A 12CO () and () atlas of circumstellar envelopes of AGB and post-AGB stars

Astronomy and Astrophysics Supplement Series, Jul 2018

We present the results of a 12CO () and () survey on a sample of 46 objects classified as AGB and post-AGB stars. We have obtained fully sampled high resolution maps of their 12CO () emission by combining visibilities from the IRAM interferometer with short spacing observations from the IRAM 30 m telescope. Properties of their circumstellar envelopes like fluxes, radii, and positions are derived from model fits to the visibilities and compared to the results obtained from 12CO () maps observed at the IRAM 30 m telescope. From the 12CO () observations we have derived mass loss rates for 38 stars and established an empirical relation between the CO photodissociation radius of an envelope and the measured radius in the () emission.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:


A 12CO () and () atlas of circumstellar envelopes of AGB and post-AGB stars

A 12CO (J = 1 ! 0) and (J = 2 ! 1) atlas of circumstellar envelopes of AGB and post-AGB stars R. Neri 2 C. Kahane 1 R. Lucas 2 V. Bujarrabal 0 C. Loup 3 0 Observatorio Astronomico Nacional (I.G.N) , Apartado 1143, 28800 Alcala de Henares, Madrid , Spain 1 Observatoire de Grenoble , BP. 53, 38041 Grenoble Cedex 9 , France 2 IRAM , 300 rue de la Piscine, 38406 S 3 Institut d'Astrophysique de Paris , CNRS, 98 bis Bd. Arago, 75014 Paris , France We present the results of a 12CO (J = 1 ! 0) and (J = 2 ! 1) survey on a sample of 46 objects classied as AGB and post-AGB stars. We have obtained fully sampled high resolution maps of their 12CO (J = 1 ! 0) emission by combining visibilities from the IRAM interferometer with short spacing observations from the IRAM 30 m telescope. Properties of their circumstellar envelopes like fluxes, radii, and positions are derived from model ts to the visibilities and compared to the results obtained from 12CO (J = 2 ! 1) maps observed at the IRAM 30 m telescope. From the 12CO (J = 1 ! 0) observations we have derived mass loss rates for 38 stars and established an empirical relation between the CO photodissociation radius of an envelope and the measured radius in the (J = 1 ! 0) emission. Atlases | stars; AGB; post-AGB | stars; circumstellar matter | stars; mass-loss | stars; evolution | radio lines; stars 1. Introduction The ultimate stage of the red giant evolution, namely their transition from the Asymptotic Giant Branch (AGB) to the Planetary Nebulae (PNe), and the mechanisms and chronology of events driving them to expell the bulk of their mass in less than a million years (in some cases 10000 years), constitutes one of the less known, albeit most exciting topics in stellar evolution. The whole evolution of the morphological, dynamical and chemical layout of envelopes from the AGB to the PNe stage through the state of a proto-planetary nebula (PPN), strongly depends on the massive ejection of stellar material during this stage. Send o print requests to: R. Neri Such a copious mass loss process necessarily leads to substantial changes in the stellar structure and appearance especially when the amount of mass ejecta becomes comparable to the mass of the star. Since the dominant process which brings back recycled stellar material to the interstellar medium is thought to be the formation of PNe, the mass loss process also greatly a ects the composition of the interstellar gas and dust, its enrichment in heavy elements and the interstellar isotopic ratios. During their high mass loss phase, the AGB stars are surrounded by dust and gas expanding envelopes which, if thick enough, are readily detectable in molecular radio line emission. In particular, high resolution observations of the rotational transitions of 12CO, in the millimetric and submillimetric ranges, yield unique data on the envelope geometry, mass loss rate and dynamics. Since the middle of the eighties, a large number of surveys, using the largest available single dish radiotelescopes, have been devoted to searches for 12CO (J = 1 ! 0) and (J = 2 ! 1) emission in the circumstellar envelopes of evolved stars, on the basis of various selection criteria. See for instance the pioneering work of Knapp & Morris (1985), the compilation of 12CO and HCN data by Loup et al. (1993), and, among the recent large surveys, that of Olofsson et al. (1993). The rst targets of these studies were the brightest infrared sources or less opaque envelopes around well known evolved stars. Since 1988, the IRAS colour-colour diagram (van der Veen & Habing 1988) has been extensively used to increase the sample of dusty AGB stars, based on their characteristic infrared colours, it has also proved to be a powerful tool to select various classes of objects. Surveys based on CO observations at higher frequencies are still rare (Young 1995; Stanek et al. 1995). Presently, the number of circumstellar envelopes detected in CO radio emission around evolved stars (AGB, proto- and Planetary Nebulae) is larger than 700. Due to the limited sensitivity and spatial resolution of radiotelescopes, the maps of CO emission around evolved stars are much less numerous (around 50) and essentially devoted to the most evolved objects (PPN or PN), which have generally been selected for some peculiar morphology (bipolar, ring like, . . . ). Since the rst maps, obtained at the end of the eighties (see e.g. Healy & Huggins 1988), most observations have been done with single dish telescopes. Recently several selected objects have been extensively mapped by interferometric techniques (see for instance Bujarrabal et al. 1997). To our knowledge, atlases of CO maps, obtained and analyzed in uniform conditions, around relatively large samples of evolved stars, are however almost inexistent. The pioneering work of Bujarrabal & Alcolea (1991) presents 12CO (J = 1 ! 0) and (J = 2 ! 1) mapping, with the IRAM 30 m telescope, of the circumstellar gas around 10 evolved stars of various chemical and variability types. More recently, the envelopes of 18 evolved stars, ranging from AGB to PN and showing various chemical types, have been mapped with the Caltech Submillimeter Observatory in 12CO (J = 3 ! 2) (Stanek et al. 1995). We present here an atlas of 12CO (J = 1 ! 0) and 12CO (J = 2 ! 1) maps of a sample of 46 AGB and postAGB stars based on IRAM 30 m telescope and interferometer observations. The 12CO (J = 1 ! 0) data come both from the interferometer and the 30 m telescope and are combined to circumvent the missing short-spacing problem. To get some additional information on CO excitation in the envelopes, we have also obtained fully sampled 12CO (J = 2 ! 1) maps with the 30 m telescope. The stars sample is presented in Sect. 2, the 30 m telescope observations and 12CO (J = 2 ! 1) data analysis in Sects. 3 and 4, the interferometer observations and 12CO (J = 1 ! 0) data analysis in Sects. 5 and 6. Statistical results on various envelope parameters (flux, size, mass loss rate, . . . ) derived from these observations are described in Sect. 7 and an individual presentation of peculiar objects is given in Sect. 8. In Sect. 9 we draw our conclusions. 2. The sample The initial criteria we have applied for the selection of a representative sample of AGB and post-AGB stars were: line strength, IR colors and chemical composition. All the stars in the sample were already known from previous observations to be relatively strong CO emitters with intensities, in units of main beam temperatures, larger than 0.5 K and 1.0 K in the (J = 1 ! 0) and (J = 2 ! 1) transitions, respectively. The circumstellar envelopes were further selected by their IR colors, to sample as evenly as possible the IR color diagram (see Fig. 1 and Table 3), and by chemical composition. Finally, to take advantage of the sensitivity and resolution of the IRAM interferometer we have restricted our sample to high declination stars which would be either unresolved, or barely resolved, at the 30 m telescope. Our initial sample contained 53 AGB and post-AGB stars: 24 O-type stars, 25 C-type stars, 2 S-type stars and 2 of uncertain composition, according to the compilation by Loup et al. (1993). The selected stars are listed in Tables 2 and 4. 3. IRAM 30 m observations 3.1. Observing conditions The 12CO (J = 1 ! 0) and 12CO (J = 2 ! 