Cool Star Lab Contributes to Discovery of “Extreme” Metal-poor Brown Dwarfs

Cool Star Lab researchers, including current and former members, have contributed to the discovery of two brown dwarfs with unusual metal abundances, suggesting that they are part of a long-sought, ancient population of brown dwarfs formed early in the history of the Milky Way Galaxy.

Illustration of a brown dwarf with a Galactic backdrop by study co-author and citizen scientist William Pendrill

Most of the stars and brown dwarfs near the Sun contain elements in the same proportion as our host star. Hydrogen and helium making up about 98% of the Sun’s mass, and all other elements – which astronomers collectively refer to as “metals” – comprising a mere 2%. Yet a tiny fraction of stars – about 0.3% – are even more deprived of heavy elements. These are referred to as metal-poor “subdwarfs”, the “sub” relating to their position below the stellar Main Sequence on the Hertzsprung-Russell color-magnitude diagram. Subdwarfs are typically the most ancient stars in the Milky Way Galaxy, formed before massive stars were able to produce heavy elements that seeded later generations of stars. They mostly populate the halo of our Galaxy, an extended, roughly spherical extended distribution of stars (in contrast to the Galaxy’s “young disk”) formed either before the Galaxy had its current shape, or as old stars were flung out by dynamical encounters with giant molecular clouds. Subdwarfs therefore provide a window into the early star formation and dynamical history of our Galaxy.

Since brown dwarfs were first conjectured nearly 60 years ago and first discovered 25 years ago, scientists have wondered whether these objects could have formed in the metal-poor environment of the early Galaxy. The mechanism(s) that create brown dwarfs remains an open question, as these low-mass objects have difficulty forming by the standard model of star formation: the gravitational collapse and fragmentation of giant molecular clouds. A molecular cloud with fewer metals makes this process even more difficult. Nevertheless, metal-poor brown dwarfs exist. The first to be identified, a metal-poor late L dwarf 2MASS J0532+8246, was discovered by our team in 2003. Theoretical analysis of this source indicates it is right on the mass boundary between stars and brown dwarfs. In fact, evolutionary models predict that most brown subdwarfs, being old, should be much colder and be members of the T and Y spectral classes. While a few “modestly” metal-poor T subdwarfs have been identified over the past decade, we had not yet found the unambiguous brown dwarf equivalents of the Galaxy’s halo population.

This has now changed with the discovery of two exceptionally metal-poor T dwarfs by the citizen science Backyard Worlds: Planet 9 project. Backyard Worlds: Planet 9 uses data from NASA’s Wide-Field Infrared Survey Explorer (WISE) and Near-Earth Object Wide-Field Infrared Survey Explorer (NEOWISE) satellites to uncover faint, cool, nearby objects. These objects are actually discovered by citizen scientists through the Backyard Worlds: Planet 9 Zooniverse platform, while the astronomers on the team confirm the discoveries through follow-up observations and analysis.

Sky images of the T subdwarf discovery WISEA J181006.18−101000.5 taken with WISE and NEOWISE in two epochs: mid-2010 and early 2017. The target is the orange source that moves slightly to the right between these epochs, a consequence of its high velocity and proximity to the Sun (from Schneider et al. 2020).

The two discoveries, WISEA J041451.67−585456.7 and WISEA J181006.18−101000.5, were both identified as high-priority sources for follow-up, as they have large proper motions (angular motion across the sky) and unusual colors. The Cool Star Lab team initially targeted WISEA J181006.18−101000.5 with the Keck/NIRES spectrograph in August 2019, but were unable to obtain spectrum of the source in the crowded field of view. However, imaging data obtained in this run allowed collaborators Eric Mamajek and Federico Marocco at the Jet Propulsion Laboratory to obtain a spectrum on month later using the Palomar/TripleSpec spectrograph. WISEA J041451.67−585456.7 was observed with the Magellan/FIRE spectrograph in February 2020 by the study’s lead author Adam Schneider (Adam Burgasser was Co-PI on the construction of this instrument).

The infrared spectra of WISEA J041451.67−585456.7 (left) and WISEA J181006.18−101000.5 (right) taken with the Magellan/FIRE and Palomar/Triplespec instruments, in black; compared to known normal-metallicity and metal-poor L and T dwarfs. The fact that these known templates provided poor matches to the data suggested that the discoveries were truly unique sources (from Schneider et al. 2020).

The spectra obtained were very unusual. While the infrared spectra of T dwarfs are distinguished by strong absorption bands of methane and water, which can only form in atmospheres cooler than about 1200 ºK, these spectra were mostly smooth, with only weak hints of methane and water, and two strong absorption but unusually shaped features at 1.1 µm and 1.4 µm. 2MASS J0532+8246 had shown similar spectral peculiarities, so Cool Star Lab graduate student Roman Gerasimov generated a new set of low-temperature, metal-poor atmosphere models of using the PHOENIX code with the San Diego Supercomputer Center to see if these could reproduce the data. Indeed, the key features of the observed spectra were confirmed: the strong absorption bands at 1.1 µm and 1.4 µm are indeed caused by methane and water in a low-temperature atmosphere; while the smooth part of the spectrum is caused by enhanced absorption from molecular hydrogen found in unusually high-pressure atmospheres, a consequence of the lack of heavy elements in the gas. The best-fit models have metal abundances 10-100 times less than that of the Sun. Roman’s analysis therefore confirmed these sources as the first “extremely” metal-poor T dwarfs.

A comparison of the same spectra to metal-poor atmosphere models generated by Roman Gerasimov. While these do not provide perfect fits, they can explain the combination of strong absorption features at 1.1 µm and 1.4 µm and the smooth spectrum at longer wavelengths as arising from a depletion of heavy elements (from Schneider et al. 2020).

