2 Observations

2.1 Basic stellar properties

Table 1 reports the properties of the four stars. Column 1 gives the name of the star, Col. 2 the distance and Col. 3 the spectral type (see van den Ancker et al. 1999a; Rostopchina 1999 and references therein). Effective temperature (Col. 4), luminosity (Col. 5) and extinction (Col. 6) have been recomputed by us using the available photometry and standard methods. Column 7 shows for each star the maximum observed variability $\Delta V$. UX Ori, WW Vul and CQ Tau show large photometric variability and deep minima, associated with a large increase in the polarization (Grinin 1994 and references therein). AB Aur is photometrically much more stable , with variations of small amplitude ($\Delta V$$\sim$0.25 mag; Herbst & Shevchenko 1999). However, there is a report in the literature of a minimum of about 1 mag in 1997 (Kawabata et al. 1998; Ashok et al. 1999; van den Ancker et al. 1999b). It is possible that the rarity of deep photometric minima in AB Aur is due to the fact that we see its disk close to face-on (see following).

Column 8 shows the inclination of the disk with respect to the observer ( $\theta=90\hbox{$^\circ$ }$ for an edge-on disk). For WW Vul, UX Ori and CQ Tau, the inclination is inferred from the degree of polarization at minimum light (Natta & Whitney 2000). For AB Aur, Grady et al. (1999) estimate $\theta <45\hbox{$^\circ$ }$ from an HST image in scattered light, while Mannings & Sargent (1997) derive $\theta \sim 76\hbox{$^\circ$ }$ from the elongation of the ¹³CO (1-0) emission. The small inclination derived by Grady et al. is confirmed by recent interferometric observations at Plateau de Bure (Dutrey et al., private communication) and we will adopt in the following $\theta=30\hbox{$^\circ$ }$.

2.2 Millimeter interferometry

Observations of WWVul were made simultaneously at 2.9 mm and 1.2 mm using the Plateau de Bure interferometer (IRAM, Plateau de Bure - France) on October 31 and November 15, 1999. Visibilities were obtained in the most compact configuration of the 5 antenna array, yielding projected baselines which range from about 64 m down to the antenna diameter of 15 m. The 46'' (20'' at 1.2 mm) primary beam field of the interferometer was centered at $\alpha_{{\rm J}2000}=$19:25:58.75 and $\delta_{{\rm J}2000}=+$21:12:31.3.

At 1.2 mm, data were taken in double sideband mode with the receivers tuned to 240.0GHz (upper sideband). At 2.9 mm, observations were made in upper sideband only, with the SIS receivers tuned to 105.0GHz. The spectral correlators covered an effective bandwidth of 420MHz, equivalent to a velocity range of 520 kms^-1 at 1.2 mm (1200 kms^-1 at 2.9 mm).

Visibilities were obtained using on-source integration times of 20 min interspersed with 4 min calibration on 1923+210. The atmospheric phase noise on the most extended baselines ranged between 7$^\circ$ and 15$^\circ$ at 2.9 mm (17$^\circ$ and 34$^\circ$ at 1.2 mm), consistent with seeing conditions (0.9''-1.0'') typical for late-fall weather conditions. The absolute flux density scale which was established on the basis of cross-correlations on the continuum of the radio star MWC349 (1.06 Jy at 2.9 mm, 1.75 Jy at 1.2 mm), is in full agreement with the interferometric efficiency and should be accurate to 5% at 2.9 mm and to about 15% at 1.2 mm. The receiver passband shape was determined on 3C 454.3 and was measured better than 5% throughout the observations.

Data calibration was performed in the antenna-based manner. Flux densities of WWVul were obtained from the visibilities using standard IRAM fitting procedures. At 1.2 mm, we derived a one $\sigma$ continuum point source sensitivity limit of 1.1 mJy/beam corresponding to a rms brightness temperature of 3.1 mK, fully consistent with a total on-source integration time of 400 min and a mean system temperature of 350 K. At 2.9 mm, we obtained a continuum sensitivity limit of 0.3 mJy/beam, roughly equivalent to a rms brightness temperature of 1 mK.

A continuum source was detected at 1.2 mm and 2.9 mm almost at the position of the array's phase tracking center. The source was well-detected on all baselines and must be smaller than 0.9''. The source is likely to be point-like, the gaussian model fitted to the visibility profile at 1.2 mm being fully consistent with the signal-to-noise level, and the mean atmospheric seeing conditions.

