![. FOC F/96 Observations of MCPNe in [O iii]](https://www.researchgate.net/profile/S-Osmer-2/publication/1839864/figure/tbl1/AS:670042102444036@1536761894158/FOC-F-96-Observations-of-MCPNe-in-O-iii_Q320.jpg)
Content uploaded by S. J. Osmer
Author content
All content in this area was uploaded by S. J. Osmer on Apr 22, 2014
Content may be subject to copyright.
arXiv:astro-ph/9806389v1 30 Jun 1998
Hubble Space Telescope Images of Magellanic Cloud Planetary
Nebulae: Data and Correlations across Morphological Classes 1
L. Stanghellini2,3,4
J. C. Blades2
S. J. Osmer2,5
M. J. Barlow6
X.–W. Liu6
1Based on observations made with the NASA/ESA Hubble Space Telescope, obtained at
the Space Telescope Science Institute, which is operated by the Association of universities
for research in Astronomy, Inc., under NASA contract NAS 5-26555
2Space Telescope Science Institute, 3700 San Martin Drive, Baltimore MD 21218, USA
3Affiliated to the Astrophysics Division, Space Science Department of ESA
4on leave, Osservatorio Astronomico di Bologna
5Present address: Department of Computer Sciences, University of Edinburgh, King’s
Buildings, Mayfield Road, Edinburgh EH9 3JZ, Scotland
6Department of Physics and Astronomy, University College London, Gower Street,
London WC1 6BT, UK
– 2 –
Received ; accepted
– 3 –
ABSTRACT
The morphology of planetary nebulae (PNe) provides an essential tool for
understanding their origin and evolution, as it reflects both the dynamics of the
gas ejected during the TP–AGB phase, and the central star energetics. Here we
study the morphology of 27 Magellanic Cloud planetary nebulae (MCPNe) and
present an analysis of their physical characteristics across morphological classes.
Similar studies have been successfully carried out for galactic PNe, but were
compromised by the uncertainty of individual PN distances. We present our
own HST/FOC images of 15 Magellanic Cloud PNe (MCPNe) acquired through
a narrow-band λ5007 [O iii] filter. We use the Richardson–Lucy deconvolution
technique on these pre–COSTAR images to achieve post–COSTAR quality.
Three PNe imaged before and after COSTAR confirm the high reliability of
our deconvolution procedure. We derive morphological classes, dimensions, and
surface photometry for all these PNe. We have combined this sample with
HST/PC1 images of 15 MCPNe, three of which are in common with the FOC
set, acquired by Dopita et al. (1996), to obtain the largest MCPN sample ever
examined from the morphological viewpoint. By using the whole database,
supplemented with published data from the literature, we have analyzed the
properties of the MCPNe and compared them to a typical, complete galactic
sample. Morphology of the MCPNe is then correlated with PN density,
chemistry, and evolution.
Subject headings: Planetary Nebulae: Morphology, Evolution – Magellanic
Clouds
– 4 –
1. Introduction
Planetary Nebulae (PNe) provide a fertile ground for studying the evolution of low
and intermediate mass (M≤8M⊙) stars. The morphology of PNe, when combined with the
physical properties of the nebulae and the central stars (CS’s), help to complete the picture
of how such stars evolve and how the evolution depends on mass, chemical content, and PN
environment. PNe are in fact stellar envelopes ejected in advanced evolutionary stages, and
carry a wealth of information on previous phases.
The morphology of PNe, as observed through narrow–band filters, traces the structure
of the ejected gas, and contains information on the time interval between ejection and
observation, in addition to the nature of the ejection itself; the final ionized gas shape
contains information on inhomogeneities during ejection. Morphological characteristics
change with both the nebular and stellar evolution, thus they carry a record of the space
and time history between the ejection and the observation. The ejecta can be perturbed,
for instance, by a fast CS wind, stellar companions, planets in the central star’s system,
interstellar medium condensations, magnetic fields, and by changes in the post–ejection
stellar evolutionary paths.
To date, several ground–based surveys of galactic PNe aimed at delineating their
morphological characteristics have been completed (Schwarz, Corradi, & Melnick 1992;
Manchado et al. 1996; Chu, Jacoby, & Arendt 1987; Balick 1987). A space–based survey of
galactic PNe has been performed by Bond et al. (1995) with the WFPC2 camera on board
the HST. Shapes of planetary nebulae have been carefully classified and cross–correlated
with nebular and stellar properties, obtaining a series of interesting results, ranging from
the segregation of PNe hosting different types of central stars based on their morphology
(Calvet & Peimbert 1983; Stanghellini, Corradi, & Schwarz 1993) to the indication that
bipolar PNe have more massive progenitors than elliptical PNe (Stanghellini et al. 1993;
– 5 –
Corradi & Schwarz 1994). However, such results need also to be tested in a distance–bias
free environment.
Only a handful of galactic PNe have individually determined distances, while the
majority have distances derived with a statistical method, based on physical assumptions,
such as, for example, that all optically thin PNe in the Galaxy share the same ionized
mass (Cahn, Kaler, & Stanghellini 1992; Kingsburgh & Barlow 1992). Not only are
the absolute stellar and nebular parameter determinations at risk when using poorly
determined distances, but even the morphologies themselves could be misclassified when
compared to one another and thought to be, for example, at the same distance from us,
since the detectability of morphological details obviously decreases with distance. The
proximity of the Magellanic Clouds have made them perfect target galaxies to study PNe
in a distance–bias free environment. However, use of the cameras aboard HST is required
to resolve the shape of the Magellanic Cloud PNe (MCPNe). The spatial resolution of
HST/FOC allows observation of MCPN morphology with similar definition to typical
galactic PNe at a distance of ≈1.5 kpc observed from the ground with 1 arcsec seeing.
Only nineteen narrow–band HST frames of MCPNe have been published to date,
either imaged with the PC1 (Dopita et al. 1996, hereafter D96) or the Faint Object Camera
(Blades et al. 1992, hereafter B92). All published data were acquired in the pre–COSTAR,
pre–first refurbishing mission epoch (<1994). The B92 paper is an essay on what can be
achieved with HST/FOC when observing MCPNe through the Hβand [O iii] narrow–band
filters. B92 showed, for the first time, spatially resolved images of four extragalactic
PNe and their morphological details. The target nebulae were chosen to be bright, so
as to trace early post–AGB evolution. Liu et al. (1995) used B92’s images to construct
detailed photoionization models for two of these nebulae, SMC N 2 and N 5, derived
nebular ionized masses and central star masses, and compared nebular expansion ages
– 6 –
with central star evolutionary track ages. The main aim of the GO observing program of
D96 was to image, through the narrow–band [O iii] filter, a number of MCPNe covering
a large domain in nebular parameter space, such as [O iii] luminosity, luminosity class
and optical thickness. Their published results included the narrow–band images and the
study of the expansion velocities and dynamical ages, taking into account the nebular
inclination on the plane of the sky in the case of non–symmetric PNe. The resulting
nebular evolution was then coupled with a study of the evolutionary status of the central
stars, by means of the logT–logL plane location. Dynamical ages and evolutionary times
were found not to follow a simple correlation if the evolutionary times were calculated on
the basis of hydrogen-burning; rather, the authors suggested in D96 that nebulae hosting
helium–burning CS’s outnumbered by 2:1 those hosting H-burning CS’s.
