be assembling a sample of white dwarfs that is an order of magnitude larger than the current state-of-the art, and we will use high-quality spectroscopy and photometry from large international facilities to identify ~1000 evolved planetary systems. Our analysis of these data as well as dedicated follow-up observations will answer the questions above, and greatly advance our understanding of the final fate of planetary systems.

The future evolution of planetary systems

As stars hosting planets evolve into red giants, and eventually into white dwarfs, they lose substantial amounts of mass, these lost layers of mass can be seen for some time as planetary nebulae. Consequently, the orbits of the planets, and all minor bodies, asteroids, and comets, will widenLink opens in a new window, with many of them moving far beyond the maximum radius that the star will reach during the red giant phase. In our solar system, Mars, the asteroid belt, and all the giant planets will escape evaporationLink opens in a new window. In other words, in ~6Gyr, the solar system will consist of a white dwarf with one rocky planet, four giant planets, and countless asteroids and comets. Thus, we expect that the 1000s of known planetary systems will eventually leave behind white dwarfs bearing the remnants of planetary systems. And by analogy, we can expect that many of the white dwarfs in the Milky Way were planet hosts in the past. Our goal is to identify ~1000 white dwarfs that have detectable evidence of surviving planetary bodies.

Left: Artist impression of the evolution of a planetary system similar to our solar system. Four rocky planets are orbiting a sun-like star, the presence of water on the fourth planet implies that it is suitable to host life (top panel). Once the host star evolves into a red giant and begins to lose mass, the orbits of the planets widen. This may lead to instabilities, orbit crossings and collisions that create vast amounts of debris, or the ejection of some of the planets (middle panel). Planets, moons, asteroids or comets can be gravitationally scattered onto highly eccentric orbits that takes them deep into the gravitational field of the white dwarf where they are disrupted by tidal forces, eventually forming a debris disc around the white dwarf (bottom panel). © Mark Garlick and University of Warwick.

Extrasolar asteroids & the bulk composition of planets

The first detections of evolved planetary systems came not in the form of their planets, but asteroids. Just as we know about sun-grazing asteroids, ever now and then an extrasolar asteroid will have its orbit perturbed by a larger body, venture too close to the white dwarf, and by tidally shredded and form a debris disc of dustLink opens in a new window and/or gasLink opens in a new window. Those dust discs are detected as infrared excess emissionLink opens in a new window, the gas discs are identified through highly unusual emission lines of metals such as Ca or Fe in their optical spectra.

As the dust and gas falls onto the the white dwarf, it pollutes its otherwise pristine hydrogen (or in some case helium) atmosphere, and becomes spectroscopically measurable. In other words, through spectroscopy of asteroid-debris bearing white dwarfs, we can measure the chemical bulk compositionLink opens in a new window of extrasolar rocky material, the building blocks of terrestrial planets like our Earth – an insight that can not be learned from “normal” extrasolar planets. "Metal-polluted" white dwarfs have been known for almost a century, their link to the remnants of planetary systems has become clear only over the past few years. This method of studying the make-up of exo-planetary material is unrivalled, as observations of planets orbiting main-sequence stars yields only information on their atmosphere composition and the bulk density of the planets.

This animation shows the white dwarf WDJ0914+1914 and its Neptune-like exoplanet. Since the icy giant orbits the hot white dwarf at close range, the extreme ultraviolet radiation from the star strips away the planet’s atmosphere. While most of this stripped gas escapes, giving the planet a comet-like tail, some of it swirls into a disc, itself accreting onto the white dwarf. © ESO/M. Kornmesser

Planets orbiting white dwarfs

Whereas 100s of white dwarfs are known to still have asteroid belts, the detection of planets has proven more difficult. The reason is that the standard methods to identify planets are much, much harder in the case of white dwarfs. Measuring the Doppler wobbleLink opens in a new window requires high-resolution spectroscopy, and most white dwarfs are too faint for current instruments (that will change with the E-ELTLink opens in a new window). Similarly, the detection of transitsLink opens in a new window requires a line of sight very closely aligned with the orbital plane of the planet. Because the chances of that are relatively low, vast numbers of stars need to be monitored photometrically, and, again because of their faintness, that has remained a challenge (but see below).