1) observations were carried out simultaneously with the IRAM 30 m telescope, Spain (Baars et al. 1987), in three runs between October 1990 and August 1993. The telescope was equipped with 3 mm and 1 mm SIS receivers operating simultaneously in single sideband mode. The antenna temperature scale was calibrated every 10 minutes by successively observing a cold load, a room temperature load and the sky. The calibration accuracy, checked on well known molecular sources (Mauersberger et al. 1989) was found to present an uncertainty of less than 20%. The main beam temperatures TMB reported in the gures and in Table 4 are related to the antenna temperature scale TA by TMB = TA/ , where , the ratio of the main beam to the forward beam e ciency, is equal to 0.45 for the 12CO (J = 2 ! 1) observations and 0.60 for the 12CO (J = 1 ! 0) observations. The backends were at both frequencies lterbanks of 512 or 256 channels 1 MHz wide, providing a velocity resolution of 2.60 km s−1 at the frequency of the 12CO of the 12CO (J = 1 ! 0) and (J = 2 ! 1) data di er, (J = 1 ! 0) line (115271.2 MHz) and 1.30 km s−1 at the since the 12CO (J = 1 ! 0) observations obtained with frequency of the 12CO (J = 2 ! 1) line (230538.0 MHz). the IRAM 30 m telescope were combined with the data In October 1990 and May 1991, the observations were from the IRAM interferometer before further processing. made in position switching mode, with a reference posi- This is discussed in Sect. 5.2. tion 15000 east to the star position. In August 1993, the To increase the signal to noise ratio, the 12CO (J = secondary mirror was wobbled every 3 s, with an ampli- 2 ! 1) spectra were smoothed to 2.6 km s−1, the velocity tude of 24000 (this technique provides very flat baselines). resolution of the 12CO (J = 1 ! 0) data, except for 8 en The pointing was checked every hour or less on nearby velopes with relatively narrow lines: R And, RAFGL482, continuum sources. It was found to remain accurate within R Aur, RS Cnc, Y CVn, V CrB, Cyg and T Cep. 500 during this time interval. To avoid distorsion due to pointing drifts, the maps were taken with short integration times (between 1 min and 3 min on each position) 4. Data analysis: 12CO(J = 2 ! 1) and completed in less than 1 hour. When longer integration times were needed to achieve the required signal to 4.1. Positions noise ratio, the maps were observed several times, with pointing checks in between, and subsequently added, us- The envelope central positions adopted to perform the ing the shift-and-add technique described in Sect. 3.2. 12CO (J = 2 ! 1) observations are those reported by In October 1990, the maps were observed with a grid Loup et al. 1993. Due to pointing uncertainties and to spacing of 500 in right ascension and declination. In May the shift-and-add technique used to combine several maps, 1991 and August 1993, the spacing was taken equal to discrepancies between these coordinates and the actual 7.500. Considering the telescope half power beam widths map centers are meaningless. In the position-velocity diaof 1300 for the 12CO (J = 2 ! 1) observations and 2100 for grams, however, the zero position of the 12CO (J = 2 ! 1) the 12CO (J = 1 ! 0) observations, these spacings ensure maps is chosen to correspond to the maximum of the 12CO that the maps are fully sampled. (J = 2 ! 1) emission. A total of 46 circumstellar envelopes have been mapped in 12CO (J = 2 ! 1) emission. 20 envelopes were mapped in October 1990, under ne weather conditions. 4.2. Fluxes and main beam temperatures The SSB system temperature was typically 600 K at 3 mm and 1000 K at 1 mm. 12 envelopes were observed in May The 12CO (J = 2 ! 1) main beam temperature TMB re1991, in less favorable conditions, with system tempera- ported in the parameter summary of each envelope and in tures of the order of 1000 K and 2000 K, respectively. The Table 4 is the average of the three central 1 MHz channels observations of 11 objects were completed in August 1993, in the spectral pro le observed towards the envelope cenin good weather conditions with system temperatures of ter. Channels a ected by a galactic CO contribution have the order of 800 K and 1200 K, respectively. In particu- been omitted. lar, the maps of 5 envelopes (IRC+10011, IK Tau, CIT6, The flux density reported in the summary and in Table IRC+20370, RAFGL3068) observed by Loup (1991) were 4 is the average value of the three central channels in the extended during this run. For 3 envelopes (RX Boo, Cyg spatially integrated 12CO (J = 2 ! 1) spectral pro le and R Cas), we reproduce and reanalyse here the observa- S(v) of each envelope. The integrated flux density pro le tions (June 1990) obtained and published by Bujarrabal has been obtained by adding the spectra with a uniform & Alcolea (1991). weight on a regular grid with a spacing d. The flux spectrum in Jy is given by 3.2. Data reduction The single-dish data were reduced using the CLASS soft- where Tij(v) is the main beam temperature observed at ware package. Only linear or parabolic baselines were sub- velocity v and positions (i; j), and where the grid spacing stracted from the spectra. When several maps of the same d expressed in radians is either 7.500 or 500 depending on object were observed, we checked that the lines intensities the map, and k is the Boltzmann constant. were in agreement and that the positions of the integrated A comparison of the spatially integrated flux spectrum intensity contours coincided within better than 1.2500 in S(v) with the flux spectrum collected in the telescope right ascension and declination. When the shift between beam towards the envelope central position, S0(v), allows two maps was found to be larger, one of the maps was comparison of the envelope extent with the beam size: if shifted relative to the other in steps of 2.500 in each di- both fluxes coincide, the envelope is unresolved by the rection. After relative recentering the maps were added to telescope beam. For a gaussian beam of half power beam obtain a single data cube. After this stage, the reductions width h (in radians), the relation between the main beam S(v) = 2k 2 X Tij(v) 2 d i;j (1) S0(v) = 2k h2 2 4 ln 2 T00(v) which is equivalent to a conversion factor of 7.35 Jy/K for the 12CO (J = 2 ! 1) line observed at the 30 m telescope. Both flux spectra, integrated S(v) and central S0(v) are plotted for comparison for each envelope (see Atlas). The 30 m telescope half power beam width at 230 GHz, was estimated to be 1300 1.500, including beam smearing e ects due to tracking errors. We also give in Table 4 the total 12CO (J = 2 ! 1) emission of the envelope (i.e. the flux integrated both spatially and over velocity). (2) 4.3. Velocities and line shapes The star systemic velocity Vlsr and the envelope expansion velocity Vexp have been derived from the full width at L = pM 2 − h2 zero level (FWZL) of the 12CO (J = 2 ! 1) lines. To minimize the e ect of noise on the derivation of the FWZL, we have de ned as the extreme line channels those for which the 20% to 90% intensity contours still show a centrally f = (L − l)=(L + l): peaked pattern. The resulting velocities are given in the parameters list of each envelope and in Table 4. The errors are typically one channel half-width (1.3 km s−1 for most envelopes, 0.7 km s−1 for the 8 narrow line envelopes mentioned above). We must keep in mind that this calculation of the expansion velocity tends to overestimate the true value of this parameter when the line pro les present conspicuous wings, the origin of which is expected to be related to local velocity dispersion or to the presence of bipolar outflows independent of the general envelope expansion. Fitting a truncated parabola systematically leads in such cases to lower estimates of the expansion velocity. This discrepancy can be signi cant in stars with low expansion velocities. In particular for objects in which the emission of the bipolar outflows is dominant, the meaning of an expansion velocity determined in this way, is just the maximal projection on the light of sight of the axial velocity. Most envelopes present sharp-edged and round-topped lines, as expected for optically thick emission coming from a spherically expanding envelope. In a few cases (OH127.8+0.0, U Cam, S Cep IRC+60427) an interstellar 12CO (J = 2 ! 