The main purpose of the present paper is to present previously unpublished narrow–
band images of MCPNe, taken with the FOC as part of the original FOC Investigation
Definition Team science program. Most of the images presented here were acquired before
COSTAR was installed on the HST. However, three MCPNe were re–observed following
the first servicing mission, using the FOC and COSTAR to check on the reliability of the
original images and the veracity of the deconvolution method used in their analysis. We
present the data as follows. In Section 2 we discuss the observations obtained with the
FOC, including target–selection criteria, scheduling of the observations, data reduction
and calibration, with emphasis on the deconvolution method. Results from the FOC
observations are presented in Section 3 for the complete HST/FOC data set, i.e., the newly
observed PNe and the ones published by B92; we discuss [O iii] images only. Additionally,
we have broadened the MCPN database by reclassifying the PC1 images from D96 with the
same morphological scheme used for the FOC data. So, in Section 3 we show (a) the final
deconvolved FOC frames, (b) the morphological classification, (c) the MCPN dimensions,
(d) the photometry, and (e) the expansion velocity and the dynamical ages, and discuss the
– 7 –
results in Section 4. The conclusions, and a discussion of possible future developments, are
in Section 5.
2. HST/FOC observations and reductions
17 MCPNe have been observed since June 1991 over 4 HST observing cycles, using the
high resolution f/96 optical chain of the Faint Object Camera (Macchetto et al. 1991). The
observations concentrated on using the narrow-band F501N filter to record the [O iii] lines,
while a few of the objects were also recorded in the narrow-band Hβfilter F486N. The Hβ
recombination line traces emission by the dominant element in the nebula, while the [O iii]
λ5007 transition is a strong, collisionally excited, cooling line which traces emission by the
usually dominant O++ ion of oxygen. The systemic radial velocities of the Small and Large
Magellanic Clouds shift the [O iii] lines by between +2 and +5 ˚
A relative to their rest
wavelengths, putting the λ5007 member of the [O iii] doublet at the peak of the filter’s
transmission. With a total bandpass of 74 ˚
A, the weaker member of the doublet at 4959 ˚
A
will not be transmitted through the F501N filter, according to the transmission curve given
in the current FOC Handbook (Nota et al. 1994).
This program was originally planned before the launch of HST, and the targets were
chosen to be easily detected through the selected filters in reasonable exposure times and
without any image saturation. The nebulae were thus chosen to be reasonably bright in
the λ5007 [O iii] line. In order to maximize the probability that the chosen nebulae were
optically thin, and thus that the ionized masses would be equal to the true nebular masses,
two additional selection criteria were used: (a) the nebulae should have a detectable λ4686
He ii line, a standard criterion for selecting out low– and medium–excitation PNe with
younger and less evolved central stars (Sanduleak, MacConnell, & Philip 1978); and (b) the
nebulae should have [O ii] electron densities less than about 5000 to 6000 cm−3. The optical
– 8 –
spectroscopic study of Barlow (1987) had indicated this to be the dividing point between
optically thick and optically thin Magellanic Cloud PNe. The analysis by Liu et al. (1995)
subsequently confirmed that SMC N 2 with ne(O ii) = 2850 cm−3and 3727/Hβ= 0.29
is indeed optically thin but they found that SMC N 5 with ne(O i i) = 3890 cm−3and
3727/Hβ= 0.79 is still optically thick.
Most of the observations were obtained in the observing cycles before the 1993
December servicing mission which repaired the imaging capability, although three LMC
objects, N 66, N 97, and N 192, have since been re-observed. As we shall explain, they
provide an important check on the analysis of the earlier data. Because of its intriguing
morphology, LMC N 66 (SMP 83) was observed on two separate occasions before the
servicing mission as well as once afterwards. Our monitoring of this object turned out
to be fortuitous because of the recently announced variability in brightness of the central
star (Pena et al. 1994). Although we use the image of N 66 to check pre–COSTAR
deconvolution performance, we do not include this object in the discussion nor in Tables
2 and 3; Vassiliadis (1996) has discussed and interpreted the ensemble of HST FOC and
WFPC images of N 66 obtained between 1991 July and 1994 February. Table 1 gives a
complete record of the FOC observations in the [O iii]λ5007 light. All observations were
taken in the 512×512 pixel format with 25-µm square pixels, corresponding on the sky to
a plate scale of 0.0223 arcsec per pixel before the servicing mission and 0.0144 arcsec per
pixel with the Corrective Optics Space Telescope Axial Replacement (COSTAR) in place
(Jedrzejewski et al. 1994).
The observations taken after the 1993 December servicing mission were obtained very
early in the observing cycle (January and February 1994) and before the correct instrument
sensitivities had been determined. Inappropriately, pre–servicing values were in use at that
time, resulting in an incorrect value for the keyword PHOTLAM being attached to the
– 9 –
data in the calibration pipeline. Unfortunately, all FOC data taken around that time have
been archived with the incorrect value. (This problem was corrected in the FOC calibration
pipeline on 19 April 1994.) We have rectified our Cycle 4 data with the proper values. In
order to guard against the risk of saturation to the post servicing mission data we obtained
a pair of F501N images for N 97 and N 192 by taking a second image with a 1 magnitude
neutral density filter, see Table 1.
There are two calibrations that are applied to the raw FOC data in the ground–system
calibration pipeline, namely a geometric distortion correction followed by a relative
calibration or flat field correction. For the data taken in the early cycles we reprocessed
the observations using the IRAF/STDAS task called CALFOC using new calibration files
as they became available, and we have reprocessed all the data originally presented in B92.
Finally, we removed by interpolation the effects of the numerous reseaux marks, which are
fiducial reference marks engraved on the detector faceplate.
Our earlier analysis (B92) had shown that deconvolution techniques could be employed
to improve the qualitative appearance of these compact and high contrast objects. This
result encouraged us to continue taking observations throughout the early cycles rather
than waiting for the 1993 servicing mission. Accordingly, we spent considerable effort in the
reduction phase to try and optimize the deconvolution and to investigate how different point
spread functions (PSFs) and the telescope focus were likely to be affecting the accuracy of
our data. Subsequently, the three post–COSTAR images were of great value in providing a
direct comparison between the data sets and in vindicating our approach.
For deconvolution we used the non–linear restoration technique of Richardson (1972)
and Lucy (1974) which has been installed in the IRAF/STSDAS software package, and
we performed testing on the optimum number of iterations for our images. In qualitative
terms, we found that after 50 iterations the images showed considerable improvement,
– 10 –
through reduction of the surrounding halo or skirt of scattered light while still retaining
the basic structure that could be seen in the unprocessed images. After a larger number
of iterations (100) the shape and form of the objects began to break down and the images
became artificially lumpy and pixilated. We chose 50 iterations for all the pre–COSTAR
observations. The same number of iterations was found by Dopita and collaborators to be
the best for deconvolving WF/PC images too (Dopita 1998, private communication).
