Cold dust around nearby stars

The discovery of large reservoirs of cold dust around nearby stars in the 1980s provided strong evidence for the existence of planetary systems other than our own. If no planets exist in such systems then the density and brightness of the dust should increase smoothly with drag forces towards the central star. However, if planets do exist interplanetary dust will interact with these larger bodies whilst spiraling in towards the star, resulting in irregular variations in both the density and the corresponding brightness distribution. Although the edge-on disk around b Pictoris had been studied extensively, it was not until 1998 that significant new results emerged (Holland et al. 1998; Jayawardhana et al. 1998; Koerner et al. 1998; Greaves et al. 1998). In particular, submillimetre-wave techniques can circumvent the difficulties of the starlight that dominates at optical and near-IR wavelengths, by imaging the faint thermal emission from cold dust grains. The much hotter star has very little flux at these long wavelengths. The submillimetre images of Vega, Fomalhaut and e Eridani showed dust disks similar in size to the Sun's Kuiper Belt of comets (Holland et al. 1998; Greaves et al. 1998) and revealed evidence of central cavities in the emission near the star possibly cleared-out by the formation of planets.

The need for high resolution imaging?

Although intriguing, these observations lacked sufficient spatial resolution to investigate smaller scale structure, such as the asymmetries and warps that are potential signatures of planetary perturbations. New data at short-submillimetre wavelengths of Fomalhaut shows the presence of a "warp" in the observed dust torus. At 450m m, where the telescope beam-size is equivalent to a resolution of 50 AU at the distance of Fomalhaut, the dust disk appears to have a distinct bend in the connecting emission between the two offset peaks (see Figure 1). The image confirms that we are looking at an almost edge-on dust disk or ring, with a cavity devoid of emission out to approximately 100 AU radius from the star. The two bright peaks result from looking "down the ends" of the ring where the effective column density is largest. The new data clearly benefited from not only some of the best submillimetre weather on Mauna Kea, but are also substantially improved in both sensitivity and calibration accuracy over the previous work (due to the new wide-band filter on the short-wave array, and the calibration work of Archibald et al. paper in prep - also see JCMT webpage for details).

Several possible theories for the origin of the asymmetry are possible. Papers in preparation (Holland et al., Wyatt et al.) will show that no smooth ring model (e.g. pericentre glow) can fit the data, and the asymmetry is unlikely to be explicable by dust generation in planetesimal collisions. The most plausible model is that the dust is concentrated in orbital resonances with a planet. A comparison with recent numerical simulations shows that the data can indeed be qualitatively modeled with a Saturn-like object hidden in a gap in the torus. In figure 2 we show a simple de-projection of the 450m m image. This was first rotated for convenience so that the peaks lie in a vertical line, and then stretched by a factor of three along the horizontal axis. This is roughly equivalent to converting the image from an inclination of 19.5 degrees to face-on, although it does not take into account degradation by the finite beam size or a non-zero disk thickness (we are currently investigating ways of doing this?). The result shows a long arc below the star and a shorter one to the upper left. There is a small gap to the left of the star, and a much larger one covers the remainder of the ring.

A unique interpretation is certainly not possible at this stage, but we show for comparison (also in figure 2) a model planetary system from Ozernoy et al. (2000). Here there is a 0.3 M(Jupiter) (roughly Saturn-mass) planet orbiting at the ring radius, and dust is collected into librating orbits about the L4 and L5 Lagrangian points. There is a good resemblance between this model and the Fomalhaut de-projection, after taking into account the large beam size. In particular, the small and large gaps and short and long arcs are clearly present.

 

Figure 2. (left) De-projected (stretched) 450m m image of Fomalhaut with telescope beam, and (right) Ozernoy model.

This work is still on-going with data being collected on other systems such as Vega and e Eridani at a wavelength of 450m m. Until the ALMA interferometer comes on-line later this decade, the JCMT and SCUBA offer a unique way to study such systems with unparalleled sensitivity.
 

Acknowledgements

We thank all our collaborators who have contributed to this research: Bill Dent, Mark Wyatt (UKATC), Ben Zuckerman, Chris McCarthy (UCLA), Rich Webb (NASA Ames), Iain Coulson, Gerald Moriarty-Schieven, Ian Robson (JAC), Dolores Walther (Gemini), Walter Gear (Cardiff), Helen Walker (CCLRC) and Harold Butner (SMTO). This research was supported in part by PPARC funding, and by NSF and NASA grants to UCLA.

References

1. Holland et al.1998, Nature 392, 788

2. Jayawardhana et al.1998, ApJ 503, L79

3. Koerner et al. 1998, ApJ 503, L83

4. Greaves et al. 1998, ApJ 506, L133

5. Ozernoy et al. 2000, ApJ 537, L1

Click here for printable version.