While these may be the first ancient, metal-starved brown dwarfs to be found, they are likely not the last. Backyard Worlds: Planet 9 citizen scientists continue to search the WISE and NEOWISE data for faint moving objects, and the recent addition of CatWISE data to the sample (containing over 12 years of infrared satellite monitoring) will make it easier to spot out faint moving sources. In addition, future deep survey data obtained with the Vera Rubin Observatory will expand our sensitivity to these intrinsically rare sources. Not only will these sources probe the early history of our Galaxy, they will also help us understand the role of elemental abundances on thermal evolution, gas chemistry, and cloud formation in cool brown dwarfs. More to come!

The discovery of these two metal-poor T dwarfs is reported in Schneider, Burgasser, Gerasimov, et al. 2020 “WISEA J041451.67-585456.7 and WISEA J181006.18-101000.5: The First Extreme T-type Subdwarfs?”, accepted for publication to the Astrophysical Journal. This paper is available on the arXiv at Other coauthors on the study not mentioned above include Jonathan Gagne, Sam Goodman, Paul Beaulieu, William Pendrill, Austin Rothermich, Arttu Sainio, Marc J. Kuchner, Dan Caselden, Aaron M. Meisner, Jacqueline K. Faherty*, Chih-Chun Hsu*, Jennifer J. Greco, Michael C. Cushing, J. Davy Kirkpatrick, Daniella Bardalez Gagliuffi*, Sarah E. Logsdon*, Katelyn Allers, and John H. Debes. (* = current/former Cool Star Lab members). This work used the Extreme Science and Engineering Discovery Environment (XSEDE) Comet cluster at the San Diego Supercomputer Center (program AST190045), which is supported by National Science Foundation grant number ACI-1548562. The original NASA press release can be found at

Citizen Scientist Discovers Dusty Debris Disk Around White Dwarf

Citizen scientist Melina Thévenot of Germany helped the Backyard Worlds/Planet 9 program discover a unique white dwarf with a dusty debris disk, and observations made by Cool Star Lab members with the Keck/NIRES instrument were critical its confirmation. The work, led by STScI astronomer John Debes, was reported in Astrophysical Journal Letters today.

Since 2017, the Backyard Worlds project has been engaging citizen scientists to search through data from NASA’s WISE mission to identify overlooked stars in the vicinity of the Sun. These have mostly been cold brown dwarfs, of which the project has found more than 1,000 too date – more than one a day! But it also picks up other dim, red things, in this case the white dwarf LSPM J0207+3331.

White dwarfs are normally “blue” due to their high surface temperatures (they are after all the cores of spent stars), but this white dwarf is surrounded by a complex disk of dusty debris, likely the result of the tidal disruption of an orbiting planet or asteroid (the same process is likely responsible for the rings around Saturn and other giant planets). This disk, heated by the white dwarf, glows in the infrared, allowing it to show up in WISE. While tidal debris disks aren’t new around white dwarfs (Cool Star Lab’s Carl Melis is specialist in this area), both the structure of this disk—which appears to be made of several distinct ring-like components—and that age of the white dwarf are surprising.

Very little was known about J0207+3331 prior to its identification by Melina in the Backyard Worlds program; only one prior paper had identified it as a high proper motion star. After an initial attempt to measure its spectrum was foiled by bad weather, it was therefore up to Cool Star Lab members Adam Burgasser & Jon Rees to get the necessary spectral data. Using the newly-commissioned NIRES instrument on Keck (during admittedly more not so great weather), Adam & Jon managed to measure the near-infrared spectrum of the source, which was largely consistent with a hot blackbody with slight uptick at the red end. This tiny bit of near-infrared excess, and the much greater mid-infrared excess in WISE photometry, could arise from several things, including an unseen brown dwarf companion (much like the first L dwarf ever discovered, GD 165B). However, we were able to show that the combined NIRES spectrum and WISE photometry were inconsistent with any white dwarf-brown dwarf combination, leaving a debris disk as the best model. (For once, Adam was happy not to find a brown dwarf!)

(Left) Analysis of our NIRES spectrum shows that any brown dwarf companion to J0207+0331 would be too small (blue dots) compared to model predictions to reproduce the observed excess, which rules out the binary model. (Right) instead, a model that includes a single white dwarf (orange line) and a debris ring system (red dashed line) can fit both spectral and photometric data (from Debes et al. 2019)

The NIRES spectrum, which contains several weak Hydrogen lines, allowed our team to determine the temperature and surface gravity of the white dwarf, and in turn its mass (0.69±0.02 solar masses) and age (3.0±0.2 billion years). By combining all of the data together, our team was also able to generate a model for the disk, which requires more than one “ring” of material with a total mass greater than a typical asteroid or comet. Both of these features are surprising: structure in the ring suggests there may be another body clearing a gap in the disk, or perhaps there have been two tidal disruption events that happened sequentially. This dust should also be cleared our “relatively” quickly (“relatively” = few 100 million years), requiring a “relatively” recent disruption.

Overall, the properties of J0207+3331 suggest that planetary systems may be continuously dismantled for billions of years after a star dies, which gives us a lot more time to study the innards of planets after tidal dissection (yech!). Moreover, the discovery of such an interesting, and relatively nearby system (only 45 parsecs, or 150 light-years, away), means that there may be many more such systems out there. Plenty of opportunity for future citizen scientists like Melina Thévenot!

Here are some links to press reports on this result:






Backyard Worlds Blog, where Melina Thévenot describes her discovery:

Want to find your own new world? Give Backyard Worlds/Planet 9 a try!