The images obtained in 1994 with COSTAR of LMC N 97 and N 192 were valuable
in establishing the veracity of the deconvolution work. In Figure 1 we show contour
plots of the deconvolved pre–COSTAR images of these two PNe with the more recent
images obtained with FOC and the COSTAR. There is excellent agreement between the
pre–COSTAR deconvolved images and the post–COSTAR direct images. The agreement
covers the overall size and shapes of the nebulae and extends to the smallest structures
that can be discerned at a scale of 0.07 to 0.1 arcsec. The consistency can also be seen in
the images that are presented later (see Figure 5, panels 1 and j, and m and n), as well as
in the good photometric agreement (see §3). The direct comparison provides confidence
that the deconvolution technique has improved the qualitative appearance of these objects
without introducing artifacts.
Routine monitoring of the image quality of HST has been carried out since launch in
order to monitor and maintain good telescope focus as well as to characterize the features of
the optical performance (Hasan & Burrows 1994). During this time, the Optical Telescope
Assembly (OTA) continued a steady contraction as gas desorbed out of the telescope
structures, thus requiring frequent re–alignments to retain the focus within 10 microns
of the nominal value, and this was not always achieved. In addition, short time period
fluctuations of the OTA PSF were discovered (Hasan & Bely 1993) which are attributed
to expansion and contraction of the secondary mirror support system causing small (≈5
– 11 –
microns) motions of the secondary mirror (breathing). Also, the internal focus of the FOC
was optimized on occasions. It is impossible to unravel the effects of these optical changes
from our observations because simultaneous PSF observations were not obtained. (Indeed,
it would have been time consuming to have attempted to calibrate these defocusing effects,
especially the breathing which can alter the PSF by small amounts over short time periods
of about 30 minutes.)
We were concerned that these image degradation problems could affect the resolution
of our data and we ran tests to see how sensitive our deconvolution results were to different
input PSFs. We experimented with a variety of Richardson–Lucy deconvolutions. Observed
PSFs were obtained from FOC calibration observations of BPM 16274 through the F501N
filter and these observations provided two PSFs, one based on observations from early in
Cycle 1 and one from Cycle 3. A search of the FOC archive yielded a third PSF star from
April 1992 observations of the SN 1987A field. Finally, we constructed a theoretical PSF
using the optical modeling work of Krist (1993). An obvious advantage of the theoretical
PSF compared with any of the observed PSFs is the infinite signal–to–noise ratio. Figure 2
shows the four PSFs described above and Figure 3 shows the results from deconvolving the
pre–COSTAR image of N 192 with each of these PSFs.
As inspection of Figure 3 confirms that, at the level to which we are working, the
Richardson–Lucy deconvolution is not very sensitive to the input PSF. Both the overall
shape and most of the small scale structures remain the same over all four images. On
closer examination there are subtle changes from one deconvolution to another, and these
changes provide an empirical assessment of the overall accuracy of the deconvolution work.
In general, high signal–to–noise ratio PSFs gave the best results (higher counts in the peak
of the PN image). This seemed a more important parameter than the closeness in time of
the PSF to the actual observation. In other words, for our data, a PSF produced from an
– 12 –
observation in 1991 worked at least as well as a PSF produced in a 1993. Probably the
breathing phenomenon is destroying any advantage the contemporary PSF may otherwise
have had.
The theoretical PSF can be adjusted to fit precisely any observation by matching the
Airy ring pattern using a stellar image in the field of the PN. Potentially, this could allow
correction for the breathing phenomena. Unfortunately, among our images only that of
N 66 has a star (in fact the central star) suitable for such matching. Figure 4 shows the
results of matching and not matching the theoretical PSF to the central star. Qualitatively
it is difficult to discern the difference between the two; however, the peak counts in the
image deconvolved with the matched Airy rings are 20 percent higher, indicating that an
improvement is achieved with this method. On the other hand, we found that use of the
theoretical PSF tends to yield a rather uneven and blotchy appearance to the image which
the observed PSFs do not. In any case, the lack of available stars in the other PN images
prevents us from using this approach for cases other than N 66. After considerable testing
we selected the 1991 observed PSF to use for all our pre–COSTAR denconvolution work.
A particular concern that we had was whether faint extended halos that might exist
around the main nebular structures would be recovered by the deconvolution process, since
such halos could potentially contain a significant fraction of the total nebular mass. For
example, a faint halo three times larger than that of the main inner structure, and with a
mean surface brightness of only 1.5% that at the central peak, would still contain as much
mass as in the inner bright nebula. We therefore experimented by artificially adding smooth
halos of varying surface brightnesses and diameters to the deconvolved FOC [O iii] nebular
images of SMC N2, SMC N5 and LMC N192. These composite images were then convolved
with an observed PSF and the resulting convolved images were deconvolved with the
Richardson-Lucy algorithm in the standard way. It was found that the artificial faint halos
– 13 –
were recovered in the deconvolved images in all cases, down to halo surface brightness levels
of 1% of the peak inner emission. We are therefore confident that any halo emission around
the brighter PNe in the sample must be below this level, although for the fainter PNe in
the sample the upper limits to any surrounding halos would be correspondingly larger.
Our re-observations of four of the nebulae, after the installation of the COSTAR corrective
optics, revealed no extended low surface brightness emission around them, confirming this
conclusion in their cases.
3. Image analysis
3.1. The FOC and PC1 data sets
Our discussion of PN morphologies, dynamical expansion times, and luminosities is
based on two data sets. The first data set is composed of the FOC images presented here
for the first time and the images illustrated by B92 (FOC set); the other set consists of the
PC1 images described by D96 (PC1 set). In all we have 27 MCPNe, excluding N 66 (see
§2). The list of observed PNe can be found, respectively, in Table 1 of this paper for the
FOC set, and in D96’s Table 1 for the PC1 set. The two sets have three objects in common,
useful to check the criteria for morphological classification, and image quality. Each set has
been selected with a defined criterion: the FOC set contains PNe with high [O iii] fluxes
and generally low optical thickness, while PC1 set presents a variety of [O iii] fluxes and
Lyman continuum optical depths. As a consequence, the two sets are not homogeneous,
and the final composite sample is not, by any means, a complete or unbiased statistical
sample of MCPNe. Nonetheless, there is purpose to analysing the composite group of PNe
in a qualitative way, in order to establish morphological trends.