2.3 ISO

The ISOPHOT observations of WW Vul were obtained April 12, 1997. The photometric measurements of WW Vul were part of an ISO programme aimed at studying the circumstellar environment of UXORs. The sequence contained a spectrophotometric measurement ( ${\rm TDT}=51300108$, observing mode PHT40), a background photometric measurement two arc minutes off target (51300109, PHT03), a photometric measurement (51300110, PHT03) and small maps at 150 and 200 $\mu$m (51300106, PHT22). The spectrophotometric measurement was done with 256s integration time in the wavelength ranges 2.5-4.8 and 5.8-11.6 $\mu$m with resolving power of about 90. The off measurement was done with 32 s integration time per filter at 3.6, 7.3, 12, 25, 60 and 100 $\mu$m. The apertures were 18 $^{\prime\prime}$ for the three shortest wavelengths, 52 $^{\prime\prime}$ for 25 $\mu$m and 120 $^{\prime\prime}$ for the two longest wavelengths. The on-target photometric measurement was done with the same filter, aperture and integration time set-up. The 150 and 200 $\mu$m maps were made with the array in spacecraft raster mode with steps of one pixel. This resulted in images with 92 $^{\prime\prime}$ pixels at 150 and 200 $\mu$m. The on-target pointings were centred on $\alpha_{\rm J2000}=19$:25:58.6 and $\delta_{\rm J2000}=21$:12:31 while the background measurement was at $\alpha_{\rm J2000}=19$:25:58.6 and $\delta_{\rm J2000}=21$:14:31.

The data were reduced with PIA (Gabriel et al. 1997). In the reductions, we followed the standard off-line processing steps with the following exceptions. For the 3.6-100 $\mu$m photometry we sub-divided the ramps into 8 parts in order to allow a better treatment of measurements where the response was drifting during the integration. For the spectrophotometry, we used also so-called dynamic calibration where every wavelength is calibrated individually against a standard star which has its flux at the corresponding wavelength close to that of WW Vul. In all cases, the statistical errors are much less than the quoted calibration accuracies of 10-20%.

The 150 and 200 $\mu$m maps do not show any clear detection of a point source. This is due to the fact that the background fluctuation is at least as strong as the flux of WW Vul at these wavelengths. Therefore, we could not extract flux values from these measurements.

The LWS observation of CQ Tau was obtained February 15, 1998 (82301814, LWS01). The observation covered the full LWS wavelength range from 43 to 196 $\mu$m. We used the automatically processed data after omitting detector SW1 which is known to have occasionally strange behaviour and which in the case of CQ Tau deviated from the other measurements.

 

 

Table 1: Stellar properties

(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Star	D	ST	$T_\star$	$L_\star$	AV	$\Delta V$	$\theta$
	(pc)		(K)	($L_\odot$)	(mag)	(mag)	(deg)
AB Aur	140	A0	9500	48	0.5	0.25	<45
CQ Tau	100	F2	7500	5	0.9	2.1	66
UX Ori	450	A3	8600	42	0.4	2.2	60
WW Vul	550	A3	8600	43	0.6	1.9	53

 

 

Table 2: Millimeter interferometry

(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)
Star	Position (J2000)		$\lambda_1$	S	$\lambda_2$	S	$\alpha_{\rm mm}$	FWHM	Mass	Telescope	Ref.
	$\alpha$	$\delta$	(mm)	(mJy)	(mm)	(mJy)		( $^{\prime\prime}$)	($M_\odot$)
AB Aur	04:52:34.3	+30:28:20.1	-	-	2.7	$10.6\pm0.4$	$3.1\pm0.3$$^\dagger$	< $2^{\dagger\dagger}$	0.02	OVRO	a
CQ Tau	05:32:54.13	+24:43:03.9	1.2	$151\pm 3$	3.4	$13.0\pm 1$	$2.6\pm 0.1$	1.0	0.03	PdB	b
UX Ori	05:04:30.00	-03:47:14.3	1.2	$19.8\pm 2.0$	2.6	$3.8\pm 0.4$	$2.1\pm0.2$	<0.5	0.05	PdB	c
WW Vul	19:25:58.74	+21:12:31.3	1.2	$9.1\pm 1$	2.9	$1.2\pm 0.3$	$2.3\pm 0.4$	<0.9	0.05	PdB	d

$^\dagger$ There are no 1.3 mm interferometric observations of AB Aur. $\alpha_{\rm mm}$ is computed using the single-dish flux $103\pm 18$ (Mannings 1994). $^{\dagger\dagger}$ from 2.6 mm
OVRO observations. References: (a) Mannings & Sargent (1997); (b) A. Dutrey, personal communication (c) Natta et al. (1999); (d) This paper.

