We have reported in this paper spatial observations we have made at 12 m of the Pictori s dust disk. Our aim was to build a self consistent model of the disk, using constraints provided by our data and other observations available i.e. the IRAS measurements, the 10 m spectrum, the photometric measurements of Backman et al. at 10 and 20 m, the scattered flux in the visible range, and to a smaller extent, the sub-mm and mm fluxes. We first tried to make a model of the inner disk, i.e., inside a 100 AU radius, corresponding to the area we could probe with our 10 m images. We were guided by thr 10 m spectrum of Knacke et al., 1993, to choose a composition of the grains, i.e. core-mantled silicate particles, following Greenberg and Li results, instead of the popular Draine and Lee interstellar silicates. In order to get the density, we had to invert the integral equation relating the observed spatial flux at 12 m, having first built a thermal model of the dust emission. The density shows an inner clearing zone in accordance with our previous results (LP94) and an asymmetry in favour of the South-West extension. This could be due to the effects of a planet on an excentric orbit, sweeping out the dust and producing these asymmetries, but further simulations of the interaction between a planet and a dust disk have to be made to be able to give strong conclusions. In order to be able to check the coherence of our density with the other observables of the disk, we kept in a first attempt, the same composition of the grains and we extrapolated our density using a decreasing power-law in the outer disk (r 100 AU). We found that the compatibility with IRAS and 20 m data was only marginal, with a dramatic dependance of the fit to the size distribution index.
As it is likely that the coolest grains may be covered by an ice mantle, we constructed a second model in which the inner grains are covered with ice beyond the ice boundary (free parameter in our models). This time, the compatibility of our model with the observations was greater, with a lower dependance of the IRAS and 20 m predictions on the size distribution index, which makes us think that this model is much more accurate and physically acceptable than the single component model (in this model, ice is responsible for the high 60 m flux measured by IRAS). Furthermore, the two component model explains why we detect in our 10 m images a dust disk as far as 100 AU which implies a relatively high temperature of the grains (so they must have a low albedo in the visible and a low absorption in the far-infrared range), as opposed with an albedo around 0.7 in the outer disk, measured by Artymowicz et al., 1989. It reproduces also the break in the scattered flux at a distance around 100 AU, due to the iced component disappearing. It allows us to make some predictions concerning future observations of the disk i.e. for instance an infrared spectrum showing a intense peak of emission around 50 m produced by the ice mantles. We have also explained how to reconcile the questionss one could have about the strong asymmetries we see in the thermal flux, and not detected in the scattered flux, showing that the thermal flux and the scattered visible flux do not probe the same regions in the disk.
However many question remain open:
The two component model we have built to try to fit all the observables is probably not unique. We based our models on previous models of interstellar/cometary dust that recommend precise values of some parameters like the porosity or the amount of organic refractory relatively to the silicate material. These values may be different in the Pic system. For instance some interplanetary particles are very porous (up to a porosity of 0.99) but others are not. The quick destruction of ice by UV sputtering remains an obstacle to the validity of models that include particles containing ice. We have considered uniform size distribution throughout the disk, but it is likely that it changes, since the collision rates vary with the distance to the star. The preliminary attemps we made to fit the observables with a single grain composition (core-mantle silicate) but a non-uniform size distribution failed, but we have not investigated all the possibilities.
© European Southern Observatory (ESO) 1997
Online publication: April 6, 1998