Brown Dwarfs and Extrasolar Planets

We are involved in a wide variety of programmes to detect and characterise extrasolar planets and brown dwarfs. We are one of only a few groups worldwide to discover nearby extrasolar planets and over the last decade have found around 40 extrasolar planets from the radial velocities of nearby stars and from imaging and spectroscopy in Orion. We have also made substantial contributions to the related and similarly flourishing area of brown dwarfs by discovering, characterising and modelling nearby and cluster brown dwarfs by a variety of techniques. We are pioneering a number of new extrasolar planet and brown dwarf projects: (1) using our own polarimeter to observe the reflected light from extrasolar planets, (2) an infrared extrasolar planet transit survey recently awarded 200 UKIRT nights (3) a new infrared radial velocity spectrograph for the Gemini Observatory (PRVS), (4) the detection and follow-up of brown dwarfs from the UKIDSS LAS survey, (5) the UKIDSS and VISTA galactic plane surveys. We have a variety of scientific objectives though long-term ones are the discovery and characterisation of other Solar Systems relative to our own and in particular the detection of extrasolar 'Earths' within the habitable zones of their parent stars as well as the determination of the low-mass mass function and its environmental dependence.

More details for our main research programmes are given below (ordered approximately in ascending mass).

PRVS - Searching for Earth-mass Planets

(Jones, Barnes, Jenkins)

The Precision Radial Velocity Spectrometer (PRVS) is key to answering the question: How common are extra-solar planets, including Earth-like planets?. Of the over 200 extra-solar planets discovered to date using a variety of techniques, the vast majority have been discovered using Doppler shift studies measuring precision radial velocities. By being the first such instrument to operate in the infrared PRVS can target low-mass stars and exploit the mass-sensitivity of the Doppler technique. In particular, PRVS will be sensitive to terrestrial-mass planets orbiting in the habitable zones of nearby stars. PRVS is a 1.0-1.8 micron 70k resolution spectrometer. It is designed to deliver short- and long-term stability of better than 1 m/s.

CAD/CAM impression of PRVS courtesy of David Montgomery, UKATC

Hunting for "extrasolar earths" via the near infrared transit method

(Pinfield, Jones, Lucas, Burningham, Hough)

We are carrying out a new large scale survey that will probe for habitable rocky planets around cool stars, as well as extrasolar giant planets. This survey employs the so called transit method - by monitoring the brightness of large numbers of stars and searching for the characteristic periodic decrease in brightness as a planet passes in front of (or transits) its host star.

Previous transit searches have been made in the optical, and have identified about 10 close in gas giant planets transiting solar type stars. These planets are known as "hot Jupiters", since they are strongly irradiated by the star. By using the new Wide Field Infrared camera on the UK Infrared Telescope, our survey is far more sensitive to planets transiting cool stars (M dwarfs) which are brightest in the near infrared. M dwarfs are 5-10 times smaller than solar type stars, and therefore smaller rocky planets can produce detectable transits. Also, the habitable zone (where liquid water may flow on an orbiting planet) around cool stars is much closer in than around solar type stars, and cool dwarf transits could be habitable rocky worlds.

Direct spectroscopic detection of extrasolar planets

(Jones, Barnes, Pinfield, Jenkins)

Direct imaging techniques will have great difficulty detecting such planets, because the planet is so much fainter than the star and because their two images are never separated on the sky by more than 0.003 seconds of arc. Fortunately, however, the starlight scattered from the planet can be distinguished from the direct starlight because the scattered light is Doppler shifted by virtue of the close-in planet's relatively fast orbital velocity (~ 150 km/sec). Superimposed on the pattern given by the planet's albedo changing slowly with wavelength, the spectrum of the planet's light will retain the same pattern of photospheric absorption lines as in the direct starlight.

Except in the unlikely case of a nearly face-on orbit, the orbital Doppler shift should cleanly separate the planet's spectral lines from those of the star. The essence of the technique is illustrated in the animation below.

Click the picture to play the animation. The light reflected from the planet contains the same pattern of dark absorption lines as the star. In the graph at top, the absorption lines are shown as dips.

The planet is at its brightest when it is on the far side of the star with its illuminated face turned toward Earth. As the planet passes on the far side of the star its spectrum appears to move from right (red) to left (blue) because of the Doppler effect. Eventually it reverses direction as the planet passes on the near side of the star, but by this time its dark side is turned towards us and the reflection signature is faint.