1) contribution appears as a narrow peak or dip in the spectrum. As far as the signal to noise ratio of the data allows to conclude, some envelopes show \unusual" line shapes, with wings or shoulders, which suggest changes in the mass loss activity of these stars (see for instance U Cam, T Dra, RS Cnc and M1-92). temperature pro le T00(v) observed towards the envelope 4.4. Sizes and asymmetries center and the flux spectrum in the central beam S0(v) is (3) (4) (5) (6) To derive a more quantitative estimate of the envelope size in the 12CO (J = 2 ! 1) line, we have tted as a function of the velocity channel a 2D gaussian to the observed spatial distribution. Plots of the tted minor and major axis (without beam deconvolution) and of the centroid shifts relative to the mean centroid position are shown for each envelope as a function of the channel velocity. In a spherical expanding envelope with a constant (or monotonically increasing) radial velocity, the radius of the region emitting at a given radial velocity Vz or −Vz with respect to the systemic velocity is proportional to p1 − (Vz=Vexp)2. A decrease of this radius with the absolute value of Vz is, therefore, expected. As shown in the gures, such a behaviour is indeed observed in most sources, although the uncertainties in the measured size is high. Once deconvolved from the telescope beam (h), the tted major and minor (M and m) axes averaged over the three central channel maps provide a measure of the envelope size (L and l) l = pm2 − h2 and a quantitative estimate of its asymmetry, given by the parameter We have carried out a worst-case beam deconvolution to lter out envelopes with intrinsic elliptical shapes and mask out those whose asymmetry is possibly due to beam distortion. For this deconvolution, the telescope beam was assumed to be elliptical (14:500 11:500) and its position angle was made to coincide with the position angle of the gaussian t. This deconvolution provides a lower limit Lmin to the actual envelope major axis and an upper limit lmax to its actual minor axis. We have then performed a conservative selection of the \non-spherical" envelopes, considering that any real departure from circular symmetry shows up even in the lower limit to the asymmetry parameter fmin given by fmin = (Lmin − lmax)=(Lmin + lmax): When fmin is lower than 0.1, which corresponds to a difference smaller than 20% between the axis limits, or when the sizes L and l are not determined with better accuracy than 5 , the envelope is considered as spherical. We choose such a highly conservative limit to rule out any asymmetry due to possible uncertainties in the determination of the telescope beam or of the envelope sizes. In such a case, we give a single size in the identi cation panel of each star and in Table 5, which is the average of the tted axes deconvolved by a circular beam of width h = 1300 s = p(M + m)2=4 − h2: Otherwise, the envelope is considered as asymmetric. Table 5 contains the envelope deconvolved minor l and major L axis (assuming a circular beam h) and the result- 5.2. Data reduction ing asymmetry parameter f . In addition, the plot of the envelope emission centroid versus velocity provides informations on the geometry and kinematics: for a spherically expanding envelope the centroid is expected to keep the same position at any velocity. On the other hand, if there is departure from overall spherical symmetry, the centroid is expected to move along an axis of symmetry. This behaviour is particularly obvious for RS Cnc, M1-92 and R Cas but exists also possibly for a few other envelopes, namely for Y CVn, Cyg, T Cep, IRC+40540 and IRC+60427. Data calibration was e ected in the baseline-based manner using the CLIC (Lucas 1991) interactive software package. We used whenever available either 3C 273 or 3C 345 as bandpass calibrator. Phase and amplitude calibrations were carried out with the sources listed in Table 1. Conventional image restoration algorithms like CLEAN produce maps without sidelobes and without features that a synthesized antenna pattern can show as a consequence of missing information. Accordingly, these methods imply interpolation or extrapolation of information that can be retrieved under \reasonable" assumptions, only when the objects are of small extent, compared to the size of the primary eld of view. 5. IRAM interferometer observations However, when the emission structures are large with respect to the beamsize of an individual dish (spacings smaller than the dish diameter are for obvious physical 5.1. Observing conditions reasons fundamentally inaccessible to measurements by an interferometer), short spacing information needs to be supplied. We circumvented this di culty by using the A total of 52 objects was mapped at the frequency 30 m telescope to measure the missing information. of the 12CO (J = 1 ! 0) line with the IRAM in- While the IRAM interferometer is ideal for achieving terferometer (PdBI) placed at Plateau de Bure, France the required high resolution, emission structures larger (Guilloteau et al. 1992). The observations consisted in than 1700 are essentially resolved out: 1700−1 down to about multi-con guration snapshots, each object being typically 400−1 is the typical range of spatial frequencies to which observed in four di erent con guration at various times our observations were sensitive. This results in interferover a period of two years (see Table 1 { December 1990 ometric maps in which regions of negative intensity surto April 1992). The objects were observed typically at two rounding small-scale structures can hardly be removed. hour angles in each con guration, each hour angle consist- For envelopes larger than 2000, including the 30 m data ing of 4-minute integrations on primary calibrators inter- recovered typically 50% of the flux missing from the maps spersed with two 20-minute periods of on-source integra- obtained from interferometer data alone. tion time. The 30 m telescope was used to obtain maps sampled The observing parameters were consistent from one at twice or more the telescope beamwidth, those maps run to the next. Two antennas were equipped with SIS that completely cover the region of the interferometer primixers, while the third had a Schottky mixer. All receivers mary beam and that greatly account for the missing flux. were operating in double side band (DSB) mode. The Short spacing visibilities were created from the single-dish cross-correlator was set to a bandwidth of either 40 MHz data and combined with the interferometer samples using or 80 MHz with a nominal channel spacing of 0.625 MHz. the data reduction scheme described below. Depending on Because of channel apodization, the e ective frequency the signal-to-noise ratio, however, only short-spacing visiresolution was 1 MHz, or equivalently 2.6 km s−1 at the bilities up to 20−25 m were kept to avoid spurious e ects frequency of the (J = 1 ! 0) transition. We used a redun- in the merging process. dant correlator setting with 64 channels per baseline, in- The following steps were taken with the GILDAS stead of 128. The on-source integration time equivalent to (Guilloteau & Forveille 1989) software package in combinsingle-baseline observing was 4.5 hr on average. The point ing data from the IRAM 30 m telescope and the IRAM source sensitivity, derived from emission-free channels, interferometer, once the initial data processing of singlewas 50 mJy on average, and was consistent with system dish and interferometer data was carried out: temperatures of 500 K and atmospheric phase decorrelations of 10 to 30 degrees. The IRAM continuum correlator, resample the interferometer data to a frequency resowhich was operating simultaneously over a bandwidth of lution of 1 MHz to be consistent with the single-dish 500 MHz consisting of 10 contiguous 50 MHz channels, was observations, used mainly for the purpose of calibration. However, as a correct the antenna temperatures for the forward gain valuable by-product, we were able to measure with the to obtain main beam temperatures, TMB, and apcentral 400 MHz the continuum flux at 112 GHz of CIT 6, ply the standard conversion factor to convert the Cep (tentatively detected) and 19480+2504, and to give single-dish data from TMB to the interferometer units 5 sensitivity upper limits for all other sources. Jy beam−1, interpolate irregularities in the spatial sampling of the single-dish data (missing or bad sampling points) using a CLEAN based algorithm. The purpose of this step is to avoid an all too poor representation of the large emission structure when the sampling was too irregular, deconvolve the single-dish maps by dividing their Fourier transforms by the Fourier transform of the 2200 beam of the 30 m telescope (assumed to be gaussian), multiply the inverse Fourier transform of the result by the 4200 primary beam of the interferometer, and Fourier transform back to the uv-plane to obtain single-dish visibility tables, recenter the single-dish and the interferometric emission centroids, determined from averaging blue (V −Vexp=3) and red (V Vexp=3) velocity maps, to the phase tracking center of the interferometer. Errors in the pointing (200 − 400 are reported for the 30 m telescope) need to be removed as accurately as possible. Since the 30 m velocity maps contribute predominantly to the large scale structure, residual un- tted emission centroids in the blue (V −Vexp=3) and certainties shoud not signi cantly a ect the results. red (V Vexp=3) velocity maps. When two-component elgenerate visibility tables that combine both single- liptical gaussians were used to t the CO brightness distridish and interferometer data from which visibility plots bution (see Sect. 6.4) we used the positions of the gaussian along circular uv trajectories are created. The single- ascribed to the inner, more compact shell surrounding the dish weight is scaled according to the mean weight of central star. the short spacings interferometer visibilities in order Except for 12 stars, most of the positions listed in to provide single-dish and interferometer uv cells of Table 4 are in excellent agreement with those given in similar weight, Loup et al. (1993). The stars for which the positional difcheck the compatibility of single-dish and interferom- ference is larger than 300 are noted in the caption to Table eter data in the common uv regions (spacings in the 4. Single-dish 12CO (J = 2 ! 1) maps were used just in 15−25 m range) and, depending on the signal-to-noise, the case of CL Mon and HD 187885 for which no 12CO account for possible flux calibration errors, C, by scal- (J = 1 ! 0) interferometer data were available. ing the single-dish visibilities to the interferometer visibilities. C was found to vary between 0.5 and 2. In general, however, the two data sets agreed to about 6.2. Fluxes and main beam temperatures 20%. The 7 objects listed in Table 2 were discarded from the original sample. 6. Data analysis: 12CO(J = 1 ! 0) 6.1. Positions We have used interferometric maps (not combined data) to determine star positions. Mostly VLA positions were used for the primary phase calibrators. We have ignored phase errors due to inaccuracies in the assumed baselines, as the distance s between the stars and their calibrator(s) was generally below 15 degrees. Actually, baseline errors B at the frequency of the 12CO (J = 1 ! 0) transition lead to a maximum phase error 2 B s=c of a few degrees, somewhat smaller than the atmospheric phase fluctuations recorded on the shortest baselines, typically 10 − 30 degrees. On average, this corresponds to absolute positional errors of about 1/10 of the synthesized beam, or equivalently 0.500. The star positions listed in Table 4 were obtained by averaging the positions of the To properly bootstrap the fluxes of the stronger calibrators (mostly 3C 84 and 3C 273), we have made measurements of these, in conjunction with planets, with the antennas operating in the autocorrelation mode. We have then referred the weaker calibrators to the rst ones in the more sensitive interferometric mode. Since more than half of the calibrators were used to calibrate more than one source (e.g. 2005+403 was used alternatively for 10 stars), remaining errors in the flux density scale of those were largely removed by cross-checking the interferometric visibility pro les on a sample of stars across the whole set of array con gurations. After readjusting discontinuities in the visibility pro les attributed to errors in the amplitude calibration, we expect the fluxes of the main calibrators to be accurate within 10%. Table 1 lists the flux of all the calibrators used in the survey. Despite an accurate relative flux calibration, the absolute calibration scale is probably not better than 15%. The spatially integrated 12CO (J = 1 ! 0) fluxes have been estimated from the combination of both, single-dish and interferometric visibility pro les. We have tted either one or two elliptical gaussian components to the real part Ident Name Region Chem. type Distance (kpc) a No (1{0) data. b No distance given by Loup et al. (1993). Photodissociation radius derived from the relationship RCO vs. CO. Distance derived from CO modeling. c No distance given by Loup et al. (1993) and CO modeling does not converge. d No parameter derived as envelope is asymmetric or ring-like. e CO modeling gives two solutions. The alternative is D = 0:76 kpc, RCO = 2:1 1017 cm and M_ = 14 10−6 M /yr. of the global visibility pro le of each star, As a result, we where D= is the HWHP projected baseline of the pro le have derived the integrated flux from the tted value at in units of the observing wavelength. In case of elliptizero spacing for each velocity channel. When either the uv cal gaussians a size is tted along the major and minor coverage was inadequate or the line strength insu cient axes. The gaussian curves corresponding to the tted ento account for possible departures from circular symmetry, velope sizes in the direction of maximum and minimum we have tted only circular gaussian components. Just in extension are shown for each star on the individual 12CO the case of U Cam, we have tted a circular ring to the (J = 1 ! 0) pages (central panel, second row) of the atlas. visibility pro les corresponding to the central velocities. As for the 12CO (J = 2 ! 1) data we have de The spatially integrated fluxes listed in Table 4 are termined the asymmetry of the envelope from the mapeak values. Each value was determined by averaging the jor and minor axes (see Table 5 { consult the atlas for flux in a small range of velocities centered on Vlsr. This the velocity dependence of the asymmetry). Double elrange of velocities is delimited by an horizontal bar in the liptical gaussian were used for 17 envelopes, three of integrated 12CO (J = 1 ! 0) flux vs velocity plots. which (IRC+10011, HD 235858, RAFGL 3068) show a Particular care was given to elds with signi cant clear asymmetry in the compact, central component, two emission far from the tracking center (e.g. 04307+6210, (RAFGL 2155 and Cyg) in the extended, outer compoCIT6, Cyg, IRC+40540) as here the quality of the single- nent, and one (IRC+40540) in both components. IK Tau dish contribution appears to be far more important than has a visibility pro le which is apparently too complex to the interferometric one for the determination of the inte- be tted by a double component pro le. grated fluxes. Except for U Cam, where we have tted a circular The main beam temperatures listed in Table 4 were ring, almost all visibility pro les are well-approximated determined directly from the peak integrated 12CO (J = by gaussian pro les. 1 ! 0) flux densities. The main beam temperature, TMB, is related to the integrated flux by S = C γ TMB, where C accounts for residual calibration errors and 7. Statistical results where γ = 4:8 Jy K−1 is the nominal point source sensitivity of the 30 m telescope at the frequency of the 12CO (J = 1 ! 0) transition. However, TMBis not exactly the main beam temperature as measured by the 30 m telescope in extended sources, it must be much larger in very extended sources. The actual number of envelopes mapped in 12CO (J = 1 ! 0) with combined data from the IRAM interferometer and the IRAM 30 m telescope and mapped in 12CO (J = 2 ! 1) with the IRAM 30 m telescope is 46. Two objects have been mapped in 12CO (J = 2 ! 1) only (CL Mon, HD 187885) and two in 12CO (J = 1 ! 0) only (RY Dra and 19480+2504). 6.3. Velocities and line shapes 7.1. Table contents If available, only interferometric data were used to determine the expansion velocity in the 12CO (J = 1 ! 0) The results of the 12CO (J = 1 ! 0) and 12CO (J = 2 ! line. As for the 12CO (J = 2 ! 