In 2019, we discovered the first evidence for a giant planetLink opens in a new window in a close orbit around a white dwarf, that is being evaporated by the strong ultraviolet radiation from the white dwarf. This led us to think again more carefully about the future of the solar system, and we realised that the white dwarf left behind by the Sun will be be so luminous in the ultraviolet that it will drive detectable amounts of mass lossLink opens in a new window from Jupiter, Saturn, Uranus and Neptune. Detectable by some alien astronomer who, in 6Gyr, points their telescope at the solar white dwarf.

The first transiting giant planet candidate has been found from the analysis of TESS observations of a white dwarfLink opens in a new window. This is extremely exciting, as this white dwarf is only ~25pc away, so such systems may not be so rare after all, and the Zwicky Transient FactoryLink opens in a new window, and later the Rubin ObservatoryLink opens in a new window have the potential to find many more.

Current research

White dwarfs are small - Earth-sized - and therefore very faint. Finding white dwarfs among the much brighter main-sequence stars has been very difficult in the past, however, the parallaxes that the ESA Gaia mission measured for about 1.5 billion stars made it possible to identify ~360,000 white dwarfs across the entire skyLink opens in a new window. We are members of the SDSS-V, DESILink opens in a new window, and WEAVELink opens in a new window collaborations, and lead the spectroscopic follow-up of the Gaia white dwarf sample. Combined, these three surveys obtain spectroscopy of over 100 white dwarfs per night. We are also partners in ZTF-II, providing time-series photometry of all the white dwarf candidates.

• The vast amount of spectroscopic and photometric data requires to automate the data flow, classification, and analysis as much as possible, and we are currently developing a software framework to seamlessly carry out these tasks.
• On average, about one per cent of white dwarfs are accreting particularly large amounts of planetary debris to allow very detailed abundance studies, and we expect to model at least 1000 such stars over the course of the project. This requires custom-computed grids of model spectra, and we are automating the generation of these grids using the model atmosphere code of Detlev Koester. An important task in the analysis of the white dwarf spectroscopy is a careful assessment of the systematic uncertainties.
• The small subset of white dwarfs that host gaseous debris disks provide the opportunity to measure the composition of the material of the disk "in situ" . We are developing a photo-ionisation model for these disks that are heavily irradiated by ultraviolet photons from the white dwarf to carry out these measurements from modelling the emission lines from the disk.
• We are screening the spectroscopic survey data obtained so far for particularly interesting systems, which will be analysed and published with high priority. An example is the discovery of 21 magnetic white dwarfs exhibiting Zeeman-split Balmer emission linesLink opens in a new window, boosting the size of this new and enigmatic class by a factor six. The origin of the emission lines is still unclear, but one models links them to close-in planets.
• Our group has been leading over a decade of far-ultraviolet spectroscopic surveys of white dwarfs to probe the frequency of evolved planetary systems, and assess their bulk compositions. We have published the largest study of the atmospheric parameters of white dwarfs with hydrogen-dominated atmospheres, and are now working on a detailed analysis of the abundances of those stars where we detect the signatures of planetary debris.
• We have been allocated several new programs on the Hubble Space Telescope to study the abundances of a new gaseous debris disc, the first young massive magnetic white dwarf harbouring remnants of a planetary system, and a legacy survey for evolved planetary systems around white dwarfs within 100pc.
• We have been allocated a study of the mineralogy of six debris discs using the James Webb Telescope, which follows in the footsteps of the first JWST observations of a dusty debris disc around a white dwarf.
• We organised a special session on "Planets not orbiting the main sequence" at the 2023 meeting of the European Astronomical Society in Krawkow, and gave a short interview on the wide range of exciting science related to white dwarfs.
• We have assembled a data base of all published abundance measurements of metal-polluted white dwarfs, and developed tools for the statistical analysis of these data. Shown below is an example of the output of this data base, illustrating the number of white dwarfs for which a particular element has been detected: whereas there are ~1400 white dwarfs with a Ca detection, only a few tens have recorded abundances for the major planet-forming elements O, Si, Mg, and Fe.
  
  
  We will add to this the measurements obtained from the SDSS-V, DESI and WEAVE spectroscopy, which will eventually result in an abundance data base that is at least as large as that of meteorites in the solar system, which has been the foundation of all planet formation models.

To facilitate the analysis of white dwarfs by the wider community, we will release the spectroscopic and photometric survey data for all white dwarfs in a structured format and we will make all our software publicly available.