arXiv:astro-ph/0607232v1 11 Jul 2006
ApJ. Draft: 09-Jul-2006
Resolving and probing the circumstellar disk of the Herbig Ae
star MWC 480 at λ=1.4 mm: Evolved dust?
Murad Hamidouche, Leslie W. Looney
Astronomy Department, University of Illinois, 1002 W Green st, Urbana, IL 61801
Lee G. Mundy
Astronomy Department, University of Maryland, College Park, MD 20742
We present high resolution 0.45′′×0.32′′observations from the BIMA array
toward the Herbig Ae system MWC 480 in the λ = 1.4 mm dust continuum. We
resolve a circumstellar disk of radius ∼170 AU and constrain the disk parameters
by comparing the observations to flat disk models. These results show that the
typical fit parameters of the disk, such as the mass, MD∼ 0.04-0.18 M⊙, and
the surface density power law index, p=0.5 or 1, are comparable to those of the
lower mass T Tauri stars. The dust in the MWC 480 disk can be modeled as
processed dust material (β ≈ 0.8), similar to the Herbig Ae star CQ Tau disk;
the fitted disk parameters are also consistent with less-evolved dust (β ≈ 1.2).
The possibility of grain growth in the MWC 480 circumstellar disk is supported
by the acceptable fits with β ≈ 0.8. The surface density power-law profiles of
p=0.5 and p=1 can be easily fit to the MWC 480 disk; however, a surface density
power-law profile similiar to the minimum mass solar nebula model p=1.5 is ruled
out at an 80% confidence level.
Subject headings: stars: individual: MWC 480 — stars: evolution, formation —
circumstellar matter — techniques: interferometric
Herbig AeBe stars (HAEBE) are young intermediate mass stars, ranging roughly from
2 to 20 M⊙ (12). They are the more massive counterparts of T Tauri (TT) stars and
– 2 –
are usually optically visible and highly variable. They have broad emission lines and dust
millimeter emission (e.g. Zuckermann 2001). Their pre-main-sequence evolution is more
difficult to study with as much detail as TT stars because their formation and evolution
processes are accelerated and more embedded; both lower mass Herbig Ae stars and the
more massive Herbig Be stars are thought to accrete disk material quickly or disperse the
disk by a stellar wind (e.g. Natta et al. 2000, N00 hereafter), which decreases disk lifetime.
Nonetheless, there is increasing observational support for the existence of circumstellar disks
around HAEBE stars (e.g. Fuente et al. 2003). This is interesting since the formation of
planetesimals and planets is expected to occur in these disks during the pre-main-sequence
evolution of the star.
The principal objective in this study is to utilize subarcsecond observations of the Herbig
Ae star MWC 480 circumstellar disk in the λ = 1.4 mm continuum, probing its structure and
morphology. MWC 480 (HD 31648) is a Herbig Ae star of spectral type A2/3ep and mass
of 2.2 M⊙at a distance of 140 pc (Th´ e et al. 1994, Eisner et al. 2004). We will compare our
results to previous observations of HAEBE stars and to their lower mass counterparts TT
stars. We use a simple Gaussian model to fit these high resolution observations and estimate
the disk parameters. Fitting a thin disk model with an inner hole to the observations allows
us to deduce more explicitly the disk parameters. Constraining the disk is difficult with the
large number of free parameters in the model. We fit the spectral energy distribution (SED)
to constrain roughly the disk model. Our disk model fits the shape of the broad SED. This
is a first step in our modeling strategy to reduce the number of free parameters. Thereafter,
we use the allowed parameter space of the SED fitting to model the resolved disk image and
better constrain the disk and best fit parameters.
2.Observations and data analysis
The observations were obtained with the Berkeley-Illinois-Maryland Association (BIMA)
array1, located at Hat Creek, California. On 2004 February 9, MWC 480 was observed at λ
= 1.4 mm while the BIMA array was operated in its extended configuration (A array). The
corresponding maximum baseline length was ∼ 1.2 km. System temperatures of the nine
antennas varied from 436 to 975 K, and the weather was very good. We observed with a
repeating cycle of 18.5 minutes on source and 3.5 minutes on the phase calibrator 0359+509.