1) lines, the velocity was 1) maps analysis are given for each object and summarized not determined by pro le tting but from edge channel in Tables 4 and 5. The rst column of Table 4 contains maps still showing emission features. Terminal Vexp and the abbreviated IRAS designation and the second gives systemic Vlsr velocities of the circumstellar envelopes were the most common name. The next two columns show the determined from channel maps prior resampling to the equatorial coordinates (epoch J2000) of the source, obfrequency resolution of the 30 m telescope. tained by tting interferometric data alone, and the es Centroid and expansion velocities measured from the timated errors. Column 5 lists either the flux density of 12CO (J = 1 ! 0) line are listed in Table 4. The same the radio continuum at 112 GHz or gives 5 upper limits. comments pointed out in Sect. 4.3 on the meaning of our Columns 6{11 and 12{17 tabulate the results obtained in estimation of the expansion velocity and on the presence the (J = 1 ! 0) and (J = 2 ! 1) line emission, respecof pro le anomalies also hold here. tively: the terminal 12CO velocity of the envelope (i.e. the expansion velocity or the maximum projected velocity), the systemic velocity with respect to the local standard 6.4. Sizes and asymmetries of rest, the spatially integrated flux density at the systemic velocity, the velocity integrated lined intensity, and We have obtained envelope sizes by tting either one or the main beam temperatures as observed towards the star two elliptical gaussian components to the real part of the position. The epochs of observations for the interferometer combined visibility pro les in the range −Vexp=3 V are in Col. 12, for the 30 m telescope in Col. 17. Vexp=3. The envelope size CO of a circular gaussian visibil- Table 5 lists the results of the gaussian ts. Columns ity pro le was derived according to CO = 2 log 2=( D= ) 3{12 and 13{17 give the results for the 12CO (J = 1 ! 0) and (J = 2 ! 1) transitions, respectively. Columns 3{7 (index 1) tabulates the flux density, the major and minor sizes, the position angle (east to north) and the degree of asymmetry of the main envelope component. When a second component was tted to the visibility pro le, the results are given in Cols. 8{12 (index 2). The major axis is marked \U" (unresolved) when the envelope was found to be roughly equal or less than the estimated beamsize of the 30 m telescope, and is marked \?" when the data was to scarse or noisy to allow any size determination. 7.2. Atlas contents The results of the 12CO (J = 1 ! 0) and (J = 2 ! 1) data analysis are presented at the end of the paper. There are two pages for each object, one for each transition. The pages are ordered by increasing right ascension. The page heading identi es the 12CO transition and the star's most common name and coding in equatorial coordinates (epoch B1950). Each page is divided in four panel rows: row 1: while the three panels on top of the 12CO (J = 1 ! 0) page show channel maps obtained from the combination of the interferometric and single-dish data, the panels on top of the 12CO (J = 2 ! 1) page show channel maps obtained from single-dish data alone. The integrated velocity intervals for the panels are roughly −Vexp V −Vexp=3, −Vexp=3 V Vexp=3 and Vexp=3 V Vexp. The range of velocities used for the central panel is indicated in the integrated flux panel. The synthesized beam obtained from the combination of the two data sets is shown in the inset. The contour levels are marked in the wedge on top of each panel and start at ve sigma. Map units are in Jy/beam and K km/s. The 12CO (J = 2 ! 1) maps were not recentered to zero o set, the central position being undetermined deviations up to 2.500 are possible. row 2: the left panel 12CO (J = 1 ! 0) shows the integrated flux obtained at zero spacing by tting gaussian pro les to the visibilities. The error bar is set to the one sigma noise level W −1=2, where W is the weight of the single-dish data at zero spacing. The central panel shows the global visibility pro le for velocities in the −Vexp=3 V Vexp=3 range and gaussian curves representing the major and minor axes of the tted proles. The uv{coverage is shown to the right. The left panel 12CO (J = 2 ! 1) shows the flux integrated over the channel maps (continuous line) and the flux towards the star position (dashed line). The central and right panels show the apparent full widths along the major and minor axes, prior to beam deconvolution. row 3: the major and minor axes of the gaussians which were tted to the global visibility pro le are shown in the left and central panels of the 12CO (J = 1 ! 0) page. The asymmetry parameter (L − l)=(L + l), where Fig. 2. Comparison of integrated (J = 2 ! 1) vs. (J = 1 ! 0) 12CO fluxes. The dotted line traces envelopes in the optically thick F230=F115 = 4 limit L and l are the major and minor axes respectively, is shown to the right. The 12CO (J = 2 ! 1) left panel shows the radial intensity pro le (continuous line) and the presumed shape of the single-dish beam (dashed line). Positional o sets in right ascension and declination where obtained by tting gaussians to the channel maps and are given in the central and right panels. row 4: the 12CO (J = 1 ! 0) panel displays positional o sets in right ascension and declination (left and center) and position angle (north to east) as obtained by tting gaussians to the global visibility pro les. The 12CO (J = 2 ! 1) panel lists the name of the source, the equatorial coordinates (epoch J2000) obtained by tting only interferometric data, the interferometric on-target time for equivalent single-baseline observations, the number of snapshots, the synthesized beam and the one sigma noise level in the 12CO (J = 1 ! 0) channel maps shown in the top panel row, the primary calibrators, the amount of flux retrieved by the interferometer, and it summarizes the results found in both the 12CO (J = 1 ! 0) and (J = 2 ! 1) transitions. We have not included the low-declination source VX Sgr in the atlas and in Table 5, as the data quality was too marginal. 7.3. Statistical error estimates Positional errors are normally distributed, whereas errors in flux density and width, as a consequence of their positivity, are not. To quantify the uncertainties in position, width and flux we have carried out MonteCarlo simulations. We modeled circular gaussian pro les sampled at twice the angular resolution of the telescope with normally distributed noise. The results (see also the Appendix) show that the probability to underestimate the width of a gaussian depends on the signal-to-noise ratio in the samples. At signal-to-noise ratios larger than 10 the probability to underestimate or overestimate the width is almost even, the probability distributions become gaussian and the uncertainties can well be approximated by standard deviations. Similar results are found for the flux density estimates. 7.4. Integrated fluxes We have rst compared the relative 12CO (J = 2 ! 1) / (J = 1 ! 0) intensities in our source sample. In Fig. 2 we show the integrated flux of both lines (Sects. 4.2, 6.2); the dotted line represents the expected ratio for optically thick emission and a common excitation temperature, which is equal to 4. As we see, the optically thick ratio agrees well with observations in most cases. In two oxygen rich objects, RS Cnc and R Cas, the observed ratio is higher than expected. The CO pro les in these stars are peculiar and do not show the parabolic or flat-topped shape expected for optically thick lines. It is then possible that in RS Cnc and R Cas the J -dependence of the opacity leads to the observed high line ratio. In other objects, like the carbon rich CIT 6 and the oxygen rich IRC+20326, the (J = 2 ! 1) / (J = 1 ! 0) relative intensity is smaller than 4. This is not likely to be due to pure opacity e ects but to a particularly low excitation of the (J = 2 ! 1) line. Note that it is not impossible that in certain cases the excitation and line-strength e ects cancel, which could yield line ratios close to the optically thick limit. However, the systematic presence of optically thick emission in the 12CO lines is con rmed by the general properties of the observed pro le shapes (Sects. 