– 14 –
3.2. Morphology and diameters
The λ5007 [O iii] narrow–band FOC images are presented in Figure 5. The
pre–COSTAR images are deconvolved, as discussed in Section 2. In Figure 5 we also
include three nebulae already published by B92, and we show both the pre–COSTAR
(deconvolved) and post–COSTAR images for N 97 and N 192. The following discussion, on
the morphology, the dimensions, and the photometry of the PNe observed with the FOC,
is based on the [O iii] images. To classify the morphology we follow the most recent and
widely used scheme by Schwarz, Corradi, and Stanghellini (1992) in its updated version
(Manchado et al. 1996). Originally, this classification scheme was conceived for Hα(or
Hβ) images, expecting this emission line to track the bulk of ionized gas in most PNe. In
the case of the FOC set, whose PNe have high excitation, the morphological differences
between high–and medium–excitation plasma tracers are not expected to be significant (for
galactic equivalents, browse through high–excitation PN images in the catalog by Manchado
et al. 1996). In the case of the PC1 set, however, the lower excitation might not be fully
delineated by the [O iii] line. Another source of inhomogeneity among the two data sets
is the different angular resolution of the two cameras used in the observations. The PC1
frames published by D96 clearly show their lower angular resolution with respect to the
FOC images published here for the first time, or by B92.
The classification scheme sorts PNe into five main groups, as defined by the outer
envelope of the PNe: Round PNe (R), Elliptical PNe (E), Bipolar PNe(B), Quadrupolar
PNe (Q), and Pointsymmetric PNe (P). Bipolar PNe are nebulae with one axis of symmetry,
and a detectable waist. Quadrupolar PNe consist of two pairs of bipolar lobes, joined at a
common waist. Pointsymmetric PNe show structures that are symmetric with respect to a
central point (in 2D). The scheme by Manchado et al. (1996) does not include “irregular
PNe”, although it includes the possibility that a PN could not be classified within the
– 15 –
above scheme, and in this case they are denoted as NC. The main classes have subclasses,
denoted by suffixes attached to the morphological main symbols. The subclasses describe
inner structures (s), multiple shells or haloes (m), ansae attached to the main structures
(a), and rings at the waist of some bipolar nebulae (r).
We can apply the morphological classification scheme to Magellanic Cloud PNe,
although we should keep in mind that the [O iii] images of these nebulae will track the
bright cores rather than outer features, such as multiple shells and bipolar/quadrupolar
lobes. In this sense, real bipolar structures may not be observed in their complete display
of lobes, rather only the inner ring may be visible. As discussed in Section 2, our analysis
of the nebulae confirmed the absence of lobes. We find instead a considerable subgroup of
objects whose outer shape is elliptical, and whose inner shape is “bipolar”, showing two
concentrations of surface brightness. Such structure is reminiscent of a projected inner ring,
that, in turn, is typical of bipolar outflows. From the asymmetry of the ring like structures
we can be quite confident that their true morphology is bipolar rather than elliptical. We
classify these PNe as bipolars (B), subclass bipolar core (bc).
Table 2 gives the morphological classification of PNe. In columns (1) and (2) we give
the discovery name and SMP (Sanduleak et al. 1978) catalog number. Column (3) gives
the HST camera used for the observation, where FOC indicates the PN whose images are
published in this paper for the first time, or by B92, and PC1 indicates the D96 images.
Column (4) defines the morphological class, and the detailed classification in parenthesis.
In two particular cases, the shape is incomplete, and we define that particular shape as the
suffix ‘inc’. In column (5) and (6) of Table 2 we give the angular and physical diameters of
the PNe, respectively in arcsec and parsecs.
Our diameter measurements are a result of a detailed photometric routine, as we
describe following. First, we define a geometrical center of the PN on the image, by hand.
– 16 –
Second, we chose for each object a set of (circular) apertures that segment the nebula into
anuli. The outermost of these apertures is set at a large distance from the apparent nebular
limb. Then we measure the flux in each aperture with the IRAF/PHOT routine, obtaining
the sky–subtracted total flux within each aperture. Going outward from the center, we find
the maximum total nebular flux. We then plot the relative encircled flux (flux within each
aperture divided by the total flux) versus aperture for each nebula, and we read out the
aperture encircling 85 % of the total flux. We define the latter aperture to be the physical
nebular radius. This procedure has been repeated for pre– and post–COSTAR images for
N 97 and N 192, obtaining satisfactory agreement.
Among the 15 diameters measured with the photometric method by us, six objects
show the presence of a ring in the relative encircled flux profile: N 4 (Bbc), L 305 (Es),
L 536 (E), L 343 (Bbc), N 18 (Bbc), and N 67 (Bbc). This photometric check is a good
method to determine which PNe actually show ring–like features. By using the above
described method, we had difficulties in finding the outer contour of the nebula N 24,
or the size at which its encircled flux become constant. This planetary has a halo/core
structure, with FWHB of the bright core measuring 0.23 arcsec. We thus do not give is
outer dimensions in Table 2, and we eliminate this object in the discussion of the results
and in Figures 6 through 10. Future observations of this particular object are in order.
Our definition of physical radius is, in principle, the same as D96’s, and has been
chosen this way for uniformity among the two data sets. Nonetheless, of the three objects in
common among the two sets, only for one (WS 12) do the two measurements agree. In the
cases of N 4 and L 536, both with ring–like profiles, the measurements are different. The
difference can be ascribed to the power–law skirts produced by incomplete deconvolution of
the PC1 set, and adds an extra uncertainty to our discussion. For the PC1 set we derived
the angular sizes from the published physical sizes (see Table 3 in D96) and the distances
– 17 –
to the Clouds quoted therein (dLMC=50.60 and dSMC=58.29 kpc). Obviously the same
distances to the Clouds have been used to derive physical sizes for the FOC set.
Below, we explore in some detail the individual morphologies of those nebulae whose
structures are not spherically symmetric. For PNe showing an asymmetric structure, we
estimate the projection angle on the plane of the sky, measured from the ratio of the
semiminor to the semimajor axes of the ring–like structures, and we have included this axial
ratios, q, in Table 3.
N 2 is a regular ellipse with an inner hole.
N 4 has an elliptical, boxy shape with evidence of an inner, projected ring. D96 classified
it as BR (bipolar/ring), which, apart from the different terminology, corresponds to our
definition.
N 5 is almost round, with an inner hole.
N 18 has a fairly round outer shell, and an inner, edge–on ring.
L 305 is elliptical; contour levels show a marked asymmetric structure in the inner parts,
as if the maximum brightness was off–center.
N 67 resembles a ring feature of a galactic bipolar PN (e. g. NGC 650, Sh 1–89, Manchado
et al. 1996); the measured inclination of the ring is approximately 45 degrees on the plane
of the sky.
L 343 is elliptical with a ring–like core.
L 536 is elliptical, with low ellipticity. D96 defined it as s(spherical), while we can actually
measure an axial ratio of about 0.8.
LM2–5 is elliptical, with an asymmetric ring–like core.
– 18 –
N 97 shows four density enhancements, and can be classified as quadrupolar.
N 24 has a very slightly elliptical outer shell, and a regular round inner shell, and can be
classified as R.
N 192 has a slightly elliptical outer shell, an irregular inner structure, and presents an
inner hole. It is classified as R.
WS 12 has elliptical contours, with an incomplete crescent–shape structure. Although
D96 classified it as BR, we could not definitely see the complete ring with our FOC image
brightness analysis.
WS 16 is a ring–like structure, of elliptical contour; we classify it as Es.