1BIMA has since combined with the Owens Valley Radio Observatory millimeter interferometer, moved
to a new higher site, and was commissioned as the Combined Array for Research in Millimeter Astronomy
(CARMA) in 2006.
– 3 –
The amplitude calibrator MWC349 was observed at the beginning of the observations. Based
on recent measurements, the assumed flux density of MWC349 during that period was 1.67
Jy, and the corresponding flux of 0359+509 was 2.67 Jy. The total observing time was 8
hours. On 2003 November 27, MWC 480 was observed with the B-array of BIMA. We used
the same observing strategy and calibrators as in A-array. The assumed fluxes were respec-
tively 1.67 Jy for MWC349 and 3 Jy for 0359+509. The system temperature ranged from
312 to 1100 K.
The correlator was configured with a bandwidth per sideband of 625 MHz. The tracer
line13CO J=2→1 was observed simultaneously with a resolution of 25 MHz. For mapping,
we have combined A and B array data with robust = 0.1 weighting. The inclusion of B
array data allows sensitivity to structures from about 0.35′′to 4′′. The data were FFTed and
CLEANed using the MIRIAD package software (Sault et al. 1995). The images were made
separately for the lower and upper sidebands. We checked the consistency of the images
then combined them to obtain the final image. The final beam size was 0.45′′×0.32′′at PA
The continuum map at λ=1.4 mm of MWC 480 is shown in Figure 1. The emission arises
from circumstellar dust around the star. The image resolves the disk that was previously
detected by Mannings & Sargent (1997, MS97 hereafter). This is the first resolved image of a
dust disk around a Herbig Ae star at λ=1.4 mm. The total flux density of the circumstellar
disk is 210±15 mJy and the peak value is 51±3 mJy/beam at R.A.(2000)=04h58m46s.27
and Dec(2000)=+29◦50′36.95′′. The error bars in the measurement represent only statistical
uncertainties; we estimate an absolute flux calibration uncertainty of 20%. Comparison with
previous lower resolution interferometric observations at λ=1.4 mm shows a good agreement
with the flux density 279±7 mJy (Mannings et al. 1997). This suggests that all emission
is detected in our observations (cf. Natta et al. 1997). Lower mass TT stars show similar
fluxes at λ=1.4 mm, in the hundreds of milliJanskys (Beckwith et al. 1990).
As a first step, we have fitted a simple Gaussian model and deduced a major axis
∼0.8′′, minor axis ∼0.7′′, and PA = 143±5◦. Thus, the deduced source diameter (FWHM)
is ∼100 AU at the source distance of 140 pc. The disk is inclined by 37±3◦, which gives
the elliptical shape. Both the disk PA and inclination show a good agreement with previous
measurements: PA=148±1◦and i=38±1◦from the12CO J=2-1 emission (Simon et al. 2000).
In addition to the disk emission, our image shows extended emission, total flux density ∼ 30
mJy, to the south-west region of the disk ∼0.7′′long (∼100AU) at PA ∼ -160◦. This may be
– 4 –
due to a jet of free-free emission. Recent far-UV observations have detected a jet/counter-jet
at a similar PA (C. Grady, 2005 private communication).