4.3 and 6.3) and we conclude that such departures from the optically thick situation are rare. There certainly is a strong correlation (see Fig. 4) between the CO luminosity and the physical extent of the envelope. Such a correlation may be partially due to the effects of the errors in the assumed distance on these parameters, since the distance value enters the determination of both the luminosity and linear size. However, such errors are not expected to exceed a factor 2, which is not enough to explain the empirical relation. Moreover, a correlation between the envelope thickness and the radius of the CO emitting region is expected if this is mainly given by CO photodissociation (see below). We accordingly think that the relation depicted in Fig. 4 between the CO luminosity and the CO radius is, at least partially, real. We also can see in Fig. 4 that oxygen rich envelopes seem to be less extended and luminous than carbon rich ones. This could be a direct consequence of the selection criteria used to set up Fig. 3. Comparison of the envelope sizes in the 12CO (J = 1 ! 0) and (J = 2 ! 1) line emission. In the density bounded limit (dotted line), when the envelope runs out of 12CO molecules, both the (J = 1 ! 0) and (J = 2 ! 1) transitions trace the same regions. The size of envelopes smaller than the 1300 beam of the 30 m telescope is generally overestimated the star sample. In fact, the selected oxygen rich sample is on average 3 closer than the carbon rich one, which consequently resulted in a typically 10 more luminous than the carbon star sample. However, we did not select nearby oxygen rich stars and farther carbon rich objects, we just chose the most intense (and better studied) sources. This distance factor mainly reflects the fact that oxygen rich stars are more abundant and then can be found closer to us. Again, if the distance errors are not very large, the separation between the di erent groups in Fig. 4 is real and we must conclude that absolute CO luminosities are larger in carbon rich stars. 7.5. Expansion velocities We have measured the 12CO (J = 2 ! 1) and, when available, the (J = 1 ! 0) terminal velocity for all the circumstellar shells in our sample. The expansion velocities are, as expected, signi cantly similar for both CO lines. Only in ten objects the di erences exceed 10%, and only in three stars, V CrB, T Cep and R Cas, the di erences exceed 20%. In all these cases the di erence between the expansion velocities derived from the 12CO (J = 1 ! 0) line and from the (J = 2 ! 1) line can be considered as negligible due to the noise level, uncertainties in the baselines or lack for spectral resolution. V CrB and T Cep show weak and narrow CO lines which, particularly at 115 GHz, makes di cult a good measurement of the line width. In spite of the peculiar pro les of R Cas which are quite different in both lines, a close inspection reveals again that the adoption of the same FWHP for both lines would not be incompatible with our data. Although most sources show, as we have mentioned, the rounded or flat-topped pro les characteristic of circumstellar envelopes, anomalies are (at some level) present in a non-negligible number of stars. Some sources show more or less prominent spikes (U Cam, RS Cnc, R Cas, OH127.8+0.0, 04307+6210, S Cep) or relatively intense wings (M1-92, IRC+10420). In some cases, like U Cam, the pro le seems in fact to be composite due to the particular structure of the envelope. Except for the two peculiar sources M1-92 and IRC+10420 with expansion velocities well above 25 km s−1, most of the velocities seem to fall all over the range found for AGB envelopes. M1-92 is known to present a high-velocity bipolar outflow that contributes to most of the pro le width. This is probably also the case of IRC+10420. Both sources are described more in detail in Sect. 8. 7.6. Sizes and asymmetries The envelope sizes measured in both lines are compared in Fig. 3. As we see, many stars show size ratios close to 1, as expected if the observed radii were given by a cut-o of the CO density at a given point. A certain number of stars, however, shows a somewhat larger size in the (J = 1 ! 0) line, which must be due to level population e ects, since this line is of course easier to excite (in contrast opacity e ects would lead to larger sizes in the (J = 2 ! 1) line). This conclusion is strengthened by the fact that most sources showing a signi cant discrepancy with the hypothesis of a common spatial cut-o for both lines, like CIT 6 and IRC+20326, also show a low (J = 2 ! 1) to (J = 1 ! 0) ratio. But we must note that others, in particular IRC+40540, are anomalous in relative size but not in relative intensity: IRC+40540 which is carbon rich, shows clearly parabolic CO pro les (see also notes on individual envelopes). Fig. 5. Comparison of measured sizes and calculated photodissociation radii in 31 envelopes. Photodissociation radii are calculated according to Loup et al. (1993). The dotted line is a linear t (see Eq. 7) Most envelopes are found to show a signi cantly circular appearance. However, the global visibility pro les show that there is a non-negligible number of stars with peculiar circumstellar envelopes. It appears that many of these are surrounded by an inner envelope and an outer shell with morphologies that are not easily interpreted owing to the limited resolution achieved here. The bulk of the CO emission is in general found in the outer envelope which mostly appears spherical symmetric. The presence of an inner envelope, in some case indirectly suggested by a central tip in the spectral pro le, suggests quite a substantial, if not abrupt, change in the mass loss rate in the envelopes of these stars testifying for a thermally pulsing activity. Some of these stars may already have left the AGB. Another nding is that a fairly large number of these envelopes shows a more or less pronounced asymmetry. The 12CO (J = 2 ! 1) observations rmly establish the existence of a pronounced morphological asymmetry in the outer envelope of RAFGL 2155 and Cyg, an asymmetry which is likely to be due to an anisotropic interstellar UV radiation eld. In some objects, however, departure from sphericity occurs already in the innermost regions. This is in particular the case of the envelopes around RAFGL 3068 and IRC+40540, and possibly also around RAFGL 3099 (not indicated in Table 5). As shown in Table 5 and in the atlas, there are hints for developing asymmetries in many other stars, but the limited dynamical range hardly allows to draw clear conclusions. 7.7. Mass loss rates and photodissociation radii We have used our 12CO (J = 1 ! 0) data to perform a self-consistent calculation of the envelope mass loss rate M_ (7) and photodissociation radius RCO, following the method described by Loup et al. (1993). The input parameters are: 1: the envelope expansion velocity measured in 12CO (J = 1 ! 0) and listed in Table 4, 2: the distance of the star derived from the bolometric fluxes, as given by Loup et al. (1993 { Table 2 herein), 3: the main beam temperature of the 12CO (J = 1 ! 0) line as seen by a 7 meter telescope (the calculation assumes that the envelope is unresolved) which we derive from the integrated 12CO (J = 1 ! 0) flux (Table 4). 4: the fractional CO abundance which was assumed to be constant and equal to 5 10−4 for oxygen rich stars and 10−3 for carbon rich stars and S-type stars (Zuckerman & Dick 1986). We show in Table 3 the calculated photodissociation radii and mass loss rates, together with the assumed distances and some other properties. In Fig. 5 we show the comparison of the computed photodissociation radii of these envelopes, RCO and their measured half-maximum radii, CO. The correlation between both parameters is obviously signi cant. In our procedure, the photodissociation radius is calculated from the CO total intensity (through the mass loss rate), therefore Fig. 4 reveals the existence of a correlation between the CO luminosity and the envelope size. In the case of optically thick emission, a relation of this kind can be expected, the radius being approximately identied with the point at which the line emission becomes optically thin and from which the emission starts decreasing sharply. However, in such a case the extent of the (J = 2 ! 1) line, which is always the most opaque, would have systematically shown a larger size than the (J = 1 ! 