My peer-reviewed publications
Below are brief descriptions of all of my peer-reviewed research papers. You can also click here to check them out in my ADS library
2020
|
"TOI-1259Ab - a gas giant with 2.6% deep transits and a bound white dwarf companion," Martin, D. V., El-Badry, K., Hodzic-Kunovac, V., et al., 2020, in prep (expected October 2020 submission)
TOI-1259Ab – a gas giant with 2.6% deep transits and a boundwhite dwarf companion |
At the TESS Ninja 2 "hackathon" hosted by the University of Sydney I worked with some folks to cross-match a GAIA-derived binary star catalogue from Kareem El-Badry with the TESS Object of Interest (TOI) list. We found one system with transiting body and a bound white dwarf companion. Only a dozen or so such systems are known, barely any of which contain a transiting planet. They are cool examples of a planet surviving the evolution of the other star through its red giant phase. At 2.6% deep transits though, this easily could have been an eclipsing binary, and so we took radial velocity measurements to confirm its planetary nature. With such deep transits, and being a "warm" Jupiter, this may be an interesting candidate for future JWST atmospheric characterisation.
|
|
"Searching for Small Circumbinary Planets I. The STANLEY Automated Algorithm and No New Planets in Existing Systems," Martin, D. V. & Fabrycky, D. C, 2020, ApJ (under review)
|
Finding transiting circumbinary planets is a challenge because they are transiting a moving target. This leads to large transit timing and duration variations, both of which inhibit standard planet detection algorithms. To overcome this, folks like Bill Welsh, Jerry Orosz and Veselin Kostov have well-trained eyes for finding them. That has been great for finding large planets with deep transits, but when it comes to small super-Earth and even earth-sized planets, you really need to be able to find shallow transits, which may need to be stacked coherently to build up a signal. This is why Dan Fabrycky and I created the STANLEY algorithm. Unlike the dog its named after, this code is indeed a good boy, and is sensitive down to Earth-sized planets under ideal conditions. This first paper is a demonstration of it in the context of the known planets. A second paper is on the way detailing a large application of the algorithm to the broader Kepler eclipsing binary catalogue. The code will also be made public on GitHub shortly.
|
|
Man I miss Kepler... That beautiful mission was the perfect circumbinary planet finder. Its long 4-year observing timespans turned out to be perfect for this type of planet which have to date only been found with 50+ day periods. TESS is a different challenge, with typically only one month observing campaigns. This paper explores the idea that sure, your planet may only pass the binary once, but it could transits TWICE (i.e. once per star). The transit timing and variable depths and durations will tell you a lot more than a single transit on a single star. We are now trying to implement this technique (not going to lie... it ain't easy!)
|
|
So back over 10 years ago, well before we found a transiting planet around it with TESS, we observed the Rossiter-McLaughlin effect of TOI-1338/EBLM J0608-59 with the CORALIE spectrograph. This was only the second such publication of this effect. The binary is aligned, perhaps the expected and boring result, but hey... so were the first few measurements of the R-M effect for hot-Jupiters, and we all know what happened there...
|
|
Discovery of the first circumbinary planet with TESS. It follows pretty much all of the trends found in the Kepler discoveries. Excitingly, this binary was discovered by our EBLM survey, over ten years prior. Furthermore, it was being followed as a part of the BEBOP radial velocity survey for circumbinary planets. This meant that we had dozens of high precision HARPS and CORALIE spectroscopic measurements to add to the TESS data. With new observations coming from the ESPRESSO spectrograph on the VLT we will have arguably the best-constrained circumbinary planet ever. Another cool aspect: the transits were first spotted by a 17 year old during his NASA internship working with Veselin.
|
2019
|
Building on EBLM V and III a further five M-dwarfs in eclipsing binaries have their mass and radius precisely characterised. Note that even though all of our M-dwarfs were discovered photometrically using the WASP telescope, the WASP light curves are not sufficient for accurate radius determination. This is why in EBLM III, V and VI we had to obtain better photometry with EulerCam, TRAPPIST and SAAO. More precisely constrained binaries are on the way with K2 and TESS.
|
|
The Geneva exoplanet group has been using the CORALIE spectrograph to search for extra-solar planets for about 20 years. This long baseline, along with a super precision for a 1.2 metre telescope of a few metres per second, allows the detection of long-period gas giant circumbinary planets. These are crucial in our understanding of the formation of extra-solar systems.
|
|
Stellar structure models require precise masses and and radii. For sunlike stars the data are numerous, but more is needed for small stars like M-dwarfs. Beyond stellar physics, this has implications for surveys of exoplanets around M-dwarfs, as our knowledge of the planet dependent on our knowledge of the star. One of the goals of the EBLM survey is to provide well-characterised M-dwarfs from our binaries, and in this paper we present 10 such stars.
|
|
The mass ratio of a close binary is one of the markers of their formation. Some theories suggest tight binaries should be pushed towards equal masses, owing to preferential accretion to the secondary star. Now we have a new observable to add to this: circumbinary planets. I investigate the mass ratios of known circumbinary hosts and compare them to general surveys of binaries. In particular, I investigate how our transit observations are sensitive to the binary mass ratio (short story: not very much).
|
|
Radial velocities were the first technique used to find exoplanets around sunlike stars, and hundreds have been found since. However, no circumbinary planets have been found around RVs. We are trying to fix this with the BEBOP survey. To make things easier, we only target single-lined binaries, so we can avoid the problem of two sets of spectral lines. The CORALIE survey was our first attempt. We were sensitive to planets but unfortunately did not find any. It seems like they are typically Saturn-mass or smaller, but we have since moved to bigger telescopes and more precise instruments, receiving 88 nights on HARPS @ ESO, Chile and 240 nights on SOPHIE @ OHP, France.