LM1–27 has an irregular inner structure, reminiscent of an incomplete ring.
N 122: although D96 found a bipolarity on the deconvolved image, we could not find it
on their raw image, which, on the contrary, shows a genuine elliptical PN, with very high
ellipticity.
N 52 is elliptical with an inner hole.
LMC SMP 72 is very hard to classify. At a first glance it could resemble a quadrupolar,
but a careful analysis shows no evidence for the second pair of rings. It can be a bipolar
with an enhanced, large ring.
N 60: our morphological classification confirms that of D96, of shperical/round shape.
N 215 is elliptical with a bipolar core.
LMC SMP 96 is elliptical with a bipolar core.
LM1–61 is round, with irregular inner brightness.
– 19 –
3.3. Aperture photometry of FOC images
Aperture photometry has been performed for the FOC images. Calibrated, but
non–deconvolved, images were used to this end. Even if most PNe are easily contained in
a 2×2 arcsec2aperture, we chose an aperture of 6 arcsec2, since pre–COSTAR images may
contain considerable energy output out to 4 arcsec from the target centers. We measured
the total counts per second within the aperture, after sky subtraction (Tab. 1, col. [8]), the
peak counts per second (Tab. 1, col. [9]) and the calibrated physical fluxes (Tab. 1, col. [10],
in erg cm−2s−1). The derived FOC [O iii]λ5007 line fluxes show excellent agreement with
the ground-based values measured by Jacoby, Walker, & Ciardullo (1990). For 21 FOC
measurements, the mean flux difference is found to be just 0.00±0.02 dex.
3.4. Expansion velocities and dynamical expansion ages
In order to evaluate the dynamical expansion ages of our PNe we need their expansion
velocities. We have used velocities based on measurements published by Dopita et al. (1985
for SMC, 1988 for LMC PNe). We should beware that Dopita et al. define the expansion
velocity as 0.911 times the FWHM of the λ5007 [O iii] line, corrected for instrumental and
thermal broadening, whereas in general (e.g. in the galactic PN expansion velocity catalogs
of Sabbadin 1984 and Weinberger 1989) the value vexp=0.50 FWHM is used for unresolved
nebulae7. We use this second choice for the expansion velocity, and corrected the velocities
of Dopita et al. as if they have been measured in this way, thus dividing them by 1.82. The
7Robinson, Reay & Atherton (1982) have shown theoretically that the FWHM linewidth
of a nebula completely enclosed by an observing aperture is equal to the line splitting
that would be observed at the nebular center in a spatially resolved observation. Munch,
Hipplelein, & Pitz (1984) have confirmed this result observationally
– 20 –
resulting nebular expansion velocities are given in Table 3, column (3) 8.
Column 4 of Table 3 lists τdyn = Rneb /vexp, the dynamical expansion ages derived from
the PN radii and expansion velocities (where Rneb is half the diameter D listed in col. 6
of Table 2). Table 3 (col. 2) also lists q, the measured ratio of the nebular semi-major to
semi-minor axes. D96 made use of this parameter to correct dynamical expansion times
for nebular inclination effects, assuming that ring-shaped nebulae are circles viewed at
an inclination angle θ= cos−1q with respect to the plane of the sky, so that measured
expansion velocities should be corrected for inclination effects by dividing them by sin
θ. However, since this would yield infinite expansion velocities and zero expansion ages
for q = 1, no correction was made for circular nebulae. We found that the use of this
scheme led to large decreases in the derived expansion ages for nearly-circular nebulae
(e.g. a factor of 2.5 for SMC N2, with q = 0.92), versus no correction at all for perfectly
circular nebulae (e.g. SMC N5, q = 1.0) and so decided not to make such a correction. We
note that for non-circular nebulae the nebular radius defined by the 85% encircled energy
definition is in any case a mean of the semi-major and semi-minor axis dimensions, so that
its use yields dynamical ages that are smaller than those that would be obtained just from
8The FWHM of a Gaussian line profile contains 76% of its total flux. Dopita et al. (1985)
defined the expansion velocity as the half-width at one tenth maximum line intensity. For
a Gaussian, the full width corresponding to this definition contains 97% of the total line
flux. For comparison, the nebular diameter definition adopted by D96 and by ourslves is
the diameter encircling 85% of the total nebular flux. For a Gaussian line profile, 85% of
the total line flux is contained with 0.4 maximum line intensity. We prefer to adopt here
the usual definition of vexp = 0.5 FWHM, but if expansion velocities coresponding to the
half width at 0.4 maximum line intensity are preferred, then the derived expansion ages in
Table 3 should be decreased by a factor of 1.15.
– 21 –
the semi-major axis dimensions. We note that barrel-shaped nebulae can yield apparent
circular shapes when viewed pole-on, and elliptical shapes when viewed equator-on. Figs. 4
and 5 of Frank & Mellema (1994) show that for such nebulae viewed pole-on the measured
expansion velocity corresponds to material along the line of sight that is expanding in the
polar (longer axis) direction, with a velocity higher than that in the equatorial direction.
Thus dynamical ages for apparently near-circular nebulae of this type may therefore be
underestimated, since they could be using too high an expansion velocity.
4. Analysis of the results
The morphologies of the 27 MCPNe in the [O iii] narrow band images are similar to
those of galactic PNe, if we consider the bright parts of the latter ones. We did not find
multiple shell PNe or faint extended lobes of bipolar and quadrupolar PNe. Similarly to
galactic PNe, we encounter round, elliptical, bipolar (ring), and quadrupolar shapes. We
did not expect that the statistical distribution among our group of MCPNe would be the
same as for galactic PNe, since we have overall selected against faint and low excitation
PNe, thus against symmetric shapes (Stanghellini et al. 1993). We found that 36% of the
studied MCPNe are round, 32% are elliptical, and 32% have bipolar or quadrupolar shapes.
The northern galactic sample (Manchado et al. 1996) has 24% round, 56% elliptical, 17%
bipolar and quadrupolar, and 3% pointsymmetric PNe. We thus confirm the existence
of three main morphological classes, round, elliptical, and bipolar/quadrupolar PNe. We
did not find pointsymmetric PNe, nor did we expect them, given the low percentage of
occurrence of this particular morphology among galactic PNe. We confirm that more
bipolar PNe can be found among high excitation objects, as was already inferred from
Zanstra analysis by Stanghellini et al. (1993). Our statistical analysis cannot proceed any
further, given that we do not have a statistically significant sample.
– 22 –
The main advantage of studying MCPNe with respect to their galactic counterparts
resides in knowing their distances. Distance–dependent physical properties, such as physical
dimensions, dynamical times, and luminosities, are readily determined for MCPNe. When
we discuss dynamical times derived from physical dimensions, we should not overlook the
fact that some nebulae are optically thick to the ionizing radiation from the central stars. If
a PN should remain optically thick for most of its evolution, its measured diameter would
not trace the dynamical evolution, but rather the evolution of the ionization front. We have
sorted our PNe according to their optical thickness, as derived from the line ratio λ3727
[O ii] / Hβ. As Kaler & Jacoby (1990) pointed out, this ratio should be higher than 0.8 and
0.35 for, respectively, LMC and SMC PNe to be optically thick. We derive the diagnostic
ratio from spectral line intensities available in the literature (Meatheringham & Dopita
1991ab; Vassiliadis et al. 1992), and report the optical thickness in Table 2, Column (7).