0) line, which is not the case in our data, see Sect. 7.4. Accordingly we conclude that the correlation shown in Fig. 5 indicates that the e ects of photodissociation dominate the CO envelope size. It is also noticeable in Fig. 5 that the measured envelope radius is always smaller than the calculated photodissociation radius. This is indeed expected, as the radius CO is equivalent to the size of the envelope at half the maximum intensity (see Sect. 6.4) whereas the photodissociation radius RCO delimits a spherical region around the star in which the total mass of CO molecules is contained. The size of this region is mainly given by the intensity of the interstellar UV radiation eld and the mass loss rate, since the molecule shielding from the external eld depends on the envelope opacity. Unquestionably, the mass loss processes that take place are certainly not smooth, steady and spherical in all the circumstellar envelopes of the sample. However, except for RY Dra, V814 Her, RAFGL 2155, Cyg and IRC+40540 which are slightly asymmetric at larger scales, for U Cam which has a detached shell, and for all envelopes with a hint for asymmetry (see Table 5), all other envelopes in the sample can reasonably be considered as spherical from an observational point of view and the simple model described above was applied to them. Thus, leaving aside all envelopes of non-spherical appearance, we derived from a sample of 31 stars an empirical formula with which we estimate the photodissociation radius RCO of an envelope from the measured CO radius CO at half maximum intensity RCO = 1:5 CO 00 + 8:400: Even though, this relation (dotted line in Fig. 5) may not be too appropriate for estimating the photodissociation radius of an envelope, we have subsequently used it to compute the photodissociation radius of envelopes of, to our knowledge, unknown distances. Accordingly, by reversing the computational scheme described by Loup et al. (1993), we derived estimates for the mass loss rates and the distances of IRC+10420, 19480+2504, 20028+3910, HD 235858 and 23321+6545 (see Table 3). The calculations, however, were inconclusive for IK Tau and CIT 6, the last one showing hints for an asymmetric structure. 7.8. Non-detections and marginal detections We have already seen in Table 2 that for 7 sources the quality of our data did not allow a proper interpretation. In some cases, galactic contamination was important, in others the CO emission was too weak. For OH39.7+1.5 and OH104.9+2.4 both contamination and weak emission are noticed. AC Her was not detected, we therefore do not con rm a previous tentative detection (Alcolea & Bujarrabal 1991) at a level of TMB(2 ! 1) 0:1 K, although our noise rms at this frequency is about 0.04 K, not low enough to draw de nitive conclusions. In some sources like R LMi and Cep the (J = 2 ! 1) line is intense, but the (J = 1 ! 0) emission is too weak. In R LMi our TMB 0:1 K is compatible with the data by Bujarrabal et al. (1989). Le Borgne & Mauron (1989) give an intensity TMB(2 ! 1) 0:15 K for Cep. Our data con rm this gure, but we found the (1 ! 0) line too weak ( 0:04 K) to be accurately mapped. We only show CL Mon in the (J = 2 ! 1) emission, the (J = 1 ! 0) data was too scarse. 8. Individual envelopes U Cam is a carbon-rich star that has probably experienced a change in its mass loss (Olo son et al. 1990). The 12CO radial visibility pro le and the integrated flux pro le have revealed a detached shell of 1000 radius surrounding the central star. Recent observations with the PdBI in the HCN (J = 1 ! 0) line emission support this picture (Lindqvist et al. 1996). 04307+6210 is classi ed as carbon-rich M-star (Groenewegen 1994, formerly associated to IRC+60144, an optically visible star of 12 magnitude). The 12CO (J = 1 ! 0) radial visibility pro le and the spiky integrated emission pro le suggest that the star is surrounded by an inner 800 and an outer 3400 shell, one of the two largest in our sample. This leads to the conclusion that variations in the mass loss of 04307+6210 have already taken place. RS Cnc is classi ed as an MS giant, a thermally pulsing AGB star whose surface is enriched from dredged up material. The star for which Ake and Johnson (1988) suggest a binary nature, is surrounded by a thick envelope with thermal line emission detected in several molecules (HCN, CS, SiO. . . ). The position-velocity diagram in both lines testi es the clear bipolar nature of the 12CO envelope. The spiky emission pro le which does not appear to follow the velocity gradient of the broader peak, recalls the position-velocity diagram of the broad and narrow emission components observed in X Her (Kahane & Jura 1996). These properties are indications that the envelope does not show the expected spherical structure in isotropic expansion, being possibly in a phase of strong evolution. The complex width and velocity-position diagrams suggest a very clumpy envelope, quite di erent from a spherical shell. The object is probably the precursor of very excited and structured PN, as NGC 6302 and NGC 7027. We determined the distance according to the method described in Sect. 7.7. We found two possible estimates D = 6:12 kpc, which is close to the value commonly adopted, and D = 0:76 kpc (see also Kastner et al. 1995). M1-92 also known as Minkowski's Footprint, is a well studied proto-planetary nebula. Both, the 12CO (J = 1 ! 0) and (J = 2 ! 1) observations reveal the clear bipolar morphology of the 600 dense shell, remnant of the AGB stage. Recent interferometric observations in the (J = 1 ! 0) line emission bear evidence of two 300 wide cavities along the nebular axis, probably formed by the passage of a bipolar bow shock (Bujarrabal et al. 1994b, 1997). This supports the idea that stellar wind interaction is the process that very likely dominates the circumstellar dynamics of post-AGB stars and of M1-92 in particular. T Cep is an oxygen-rich Mira variable. The 12CO (J = 2 ! 1) velocity position diagram (the 12CO (J = 1 ! 0) data is probably sensitivity limited) shows the northwest south-east signature of an axial velocity gradient, the velocity increasing steadily with distance from the central star. CIT 6 is a carbon-rich star, the most intense and one of the most extended sources in our catalog. The envelope appears slightly asymmetric but no systematic variation with the velocity of the emission centroid is found. Our interferometric maps show the CO emission to be distributed in two components, in a compact one and in a weak and extended halo, both centered on the same position. Other sources show a similar behaviour (see for instance, IRC+10011, 04307+6210, IRC+20326, RAFGL 2155, IRC+20370, Cyg, and RAFGL3068), although less clearly. RAFGL 3068 is a well-known carbon-rich star (Sopka et al. 1989) with strong emission 12CO pro les of nearly parabolic shape. The radial 12CO (J = 1 ! 0) visibility pro le is characteristic of a large outer circumstellar nebulosity surrounding a compact inner envelope. The outer shell is presumably detached (this is suggested by the ringing features in the visibility pro le) and strongly bipolar. Y CVn is classi ed as a carbon-rich star, the brightest The position angle which is almost xed at −30 deg is J-type star known in the optical. Izumiura et al. (1996) re- likely to be due to an uneven sampling of the uv-plane. port the detection with ISOPHOT of a very large, slightly asymmetric, detached dust shell surrounding the central star. Though the envelope is not asymmetric (f < 0:1) according to our analysis, the (J = 2 ! 1) positionvelocity diagram show presumably the kinematical signature of a faint asymmetry developping in the innermost regions around the star. RAFGL 3099 is a carbon-rich Mira variable. Though the 12CO (J = 1 ! 0) emission traces essentially a spherical morphology, as suggested by the visibility pro le, the (J = 2 ! 1) emission is slightly asymmetric. Moreover, channel maps obtained at the nominal velocity sampling of the interferometer show a linear but faint signature in the position velocity diagrams typical of an expanding circumRAFGL 2155 is a carbon-rich star which shows a sys- stellar envelope with axial symmetry (south-east { northtematic behaviour in the emission centroid position as a west). function of velocity, for both transitions, with a general displacement from south-east to north-west of a few arcsec. As in other sources, this behaviour probably corresponds to the presence of a circumstellar component in bipolar expansion. R Cas is an oxygen-rich star presenting an aspherical con guration (Tuthill et al. 1994), probably related to the influence of a binary companion. The star is surrounded by a large, optically thick circumstellar envelope whose radial 12CO (J = 1 ! 