In this animation, the strength of the planet's spectrum is exaggerated by a factor of 3000. Set your movie player to loop indefinitely for best results. (Schematic animation courtesy of Andrew Collier-Cameron, University of St Andrews)

Radial Velocity Planet Searches

(Jones, O'Toole, Barnes, Galvez-Ortiz, Jenkins, Pinfield)

We are engaged in a number of radial velocity projects, in particular, the Anglo-Austrailian Planet Search is targeting 250 nearby stars brighter than V=7.5 in the Southern Hemisphere. A Jupiter-like planet exerts a small gravitational pull on its parent star, causing the star to wobble with a velocity of 1 to 100 meters per second depending on the orbital distance and mass of the planet. This motion can be detected via the Doppler Effect. The light emitted by a star moving toward the Earth will be Doppler shifted to shorter (bluer) wavelengths, while a star receding from the Earth will emit light shifted to longer (redder) wavelengths. The effect is extremely subtle and has no effect on the apparent colour of the star. A star with a Jupiter-mass planet will be revealed by the periodic Doppler shift of its light. After one or two orbital periods the information from the Doppler measurements allows us to calculate the orbit and the mass of the unseen planet. Our current measurement precision is 3 meters per second (a brisk walk). For comparison, Jupiter causes the Sun to wobble with a velocity of 12.5 meters per second over a 12 year period. Saturn induces a 2.7 meter per second wobble on the Sun with a 30 year period. The other planets are too small to produce a measurable effect on the Sun.

The signal of an extremely eccentric extrasolar planet detected by Anglo-Australian Planet Search (Jones et al. 2006, MNRAS, 369, 249)

Imaging planets and substellar companions

(Galvez-Ortiz, Pinfield, Clarke, Jones, Dolling, Zhang, Burningham, Folkes, Lucas, Hough)

New cutting edge adaptive optics techniques offer the potential to directly image exoplanets, despite the huge star/planet brightness ratio. There already exist a small number of examples of possible imaged exoplanets, however the exact nature of these detected objects remains controvertial. This project is identifying optimal targets for AO imaging searches (e.g. using NACO SDI on VLT, HiCIAO on Subaru, NICI on Gemini, and future facilities like JWST and ELTs). These optimal targets are young nearby cool stars (late M and L dwarfs) in kinematically distinct "moving groups". Their proximity, youth, intrinsic faintness and well constrained age and composition, make them optimal targets for AO exoplanet searches, if indeed they host such companions. The new sample should allow Jupiter mass exoplanets to be detected at 1AU separation from their host stars. The programme is making use of substantial amounts of time on large telescopes (VLT, Gemini, E2.2), and is currently supported by a Marie Curie Fellow (Galvez).

Direct Observation of Extrasolar Planets with PLANETPOL

(Lucas, Hough)

Following the detection of extrasolar planets by the radial velocity method in the mid-1990s a global effort began to try and detect these planets directly. The radial velocity method pioneered by Mayor & Queloz and Marcy & Butler provides a great deal of information about planetary orbits, including the fact that many planets orbit very close to the parent star (which are known as ``hot Jupiters''). However this indirect technique leaves an uncertainty in the mass of the planets and their orbital inclination, measuring only M.sin(i). It also provides no information about the composition of the planets themselves.

A number of groups are attempting to remedy this by directly observing extrasolar planets using high resolution interferometric imaging to pick out faint companions, or high resolution spectroscopy to distinguish reflected light from the planet using the wavelength shift caused by the Doppler effect.

PLANETPOL is our attempt to directly detect hot Jupiter planets via the polarisation of their reflected light, see Hough, Lucas, Bailey & Tamura; 2003, SPIE 4843, 517. PLANETPOL is an optical aperture polarimeter funded by PPARC, which was successfully commissioned at the University of Hawaii 2.2-m telescope in October 2003. Since the planet is spatially unresolved from the star, the polarisation fraction from hot Jupiter systems is only a few parts in a million (see Seager, Whitney & Sasselov 2000, ApJ 540, 504). When PLANETPOL is mounted at the Cassegrain focus of a 4-8m telescope it receives sufficient photons from a V=5 star to measure such a signal in less than 1 hour. A number of larger polarisation signals are also detected: eg. telescope polarisation, instrumental polarisation, interstellar polarisation and sky polarisation. However, only the signal from the planet will rotate at the planet's orbital period, which is known from radial velocity measurements. PLANETPOL has a sky channel to measure sky polarisation so that this can be subtracted if necessary (it is insignificant on Dark time). The object and sky channels are both dual beam systems, using 3-wedge Wollaston prisms to send the light to four separate thermoelectrically cooled Avalanche Photodiode detectors. The modulating element in each channel is a Photoelastic Modulator or PEM, which provides a retardance oscillating sinusoidally from zero to Pi at 20 kHz. The polarised signal is recovered from the first harmonic at 40 kHz using lock-in amplifiers. These PEMs provide the cleanest polarisation signal of any type of polarimeter, providing an instantaneous measurement of the I and Q or I and U components of the Stokes vector with each detector. The dual beam nature of the system merely serves to double the throughput.

Owing to the differential nature of PEM-based detection and the use of large apertures the instrument can operate in non-photometric conditions and in poor seeing.

PLANETPOL has the potential to:

  1. Measure sin(i) and hence the planetary mass, being sensitive at any orbital inclination (i).
  2. Determine planetary albedo and radius. Detection at BVRI will provide good precision for these observables.
  3. Determine the optical properties, size and perhaps composition of the reflecting particles in the planet's atmosphere.