This measure of thickness is rather crude, in that it does not take into account variations
of the diagnostic ratio with density, thus a small fraction of PNe labeled as thin in Table 2
might be in fact thick. The results of Table 2 agree for the most part with Dopita and
Meatheringham’s (1991ab) photoionization models optical thickness, which we do not use
therein to avoid model dependence.
Among those PNe whose diagnostic spectral lines are available in the literature, we
find that (a) about half the elliptical PNe are optically thin, (b) most round PNe are
optically thin according to the above criterion, and (c) only one asymmetric (bipolar) PN
is optically thin. Obviously the fact that the majority of PNe in the FOC set are optically
thin to the ionizing radiation strongly depends on the target selection of those planetaries,
but the thickness/thinness of each morphological class was not selected a priori. Since most
bipolar/quadrupolar PNe are thick to ionizing radiation, their measured physical size can
be an underestimate of the real size, and the dynamical time could be actually larger than
calculated.
– 23 –
In Figure 6 we plot the histogram distributions of three main nebular properties:
physical dimensions (top), expansion velocities (middle), and dynamical expansion ages
(bottom). Each morphological class is represented in a different way (see caption). we infer
the following properties: (a) bipolar PNe have dimensions larger than 0.2 pc, this result,
although based on very few objects, is an important confirmation of a similar situation
existing for galactic PNe (Stanghellini 1995); (b) bipolar PNe in our sample have physical
dimensions within a narrower range than elliptical and round PNe.
In Figure 7 we examine the time evolution of the PN sizes for three major morphological
classes: round, elliptical, and asymmetric (bipolar and quadrupolar) PNe. We did not
include those PNe whose angular size is a measured upper limit (see D96). The physical
dimensions correlate linearly with the dynamical ages, as expected, with scatter due to the
velocity distribution. In particular, elliptical and bipolar/quadrupolar PNe define a very
tight correlation, with coefficient Rxy=0.92.
We can use the physical dimensions and dynamical age as independent variables to
reveal correlations with other physical parameters across morphological classes. Due to the
limited size of our sample and the selection criteria of the targets, the range of physical
diameters is rather restricted, and each morphological class is not statistically represented.
When using the dynamical age as an indication of the evolutionary timescale, we should
not overlook the fact that it merely indicates the time lapse between the envelope ejection
at the TP–AGB phase and the observing time, and assumes a constant expansion velocity
without acceleration (or deceleration) or the shell. τdyn is a very useful variable for order of
magnitude correlations, but it does not indicate the exact lifetime of a PN. Furthermore,
since zero age post–AGB tracks generally correspond to a defined central star temperature,
one can really never finely tune these tracks to the observed dynamical times, and a direct
comparison among the two sets of parameters, the theoretical ones and the empirical,
– 24 –
should not be used without precautions (K¨aufl, Renzini, & Stanghellini 1993). On the other
hand, dynamical ages measured for MCPNe are generally more homogeneous than those
measured for galactic PNe since their distances are better known and the dimensions and
the velocities of the MCPNe both correspond to the high excitation body of the PN.
Electron densities, measured from forbidden line ratios, have been plotted in Figure
8 against the physical dimensions of the MCPNe. The general trend shows a decreasing
electron density with increasing physical size, with the exception of L 305 and N 67, whose
loci are in the upper–right part of the diagram.
In order to study the fading of PNe with evolution, accordingly to their shapes, we
have analyzed the [O iii] surface brightness. The [O iii] luminosities from which we derive
the surface brightness have been calculated from the total fluxes observed form the ground
(Jacoby et al. 1990), the Cloud distances, and the extinction constant (Boffi & Stanghellini
1994, and references therein). The correction for extinction has been performed using the
galactic extinction curve (Osterbrock 1989), which around 5007 ˚
A has a similar shape to
the curve derived for the Magellanic Clouds (Hoyle & Wickramasinghe 1991). Figure 9
aims at disclosing possible evolutionary effects on the surface brightness for PNe of different
shapes. The [O iii] luminosity depends on the stellar energetics, and secondarily on the
oxygen content and on the effects of nebular evolution (Richer 1993). It is thus a good
guide for tracing the intrinsic stellar luminosity.
What we see in plot 9 is that the round PNe (symbols are as in the other Figures)
are not to be found at low surface brightness, as opposed to elliptical/bipolar/quadrupolar
PNe. One reason for the split in fading behaviors could certainly be a difference in the
ionized masses, which, in turn, could be an indication for a difference in the mass of the
progenitors (see Fig. 8 in Boffi & Stanghellini, 1994). On the other hand, a systematic
difference of velocity fields among the round PNe and all other shapes can also produce
– 25 –
such a separation as in Figure 9. Indeed, the relation between the axial ratio q and vexp
shows that extreme asymmetric PNe do evolve faster. But the velocity difference alone
does not explain the discrimination among morphological types of Figure 9. Unfortunately,
given the small size of the sample, and that most of the PNe in the Figure are optically
thick to ionizing radiation, we can not conclude that we are observing two groups of PNe
with different progenitor masses. What we are probably seeing here is that more massive
stars evolve faster through the high luminosity post–AGB phase, and they are fading at
the time of observation. On the other hand, stars with low–mass progenitors evolve slowly,
thus retaining their high luminosity for a longer time. Should this interpretation be right,
we can conclude that most bipolar/quadrupolar and elliptical PNe in our sample have
high mass progenitors, while most round PNe have low–mass progenitors, in addition to
lower velocities. Only with a much larger and homogeneous sample of MCPNe we could
investigate this important aspect of PN evolution to its fullness. Accurate modeling of the
surface brightness evolution of expanding shell and ring PNe are also required to provide
the necessary background to complete the picture.
In order to confirm the nature of the asymmetric PNe in our sample, and to test
the correlation between morphology and chemical enrichment, as found by Peimbert and
collaborators (e.g., Calvet & Peimbert 1983, Peimbert & Torres-Peimbert 1983), in Figure
10 we plot the N/O abundance ratio against dynamical expansion time for all PNe for which
the N/O ratio is available (Richer 1993). Symbols are for the different nebular shapes, as in
the other Figures. Since the (revised) N/O abundance ratio constraints for Type I PNe are
different for SMC, LMC and Galactic PNe (Kingsburgh & Barlow 1994, Peimbert 1997), we
have artificially decreased the N/O abundance ratio of SMC PNe by an appropriate factor,
so that they can be directly compared (∆ log N/O=0.24). The horizontal line in Fig. 10 is
at the appropriate level so that PNe above the line are of Type I (see note aon Table 2
for PNe identification). We find that all round PNe in our sample are non–Type I, all but
– 26 –
one bipolars are Type I, and elliptical PNe are equally divided among the two Peimbert
Types. We do not see any evolution in the N/O abundance. The number of objects is so
low to leave the possible consequences unexplored for now. What we can infer from the last
two Figures is that round and bipolar/quadrupolar PNe form two distinct “enrichment”
groups, while more investigation is necessary to determine whether the ellipticals are an
intermediate sequence or a different evolutionary stage of either round or bipolar nebulae.