At the resolution of Figure 1, we did not detect any13CO J=2→1 emission; presumably
we resolve out any disk emission, and we are only sensitive to line brightness of 65 K (1σ
level) at this resolution, with a channel width of 5.3 km/s.
sensitivity at large scales, we use the B array data with a beam-size of 1.5′′×1.3′′. In that
case, we have detected low-level13CO J=2→1 emission (∼ 4σ) that is offset by ≈ 1′′toward
the south-east. The flux density at the peak is 1.6±0.3 Jy/beam km/s within the velocity
range 3.1 and 8.4 km/s, bracketing the Vlsr of 5.6 km/s (Qi 2001). Qi (2001) detected
blue-shifted and red-shifted line components of13CO J=1→0. The13CO is important to
trace the outer regions of the disk. It is mostly detected in larger circumstellar structures
than our maximum detectable structures ∼ 560 AU. Simon et al. (2000) have reported the
detection of12CO J=2→1 with a diameter ∼1100 AU. Even at our lower resolution map, we
have still presumably resolved out much of the disk gas emission, and we are only sensitive
to line brightness of 4 K (1σ level), with a channel width of 5.3 km/s.
In order to determine our
Assuming optically thin dust emission at millimeter wavelengths with a constant tem-
perature, we can estimate the disk mass
where Fν is the flux, d the distance to the source, and κν=κ0(ν/1200µm)βthe dust grain
opacity (e.g. Beckwith & Sargent 1991; Pollack et al. 1994). We adopt β=1 (e.g. Ossenkopf
& Henning 1994) and opacity coefficient κ0= 0.1 cm2g−1(Hildebrand 1983). In the Planck
function Bν(TD), we assume a constant disk temperature as suggested by N00 for an A3
star, TD=23 K. We find a disk mass MD= 0.040 M⊙and a ratio of the disk to stellar mass
MD/M∗= 0.018. Both of these values are within the ranges found by N00 for Herbig Ae
Figure 2 shows the SED (Spectral Energy Distribution) of MWC 480, from the infrared
to millimeter wavelengths, including our observation at λ = 1.4 mm. The solid line is the
SED obtained with power-law disk model §4.1. A straight line fit to the measured dust
emission yields a millimeter spectral index α ≃ 2.48 (with Fλ∝ λ−α), within the wavelength
range λ=0.45-2.7 mm. This index value is consistent with the typical spectral indexes for
disks (Beckwith et al. 2000).
Free-free emission contribution to the long wavelengths region of our SED is negligible
(e.g. Rodmann et al. 2006). It is mostly dominant at centimeter wavelengths. Testi et al.
(2003) deduced a contribution of the free-free emission in the Herbig Ae star CQ Tau of
≪ 10% at 7 mm.
– 5 –
4. Disk Modeling
In order to deduce the disk physical characteristics, we compare our observations to
a flat disk model. We assume a flat, geometrically thin circumstellar disk with a hole in
the center of radius ri. The disk surface density is defined as Σ = Σ0(r/1 AU)−pand the
temperature T = T0(r/1 AU)−q, where r is the disk radius, T0and Σ0are respectively the
surface density and the temperature at the radius r = 1 AU (e.g. Looney et al. 2000).
Although this is a crude model, several studies have shown that circumstellar disks around
Herbig Ae stars can be adequately characterized by a temperature power law and a surface
density power-law (e.g. Testi et al. 2003). Note, we do not assume optically thin emission
for this model.
Our modeling strategy is to make two complementary simulations. First, we model the
SED in order to reduce the number of free parameters by fitting the shape of the broad
observed SED. Second, we use the deduced parameters from this SED modeling for the
simulation of the BIMA resolved disk image of MWC 480 at λ = 1.4 mm.
4.1. The Spectral Energy Distribution models
With the number of free parameters in the disk model, it is not possible to determine
a unique model from the λ=1.4 mm data alone. Thus, we fix some parameters a priori:
and PA=145◦, based on the results obtained in §3. In addition, fitting the SED
(Figure 2), that is the integrated flux at various wavelengths, will allow us to constrain the
inner radius ri, the temperature T0and its power-law index q, and the opacity index β, as
these parameters affect the SED’s shape and flux level. We leave the disk mass MDas a free
parameter to optimize in the fit. As part of our modeling strategy, we have fixed p at three
discrete values 0.5, 1 and 1.5. Table 1 summarizes the parameter space explored in our SED
Table 1. Range of variations of the disk model parameters.