Probing new temperature regimes between stars and planets

(Pinfield, Burningham, Jones, Lucas, Pavlenko, Zhang, Murray)

Our group is heavily involved with the ``UKIRT Infrared Deep Sky Survey'' (UKIDSS). This is a major new near infrared survey covering 20% of the sky. The UKIDSS Large Area Survey (LAS) will cover 4000 square degrees, and probe several magnitudes deeper than the 2-Micron All Sky Survey, searching an unprecedented volume for brown dwarfs.

The LAS will discover large numbers of L dwarfs (2300-1300 K) and T dwarfs (1300-700 K). However, it could also discover about 35 objects even fainter and cooler than the known L and T classes. For temperatures below 700 K atmospheric ammonia and water clouds are expected, and any resulting changes in spectral character would warrant a new spectral class, pre-emptively called the Y dwarfs.

These new record breaking Y dwarfs could include brown dwarfs that formed when the galaxy was young, as well as brown dwarfs with masses in the planetary regime - so called free-floating planets. We are carrying out follow-up programs (8m spectroscopy, photometry and astrometry) to confirm and study UKIDSS Y dwarfs, and reveal the nature of these objects.

The J-H, Y-J 2-colour diagram for sources we expect in the LAS. The location of A-M (III-V) stars, late M, L and T dwarfs and z=2-7 QSOs have been determined using colours synthesized from available spectroscopy. And expected Y dwarf colours have been synthesized from the latest theoretical model spectra by the Lyon group (priv. comm.), the Marley group (priv. comm.; see Marley et al 2002, ApJ, 568, 335), the Burrows group (Burrows et al., 2003, ApJ, 596, 587) and Tsuji (private comm.). We note that we have adjusted the Lyon J-H colours by -0.3 so that they pass through the known T dwarfs (Pinfield et al. 2007 in prep).

Understanding brown dwarfs via "benchmark objects"

(Pinfield, Burningham, Jones, Napiwotzki, Lucas, Galvez-Ortiz, Folkes, Day-Jones, Zhang, Clarke, Pavlenko, Jenkins)

Brown dwarf atmospheres are very complex. With their relatively low temperatures, numerous molecular species appear and atmospheric dust grains condense out. It is very difficult to accurately model such atmospheres, and it is thus an ongoing challenge to observationally constrain the physical properties of brown dwarfs - a task that is vital if we are to measure the disk IMF and formation history.

The group is carrying out a number of programs to identify brown dwarfs whose physical properties may be inferred independently of spectral analysis - so called "benchmark brown dwarfs". By finding brown dwarfs with known age, composition and distance, it becomes possible to determine their temperature, surface gravity and metallicity without reliance on atmospheric models. These benchmarks thus become observational reference points in temperature-gravity-metallicity space.

High resolution spectrum and model (dotted) of the archetypal late-type dwarf GJ752B. (Jones et al., 2005, MNRAS, 358, 105 )

We are specifically searching for benchmarks brown dwarfs as

Brown dwarfs in star formation regions

(Lucas, Pinfield, Weights, Burningham)

This group was responsible for the first detection of free floating planetary mass objects (IPMOs) in the Trapezium Cluster in Orion, in collaboration with Pat Roche of Oxford University. (Publications: Lucas & Roche 2000 MNRAS 314, 858; Lucas, Roche, Allard & Hauschildt 2001, 326, 695). Similar objects were discovered in the Sigma Orionis cluster at the same time (Zapatero-Osorio et al. 2000) and similar planetary mass candidates have since been found in several nearby star formation regions.

Our research in this area is continuing, using the twin Gemini telescopes to search for less massive IPMOs in Orion and other star formation regions. The main aim of this research is to further our understanding of star formation and establish whether there is a minimum mass for objects forming via the star formation process. We are also engaged in further follow-up spectroscopy in order to better determine the temperatures, ages, masses and surface gravities of the IPMOs.

Studies in regions where star formation is still ongoing have the disadvantage that source masses cannot be determined with high precision. However this is offset by the benefits of observing the star and brown dwarf formation process in action. Features such as the Initial Mass Function (IMF), the velocity dispersion of brown dwarfs, brown dwarf accretion disks and the sub-clustering of these sources all help us to piece together the star formation process.

Nearby and distant brown dwarf searches

(Jones, Pinfield, Lucas, Galvez-Ortiz, Barnes, Zhang, Folkes, Weights, Day-Jones, Clarke)

The group has several programs to search for and study very nearby brown dwarfs. Such objects are amongst the brightest (apparent) in their spectral class, and thus provide ideal targets for detailed study including

We have also begun a program to search for distant brown dwarfs within the galaxy, to investigate if the sub-stellar IMF varies spatially within the galactic disk. The challenge here is to identify L dwarfs in the galactic plane, which is difficult because there are many sources of photometric contamination that share L dwarf colour. But by combining optical and near infrared photometry with astrometry and morphological analysis, we are able to select optimal L dwarf samples in the plane. We plan to use these techniques to identify L dwarfs out to distances of about 500pc in the UKIDSS Galactic Plane Survey.

CAR home