5. Conclusions
We have presented a set of narrow–band images of 15 Magellanic Cloud Planetary
Nebulae acquired with the FOC, aboard the Hubble Space Telescope. Deconvolution
techniques, and comparison to post–COSTAR FOC images of three of the PNe, show that
excellent image quality can be achieved from pre–COSTAR images. We have measured
the nebular angular diameters, allowing the calculation of dynamical expansion ages, when
combined with the known distances to the Magellanic Clouds and previously measured
expansion velocities. We have used the published PC1 MCPN images from D96, and other
relevant physical parameters from the literature, to obtain a total group of 27 extragalactic
planetary nebulae with known distance, morphology, and dynamical age, by classifying
all the PNe with the same morphological scheme. The main scientific content of this
paper is the presentation, discussion, and analysis of the new MCPN data acquired with
HST/FOC. We also attempted a limited analysis of nebular properties across morphological
classes. The results suffer from low statistics, especially within each class. Nonetheless,
we find some trends that would confirm previous studies on galactic PNe, mainly, that
symmetric and asymmetric PNe seem to belong to different brightness group (in the λ5007
[O iii] line), possibly indicating that they belong to different mass groups. In order to
have greater consistency in these results, we would need to discuss at least 20 objects for
– 27 –
each morphological class, that is, a sample of a hundred MCPNe observed with the HST
cameras. Moreover, although the morphological classification is feasible using pre–COSTAR
images, the photometric measurements can suffer considerable errors; it is thus necessary
to repeat and extend the analysis to post–COSTAR images of MCPNe. More insight into
the evolutionary paths of different nebular shape classes could be achieved by investigating
the morphological properties of the MCPNe together with their central stars. The analysis
of the sample of MCPNe presented in this paper, together with their central stars, is in
progress and will be published in the future.
Thanks to D. Shaw for discussions on the correlation between nebular morphology and
stellar evolution in the Clouds, and to M. Dickinson for helping with the IRAF routines.
The referee of this paper, M. Dopita, is thanked for his insight and very useful comments.
L.S. gratefully acknowledge the hospitality at the Space Telescope Science Institute, where
this work was completed. J. C. B. acknowledge support from NASA through the contract
NAG5-1733.
– 28 –
REFERENCES
Balick, B. 1987, AJ, 94, 671
Barlow, M. J. 1987, MNRAS, 227, 161
Blades J. C., et al. 1992, ApJ, 398, L44 (B92)
Boffi, F. R., & Stanghellini, L. 1994, A&A, 284, 248
Bond, H. E., Schaefer, K. G., Fullton, L. K., & Ciardullo, R. 1995, BAAS, 187, 8005
Cahn, J. H., Kaler, J. B., & Stanghellini, L. 1992, A&AS, 94, 399
Calvet N., & Peimbert M. 1983, Rev. Mex. Astron. Astrof., 5, 319
Chu, Y–H., Jacoby, G. H., & Arendt, R. 1987, ApJS, 64, 529
Corradi, R. L. M., & Schwarz, H. E. 1994, A&A, 293, 871
Dopita, M. A., Lawrence, C. J., Ford, H. C., & Webster, B. L. 1985, ApJ, 296, 390
Dopita, M. A., Meatheringham, S. J., Webster, B, L., Ford, H. C. 1988, ApJ, 327, 639
Dopita, M. A., & Meatheringham, S. J. 1991a, ApJ, 377, 480
Dopita, M. A., & Meatheringham, S. J. 1991b, ApJ, 367, 115
Dopita ,M. A., et al. 1996, ApJ, 460, 320 (D96)
Frank, A., & Mellema, G. 1994, ApJ, 430, 800
Hasan, H., & Burrows, C. J. 1994, Proc. SPIE, Vol 1945, p. 36
Hasan, H., & Bely, P. Y. 1993, BAAS, 183, 113.06
Hoyle, F., & Wickramasinghe, N. C. 1991, “The Theory of Cosmic Grains”, (Dordrecht:
Kluwer)
Jacoby, G. H., Walker, A. R., & Ciardullo, R. 1990, ApJ, 365, 471
– 29 –
Jedrzejewski, R. I., Hartig, G., Jakobsen, P., Crocker, J. H., & Ford, H. C. 1994, ApJ, 435,
L7
Kaler, J. B., & Jacoby, G. H. 1990, ApJ 362, 491
K¨aufl, H. U., Renzini, A., & Stanghellini, L. 1993, ApJ 410, 251
Kingsburgh, R. L., & Barlow, M. J. 1992, MNRAS, 257, 317
Kingsburgh, R. L., & Barlow, M. J. 1994, MNRAS, 271, 257
Krist, J. 1993, ADASS II, A. S. P. Conf. Ser., Vol. 52, R. J. Hanisch, R. J. V. Brissenden,
& J. Barnes (eds.), p. 536
Liu, X.-W., Barlow, M. J., Blades, J. C., Osmer, S., & Clegg, R. E. S. 1995, MNRAS, 276,
167
Lucy, L. B. 1974, AJ, 79, 745
Macchetto, F., et al. 1991, ApJ, 369, L55
Manchado, A., Guerrero, M., Stanghellini, L., & Serra–Ricart, M. 1996, “The IAC
Morphological Catalog of Northern Galactic Planetary Nebulae” (Tenerife: IAC)
Meatheringham, S. J., & Dopita, M. A. 1991a, ApJS, 75, 407
Meatheringham, S. J., & Dopita, M. A. 1991b, ApJS, 76, 1085
Munch, G., Hipplelein, H., & Pitz, E. 1984, A&A, 135, L11
Nota, A., Jedrzejewski, R., Greenfield, D. P., & Hack, W. 1994, “Hubble Space Telescope
Faint Object Camera Instrument Handbook” (post–COSTAR), version 5.0
Osterbrock, D. E. 1989, “Astrophysics of Gaseous Nebulae and Active Galactic Nuclei”,
(Mill Valley: University Science Books)
Peimbert M., & Torres–Peimbert, S. 1983, IAU Symp. 103, “Planetary Nebulae”, p. 233
– 30 –
Peimbert M. 1997, in IAU Symp. 180, “Planetary Nebulae”, H. Habing and H. Lamers
(eds.), in press
Pena, M., Torres–Peimbert, S., Peimbert, M., Ruiz, M. T., & Maza, J. 1994, ApJ, 428, L9
Richardson, W. H. 1972, J. Opt. Soc. Am., 62, 55
Richer, M. G. 1993, ApJ, 415, 240
Robinson, G. J., Reay, N. K., & Atherton, P. D. 1982, MNRAS, 199, 649
Sabbadin, F. 1984, A&AS, 58, 273
Sanduleak, N., MacConnell, D. J., & Philip, A. G. D. 1978, PASP, 90, 621
Schwarz, H. E., Corradi, R. L. M., & Melnick, J. 1992, A&AS, 96, 23
Schwarz, H. E., Corradi, R. L. M., & Stanghellini, L. 1993, in IAU Symp. 155, A. Acker &
R. Weinberger (eds.),(Dordrecht: Kluwer), p. 214
Stanghellini, L., Corradi, R. L. M., & Schwarz, H. E. 1993, A&A, 279, 521
Stanghellini, L. 1995, in “Asymmetric Planetary Nebulae”, Annales of the Israel Physical
Society, Vol. 11, eds. A. Harpaz & N. Soker, p. 17
Vassiliadis, E. 1996, ApJ, 465, 748
Vassiliadis, E., Dopita, M. A., Morgan, D. H., & Bell, J. F., 1992, ApJS, 83, 87
Weinberger, R. 1989, A&AS, 78, 301
This manuscript was prepared with the AAS L
A
T
EX macros v4.0.