|
|
Alex Teachey and David Kipping recently put forward the strongest evidence to date for an extra-solar moon: Kepler 1625b-i. It's a pretty weird moon, about the size of Neptune, but exoplanets are weird so why not moons too? I'm pretty agnostic to the existence of the moon, I reckon more observations are needed. This motivated us to write this paper. We show how misaligned moons, such as the nominal (but poorly constrained) orbit of Kepler 1625b-i, only transits ~40% of the time. This moon can next transit in May 2019, and our conclusion is don't be surprised if it doesn't transit, and don't use a non-transit to try to kill it.
|
2018
|
In this review article I summarise observational and theoretical efforts to date regarding planets in multiple star systems. This includes both circumbinary planets and circumstellar planets in wider binaries. My chapter is available freely on arXiv, or alternatively anyone with a spare $1099 could buy it on Amazon (but you get free shipping, woohoo!)
|
2017
|
The WASP planet survey is looking for transiting hot Jupiters - giant planets orbiting very close to their host star. Hot Jupiters are similar in size to M dwarf stars, and hence from transits alone it is hard to know the difference. Radial velocity follow-up breaks the degeneracy by measuring the mass, which is very different in the two cases. Many people would throw away the M dwarf's as false positives but Amaury Triaud used them to form the EBLM program. This is a wonderful comparison sample with the hot Jupiters, and also have implications on the formation and evolution of close binaries and tidal interactions. This is the first major data release of spectroscopic orbits, which will be roughly doubled next year.
|
|
In the EBLM program one target was particularly sexy: EBLM J0555-57. It has a mass 85 times that of Jupiter, which is similar to the TRAPPIST-1 host star, but a radius only as big as Jupiter. It is only just above the hydrogen burning limit. It is one of the densest non-stellar remnants known and the smallest star ever measured (or, as Fox News called it, the "wimpiest").
|
|
It turns out that a lot of the ideas regarding the transit probability of circumbinary planets can also be applied to circumstellar planets in binaries, where the planet only orbits one of the two stars. If the binary is close, say an AU apart, then the planetary orbit will precess on a timescale that is relevant to long surveys like Kepler. This makes its transit signature come and go and probably harder to detect, which may explain why none have been found so far. I then applied the same theory to exomoons, which also have never been observed, and showed them to exhibit similar characteristics.
|
|
Because circumbinary planets transit a moving binary target, and move through orbits that are themselves varying over time, your chance of observing a transit becomes inherently time-dependent. In this second paper on circumbinary transit probabilities I calculate when exactly the planet and binary orbits overlap, and apply this to re-observations of known planets. From this I calculated that Kepler has missed perhaps up to 75% of circumbinary planets that will later transit. This was picked up on by the New Scientist and a French science podcast.
|
2016
|
Kozai-Lidov cycles are synonymous with misaligned triple systems, but what about misaligned circumbinary planets? It turns out that the Kozai-Lidov effect is not applicable to circumbinary planets because it requires the outer orbit to have a lot of angular momentum, and circumbinary planets are just too small. Amaury and I had suspect this for a while but finally published clear proof of it, as well as some interesting intermediate cases where the outer orbit has similar angular momentum to the inner.
|
2015
|
All of the circumbinary planets had been found around eclipsing binaries, and most eclipsing binaries have very tight periods, less than a few days. It therefore seemed surprising that only wider binaries (periods greater than seven days) were known to host planets. With Tsevi Mazeh and Daniel Fabrycky we showed that this dearth of planets around the tightest binaries was compatible with a popular theory of tight binary formation by Kozai-Lidov cycles under the influence of a distant stellar companion. It was nice to use an observable (or in this case, non-observable) from exoplanets to advance our understanding of star formation.
|
|
As early as my honours thesis in 2012, Amaury Triaud and I suspected that circumbinary planets were more likely to planets than planets around single stars. In this paper I sat down, tried to remember my undergraduate mathematics training and derived the analytic probability of circumbinary planets transiting, proving our suspicions. Amaury and I interpreted this to show the benefit of circumbinary planets and argued their importance in the broader astrophysical context.
|
|
Transits generally probe circumbinary planets that are close to their host star. Astrometry, in particular that with high precision of the recently-launched GAIA satellite, can probe circumbinary planets on wider orbits. With Johannes Sahlmann and Amaury Triaud we predicted up to hundreds of future discoveries of wide circumbinary gas giants and circumbinary brown dwarfs, discoveries that will no doubt revolutionise the field.
|
2014
|
All planets had been discovered around eclipsing binaries, but most binaries do not eclipse. With Amaury Triaud we posed a very simple question: can you find planets transiting non-eclipsing binaries? The answer is yes, if the planet and binary orbits are misaligned. We predicted many should be detectable within the Kepler data. We also analysed the observed trends of circumbinary planets and showed which ones are meaningful and posed explanations for them.
|
|
Using the earlier results of the 2013 paper on circumbinary transit timing variations, Dave Armstrong led a study to show that the abundance of circumbinary gas giants was surprisingly calculated to be at least as abundant as gas giants around single stars. I synthesised various distributions of circumbinary planets and simulated their transit timings. This tested the detection capabilities of the algorithm. A prediction of our work was that if many misaligned circumbinary planets exist that have presently elluded detection, then the underlying abundance would be even higher than that around single stars. A tantilising possibility indeed...
|
2013
|
Dave Armstrong was already working to write an algorithm to automatically detect and constrain the abundance of transiting circumbinary planets. The transits are more complex than those around single stars because the stars are themselves moving and the planet's orbit is not static. Together we derived analytic limits on the transit timing variations.
|
