Direct Imaging Surveys

Giovanni Battista Riccioli is usually credited with the discovery of the first binary star, resolving Z Ursae Majoris into its two wide components, Mizar and Alcor. However, the confirmation that close stellar pairs were physically-associated binary systems was actually a later by-product of the quest to measure stellar parallax. Absolute astrometry imposes severe observational requirements. A potential means of circumventing some of these obstacles was originally proposed by Galileo (1630) and restated, over a century later, by James Bradley (1747): target pairs of unequal-magnitude stars that lie in close proximity; if the two stars have similar luminosities, the fainter star lies at a larger distance, and the angular separation should exhibit an annual variation due to larger parallax of the brighter (nearer) star. The program had wait for large-scale implementation until the end of the eighteenth century, when William Herschel instituted an extensive program of double star observations with his 20-foot reflector. However, rather than resulting in the measurement of stellar parallaxes, this program revealed that many close pairs exhibited secular motions consistent with orbital motion (Herschel, 1803). Herschel was the first to refer to these physically associated stars as "binary systems".

Herschel's measurements were made by eye, and visual observations played a major role in binary-star astrometry until the mid-twentieth century. The advent of astronomical photography in the late-nineteenth century provided a means of probing to fainter magnitudes and lower flux ratio () systems. Photographic proper motions surveys, notably the Lowell survey (Giclas et al, 1958) and Luyten's surveys with the Palomar 48-inch Schmidt (Luyten, 1980; 1981), proved highly effective at identifying wide binaries, although halation rings around bright stars limited the sensitivity at moderate angular separations (A < 60 arcseconds).

The search for even lower mass, sub-stellar, companions to stars received new impetus in the early 1990's with the introduction of large-area digital detectors that were sensitive in the near-infrared (1-2 ¡m) part of the spectrum. The late 1990's and early 2000's saw the completion of the first deep, wide-field sky surveys: the Deep Near-Infrared Southern Sky Survey (DENIS; Epchtein et al., 1997) and the

2MASS (Skrutskie et al., 2006). Among the primary science goals for these surveys was the discovery of brown dwarfs4.

DENIS and 2MASS, along with the optical/far-red Sloan Digital Sky Survey (SDSS; Stoughton, 2002), account for the overwhelming majority of known brown dwarfs. A subset of known brown dwarfs (~5%; Gizis et al., 2002) are wide (A > 30 arcseconds) common proper motion companions of nearby main sequence stars. Brown dwarf secondaries can lie over a thousand astronomical units (AU) from their primaries, distances many times greater than the size of our own planetary system5. Such large separations also occur in binaries with low mass stellar companions; for example, the M5 dwarf, Proxima Centauri, lies more than 40,000 AU from a Cen AB.

Imaging brown dwarf companions at smaller separations, comparable to the < 30 AU planetary region in our own Solar System, is challenging. Even in the near-infrared, where brown dwarfs are at their brightest, they are still > 1000 times fainter than Sun-like stars. For the nearest stars, within 10 parsecs of the Sun, 30 AU spans 3 arcseconds, meaning that any sub-stellar companion within such an orbital separation is embedded in the seeing halo of its host star. Detecting extra-solar planets is even more challenging since their near-IR luminosity is another factor of 1000 fainter.

The existence of seeing haloes around stars is a direct consequence of Earth's turbulent atmosphere. As light from a star enters the atmosphere, it is refracted along its path by multiple pockets of air at slightly different temperatures and indices of refraction, leading to a smearing of the stellar image, typically ~ 1 arcsec-ond at good observing sites. However, recent developments in telescope technology and fast-processing algorithms have given astronomers an edge. A novel technique, aimed at real time correction of atmospheric turbulence has been implemented at many observatories. "Adaptive optics," or AO has been used in remote sensing applications by the U.S. Air Force since the 1970s, and found its way into astronomy in the early 1990s. The technique dramatically sharpens images blurred by the turbulent atmosphere, reducing the apparent angular size of a point source (i.e., stars, brown dwarfs) to the diffraction limit A/D of a telescope, where A is the observation wavelength and D is the telescope's diameter (Fig. 5.6). Additional scattered light suppression and contrast enhancement may be achieved with the use of a corona-graph: a specially fabricated opaque or partially transmissive mask that blocks light from the primary star to reveal fainter objects in its vicinity. The latter technique

4The interest in brown dwarfs transcended their importance as a link between the realms of stars and planets. Because of their intrinsic faintness and virtually unknown properties as a galactic population, brown dwarfs were prime candidates to solve the problem of the "missing mass," a.k.a. "dark matter," in the Universe. However, it quickly became clear that the rate at which brown dwarfs were being discovered in DENIS and 2MASS fell far short of what was needed to account for a significant fraction of the unseen 99% of the mass in the Universe.

5The radius of our planetary system, as set by the semi-major axis of the orbit of the outermost planet, Neptune, is 30 AU.

1"

1"

Fig. 5.6. Image of the binary star HD 18940 with the adaptive optics system on the Palomar 5 meter telescope. (a) The adaptive optics system is turned off. The binary is unresolved because of atmospheric turbulence. The scale bar indicates the approximate width of the seeing. (b) The adaptive optics system is turned on. The binary is clearly resolved with an angular separation of 0.167 arcseconds.

Fig. 5.6. Image of the binary star HD 18940 with the adaptive optics system on the Palomar 5 meter telescope. (a) The adaptive optics system is turned off. The binary is unresolved because of atmospheric turbulence. The scale bar indicates the approximate width of the seeing. (b) The adaptive optics system is turned on. The binary is clearly resolved with an angular separation of 0.167 arcseconds.

was invented by Bernard Lyot, who used a simple circular opaque spot to observe the solar corona in 1930.

The use of adaptive optics and a coronagraph led to the discovery of the first unambiguous brown dwarf Gl 229B (Nakajima et al, 1995) with the 1.5 meter telescope at Palomar Observatory. At an angular separation of 7.8 arcseconds from its stellar primary, Gl 229B is 13 magnitude fainter in the 0.8^m I-band and 10 magnitudess fainter at 2^m. Yet while the companion is easily discerned in the AO-corrected image (Fig. 5.7), it is lost in the glare of the primary in 2MASS seeing-limited images. Brown dwarf companions 10,000 times fainter than their host star can now be detected at separations as small as 0.5 arcseconds using AO on the largest (8 to 10 meters in diameter) ground-based telescopes. Comparable contrast at 1 arcsecond can be achieved with the Hubble Space Telescope orbiting above Earth's atmosphere.

Direct imaging allows the most complete characterization of the photospheres of sub-stellar companions, rendering the resolved secondary available for spectroscopic observation. Current observations can resolve separations as small as 5 AU for stars within 10 parsecs, reaching the outer boundaries of the range currently probed by radial velocity surveys for exoplanets. For the moment, image contrast limits potential detections to high mass (5-15 MJup) exoplanets; however, upgrades to existing systems and progress in coronagraphic techniques will push deeper into the planetary-mass realm, as discussed further in Sect. 5.5.

Fig. 5.7. Discovery (left) and follow-up (right) images of the brown dwarf Gl 229B, taken with an AO system on the Palomar 1.5 meter telescope, and with the Hubble Space Telescope, respectively. Both images are taken at a wavelength of 0.8 microns.

Was this article helpful?

0 0

Post a comment