2.4 10 micron spectrophotometry

Multi-epoch 10 $\mu$m  spectrophotometry was obtained for CQ Tau on 5 November 1997 UT, 6 November 1997 UT, and 25 September 1998 UT using the Hi-Efficiency Faint Object Grating Spectrometer (HIFOGS, Witteborn et al. 1995) at the Wyoming Infrared Observatory (WIRO) using conventional infrared observing techniques. The HIFOGS spectral range spans the 7.5-13.4 $\mu$m atmospheric window with 120 discrete Bi:Si detectors at an approximately constant resolution of $\Delta \lambda = 0.05$ $\mu$m per detector. HIFOGS used a 3 $^{\prime\prime}$ circular aperture and a 30 $^{\prime\prime}$ chop throw. The spectra of CQ Tau were flux calibrated with the bright IR standard star $\alpha$ Tau, using the well determined flux spectrum by Cohen et al. (1996). Differences in atmospheric transmission between CQ Tau and $\alpha$ Tau spectra were corrected using the ratio of telluric transmissions calculated by ATRAN (Lord 1993) for a total column of 4.5 mm of H₂O for WIRO. The method of telluric correction was confirmed by dividing the same standard star taken at several different air masses by its measurement at the lowest air mass; applying telluric corrections to these ratios yielded $1.00\pm 0.02$for the photometric nights 6 Nov. 97 UT and 25 Sep. 98 UT. Standard star measurements on 5 Nov. 97 UT showed large photometric uncertainties of up to 12%, and telluric corrections good to 2%-5% in the shape of H₂O (7.5-7.8 $\mu$m ), O₃ (9.4-9.8 $\mu$m ), and CO₂ (13.0-13.4 $\mu$m ) bands. The three nights of 10 $\mu$m spectra of CQ Tau that were obtained showed no significant variations in either spectral shape or flux level of the silicate resonance and mid-IR continuum to a level of the photometric accuracy of the observations. From the three night spectra of CQ Tau, a single statistically-weighted average spectrum was calculated with a resulting signal-to-noise greater than 20.

Multi-epoch HIFOGS 10 $\mu$m  spectrophotometry was also obtained for UX Ori on two observing runs at WIRO in October 1996 and November 1997. The photometric stability of the sky was not as good for UX Ori as for CQ Tau. The sky conditions during the October 1996 HIFOGS observing run were consistently better than during the November 1997. One night, 11 October 1996 UT was of high photometric stability, with standard star spectrophotometry consistent within 2%. The best night during the second run, 8 November 1997 UT, had spectrophotometric consistency at a level of about 8%. The spectra for a given run were scaled to the flux level measured on the best night, using the following flux scaling factors: 1.20 for 5 October 1996 UT, 1.05 for 6 October 1996 UT, 1.00 for 11 October 1996 UT, 1.12 for 12 October 1996 UT, 1.00 for 7 November 1997 UT, and 1.30 for 8 November 1997 UT. After scaling, a stastically-weighted average spectrum for each epoch was computed from all the nights, and degraded to $\Delta \lambda \approx 0.1$ $\mu$m per detector for October 1996 and to $\approx$0.2 $\mu$m per detector for November 1997. The resultant signal-to-noise ratios were 20-30 across the silicate feature, except for the ozone band (O₃ from 9.4-9.8 $\mu$m), and 5 or better at the short and long wavelength ends of the atmospheric window. In November 1997, the flux density of the silicate feature and mid-IR continuum (defined shortward of 8.2 $\mu$m and longward of 12.5 $\mu$m ) was multiplied by a factor of 1.11 to match the flux density of October 1996 spectra. The difference in fluxes between the two epochs is slightly larger than the photometric uncertainty.