0) visibility pro le show little if no deviaIRC+10420 is an oxygen-rich star surrounded by a heavy tion from circular symmetry (Bujarrabal et al. 1994a). The cool envelope. The nature of the star is still controver- data, however, reveal a remarkable kinematical continuity sial: Hvrinak et al. (1989) propose a post-AGB star evolv- (south-east { north-west) in both the 12CO (J = 1 ! 0) ing towards a young planetary nebula (PN), Oudmajer and 12CO (J = 2 ! 1) position velocity diagrams which et al. (1994) a very luminous hypergiant undergoing an clearly supports the idea that the circumstellar dynamics extremely rapid evolution towards a cool Wolf-Rayet star. is con ned to an axially symmetric structure. 9. Summary We have carried out an atlas of the 12CO (J = 1 ! 0) and (J = 2 ! 1) line emission in a selected sample of AGB and post-AGB stars. The survey observations of 46 circumstellar envelopes reported in this paper signi cantly improve the data available from previous 12CO surveys. The main results we obtained from the work are: We have determined interferometric positions for 44 envelopes. Interferometric positions are not available for CL Mon and HD 187785. As a rule, the absolute positional accuracy in the data is estimated to about 0:500. We have set 5 upper limits to the radio continuum at 112 GHz of 42 stars and determined the radio flux of CIT 6 and 19480+2504 to 13 mJy and 28 mJy, respectively. We tentatively detected Cep in the radio continuum. We have measured terminal and systemic velocities of 44 envelopes. Except perhaps for V CrB, T Cep and R Cas, the di erence in the 12CO (J = 1 ! 0) and (J = 2 ! 1) terminal velocities can be considered as negligible. The line shapes show for the most part the typical signature of optically thick envelopes. There are, however, line pro les with a strong central spike characteristic of a two-component envelope that in some cases may be related to the presence of detached shells as in U Cam, RS Cnc, R Cas, OH127.8+0.0, 04307+6210, S Cep and possibly T Dra. In addition, we found two line pro les with relatively strong wings, an indirect indication for aspherical mass loss activity, in M1-92 and IRC+10420. We have determined 12CO (J = 1 ! 0) and (J = 2 ! 1) flux densities and integrated intensities for almost all the stars in the sample. Consistency in the (J = 1 ! 0) results was achieved by minimizing possible calibration errors in the single-dish data, that is, by readjusting appropriately the single-dish flux density scale to the interferometric one. In a number of stars we found discrepancies with previous works. We analyzed the morphology, size and asymmetry of 46 envelopes. Within the sensitivity of our observations, AGB and post-AGB envelopes can be considered for the most part as of spherical appearance. There is, however, evidence for an inner shell surrounding the central star and/or for a signi cant envelope asymmetry in about 30% of the envelopes of our sample. The position-velocity diagrams of the oxygen rich RS Cnc, M1-92 and R Cas show a remarkable kinematical continuity in both the 12CO (J = 1 ! 0) and 12CO (J = 2 ! 1) line emission, an indication that the circumstellar dynamics in these stars is con ned to an axially symmetric morphology. Further envelopes but with less pronounced kinematical signature are found in Y CVn, RAFGL 2362, T Cep and 23321+6545. We have calculated mass loss rates and 12CO photodissociation radii for 38 stars. We have also established an empirical relation between the 12CO photodissociation radius of an envelope and the measured CO extent at half maximum intensity. This results indicates that very probably the CO extent is given by the envelope radius inside which CO is not photodissociated by interstellar UV radiation. Appendix The derivation of estimates for the uncertainties in flux S, position , and width of a one-dimensional gaussian brightness distribution centered at = 0 and sampled at locations j = j B =2 (j = 0 : : : N , B is the beamwidth) with additive noise of zero mean and variance 2 N requires the use of non-linear regression. A Gaussian prole is indeed non-linear in and according to B= B −2 and B= 2 B −3: (9) Though non-linear regression is required to derive estimates of the uncertainties in S, and , we used an approach suggested by the iterative Gauss-Newton method (Bates & Watts 1988) to derive a linear approximation to these estimates in the limit of high signal-to-noise (= S= N ). In this case the one sigma con dence limit for the position in the limit of high N can be written to −2 = −2 N N lim X( Bj = j )2 N!1 j=0 with similar expressions for S−2 and −2. According to equation (10) the uncertainties in S, and for the gaussian brightness distribution B are given by −1 = S −1 N p −1 = 4 2 log 2 0 N X j4=4(jR)2 j −1 R−1 −1 S N 0 N X j2=4(jR)2 j 1−1=2 Guilloteau S., Forveille T., 1989, Grenoble Image and Line −1 = 2 log 2 N−1 R−1 −1 S DOabtsaerAvantaoliyresisdeSyGsrteenmo,bIleRdAoMcu&meGntroupe d'Astrophysique 0 N 1−1=2 Healy A.P., Huggins P.J., 1988, AJ 95, 866 @Nl!im1 Xj 4−(jR)2A (13) IHzruimvniuakraBH.J.,.,HKawshoikmSot.,oVOo.l,kKKa.wMa.r,a1K98.,9,YAampJam34u6r,a2I6.,5Waters L.B.F.M., 1996, A&A 315, L221 where R = B= . Note that in the R = 1 limit (i.e. Kahane C., Jura M., 1996, A&A 310, 952 an unresolved object) the uncertainties can roughly be Kastner J.H., Weintraub D.A., 1995, ApJ 452, 833 estimated to S = N , = ( B=2)( N =S) and Knapp G.R., Morris M., 1985, ApJ 292, 640 = B( N =S). The errors in the estimates are reason- Loup C., 1991, Thesis, University Joseph Fourier, Grenoble ably small for samples with snr 10. In analogy, sta- Loup C., Forveille T., Omont A., Paul J.F., 1993, A&AS 99, 291 tistical estimates for S, and according to the two- Lucas R., 1991, Continuum and Line Interferometer dimensional case were approximated to S = 0:7 N , Calibration, IRAM document = ( B=2)( N =S) and = B N =S and found to be Lindqvist M., Lucas R., Olofsson H., Omont A., Eriksson K., better than 20% for data with snr 10. Gustafsson B., 1996, A&A 305, L57 Mauersberger R., Guelin M., Martin-Pintado J., Thum C., Cernicharo J., Hein H., Navarro S., 1989, ApJ 79, 217 References Olo son H., Carlstrom U., Eriksson K., Gustafsson B., Willson L.A., 1990, A&A 230, L13 Ake T.B., Johnson H.R., 1988, ApJ 327, 214 Olofsson H., Eriksson K., Gustafsson B., Carlstrom U., 1993, Alcolea J., Bujarrabal V., 1991, A&A 245, 499 ApJS 87, 267 Baars J.W.M., Hooghoudt B.G., Mezger P.G., deJonge M.J., Oudmaijer R.D., Geballe T.R., Waters L.B.F.M., Sahu K.C., 1987, A&A 175, 319 1994, A&A 281, L33 Bates D.M., Watts D.G., 1988, Nonlinear regression analysis Sopka R.J., Olofsson H., Johansson L.E.B., Nguyen Q-Rieu, and its applications. J. Wiley & Sons Zuckerman B., 1989, A&A 210, 78 Bujarrabal V., Alcolea J., 1991, A&A 251, 536 Stanek K.Z., Knapp G.R., Younbg K., Phillips T.G., 1995, Bujarrabal V., Fuente A., Omont A., 1994a, A&A 285, 47 ApJS 100, 169 Bujarrabal V., Alcolea J., Neri R., Grewing M., 1994b, ApJ Tuthill P.G., Hani C.A., Badwin J.E., 1994, in: Very high an436, L169 gular resolution imaging, Robertson J.G. and Tango W.J. Bujarrabal V., Alcolea J., Neri R., Grewing M., 1997, A&A, (eds). Kluwer, p. 395 320, 540 van der Veen W.E.C.J., Habing H.J., 1988, A&A 194, 125 Groenewegen M.A.T., 1994, A&A 290, 207 Young K., 1995, ApJ 445, 872 Guilloteau S., Delannoy J., Downes D., et al., 1992, A&A 262, Zuckerman B., Dyck H.M., 1986, ApJ 304, 394 624 1( OC )0 ! 2( OC 1( )0 2( OC R :021+38 nAd 1( 2( OC 1( )0 2( 1( 2( OC 1( )0 2( OC 1( 2( OC yCg 1( )0 2( 1( 1( )0 2( 2( 91:50 1( 2( 91:54+30 LG ARF 91:54+30 2( 91:5+40 LG ARF 91:5+40 1( 2( 1( 2( 1( 2( 1( 2( 1( 2( 1( 2( 32:16+ LG ARF 32:16+ 1( 2( 32:5+10 LG ARF 32:5+10 1( 2( 1( 32:+65 321+654 1( OC )0 2( 32:49+61 RI

This is a preview of a remote PDF: https://aas.aanda.org/articles/aas/pdf/1998/10/ds1428.pdf

R. Neri, C. Kahane, R. Lucas, V. Bujarrabal, C. Loup. A 12CO () and () atlas of circumstellar envelopes of AGB and post-AGB stars, Astronomy and Astrophysics Supplement Series, 1-64, DOI: 10.1051/aas:1998213