– 31 –
Figure Captions
Figure 1. Contour plots of the LMC N 192 (a, b) and N 97 (c, d) pre– and post–COSTAR
images.
Figure 2. Various point spread functions (PSFs).
Figure 3. Deconvolution of the pre-COSTAR image of N 192 by using each of the PSFs in
Fig. 2.
Figure 4. Results of matching and not matching the theoretical PSF to the central star,
in N 66.
Figure 5. Narrow–band [O iii] FOC images of Magellanic Cloud planetary nebulae.
Figure 6. Parameter distribution of nebular size in pc. (top), nebular expansion velocity
in km s−1(middle), and dynamical expansion time in yr (bottom panel) for round (solid),
elliptical (dashed) and bipolar/quadrupolar (shaded histogram) MCPNe.
Figure 7. Maximum nebular dimension versus dynamical expansion age for MCPNe which
are round (open circles), elliptical (solid circles), and bipolar/quadrupolar (squares).
Figure 8. Nebular electron density versus physical dimension for PNe. Symbols are as
in Figure 7. Diameters of optically thick PNe, and of PNe with unknown thickness, can
actually be larger than observed, and indicated with arrows.
Figure 9. [O iii] surface brightness versus dynamical expansion time. Symbols are as in
Figure 7.
Figure 10. Nebular N/O abundance ratio versus dynamical expansion time. Symbols are
as in Figure 7. The solid line represents the dividing line between Type I (top quadrant)
and non–Type I PNe (bottom quadrant), as described in the text.
– 32 –
Table 1. FOC F/96 Observations of MCPNe in [O iii]
Name RAaDecaFilter Date texp counts peak counts log F
(J2000) (J2000) (UT) [s] [s−1] [s−1] [erg cm−2s−1]
SMC
N 2 00:32:38.8 –71:41:59 F501N 1991 Jul 09.62 995.9 407.27 0.442 –11.70
N 4 00:34:22.0 –73:13:21 F501N 1993 Apr 28.92 496.8 137.05 0.156 –12.19
N 5 00:41:21.8 –72:45:19 F501N 1991 Jul 09.76 995.9 334.65 0.413 –11.79
N 18 00:46:59.6 –72:49:39 F501N 1992 Nov 25.31 996.9 154.07 0.168 –12.14
L 305 00:56:30.9 –72:27:01 F501N 1993 Apr 26.92 1996.8 87.75 0.360 –12.42
N 67 00:58:37.3 –71:35:49 F501N 1993 Jan 09.85 1996.9 83.33 0.064 –12.40
L 343 00:58:42.6 –72:57:00 F501N 1993 Jul 10.26 1995.8 122.98 0.126 –12.22
L 536 01:24:11.8 –74:02:34 F501N 1993 Jul 06.66 995.8 42.82 0.130 –12.68
LMC
N 97 05:04:51.9 –68:39:10 F501N 1992 Nov 18.12 1803.2 389.99 0.282 -11.72
F501N 1994 Feb 02.81 1995.9 –11.77
F501Nb1994 Feb 02.84 995.9 –11.73
N 24 05:06:09.3 –67:45:29 F501N 1992 Dec 12.94 996.9 328.53 0.449 –11.79
N 192 05:09:37.3 –70:49:09 F501N 1993 Mar 03.01 996.8 358.85 0.145 –11.74
F501N 1994 Feb 06.81 1995.9 –11.77
F501Nb1994 Feb 06.84 995.9 –11.74
WS 12 05:10:50.0 –65:29:31 F501N 1993 Apr 28.86 1996.8 392.38 0.118 –11.72
LM 1–27 05:19:20.7 –66:58:07 F501N 1993 Jun 20.95 1996.8 187.76 0.049 –12.02
N 52 05:28:41.2 –67:33:39 F501N 1992 Nov 18.91 996.9 318.94 0.220 –11.79
N 66 05:36:20.8 –67:18:08 F501N 1991 Jun 26.94 540.3 365.20 0.130 –11.74
F501N 1993 Jul 10.19 1995.8 325.94 0.247 –11.81
F501N 1994 Feb 05.30 1995.6 –11.80
F501Nb1994 Feb 06.33 995.9 –11.76
LM 1–61 06:10:25.5 –67:56:21 F501N 1992 Nov 18.84 1778.3 281.64 0.100 –11.85
aCoordinates: STScI Guide Star Selection System
bthese filters are really F501N+F1ND
– 33 –
Table 2. Morphology, Diameters, and Optical Depths
Name SMP camera Morph.aθD opt. depth
[arcsec] [pc]
SMC
N 1 1 PC1 R .241 .068 thin
N 2b2 FOC E(Es) .637 .180 thin
N 4 3 FOC,PC1 B(Bbc) 2.64 .747 thin
N 5 5 FOC R .621 .176 thick
N 6b6 PC1 R .304 .086 thin
N 18 10 FOC B(Bbc) 2.64 .747 · · ·
L 305b21 FOC E(Es) 3.00 .846 thin
N 67b22 FOC B(Bbc) 2.74 .774 thick
L 343 23 FOC B(Bbc) 2.61 .738 · · ·
L 536b28 FOC,PC1 E 3.50 .990 thin
LMC
· · · 2 PC1 R .408 .100 thick
N 78 8 PC1 R · · · · · · thin
LM 2–5b20 PC1 B(Bbc) .823 .202 thick
N 97bc 21 FOC Q 1.15 .281 thick
N 24 23 FOC R d d thin
N 192c32 FOC R 1.083 .266 thick
WS 12 35 FOC,PC1 E(inc) 1.59 .391 thick
WS 16 40 PC1 E(Es) .783 .192 thick
LM 1–27 45 FOC E(inc) 2.36 .578 thick
N 122b47 PC1 E? .412 .101 thin
N 52 66 FOC E(Es) 1.08 .266 · · ·
· · · 72 PC1 B · · · thin
N 60 76 PC1 R · · · · · · thin
N 69 85 PC1 R · · · · · · thin
N 215b87 PC1 B(Bbc) 1.01 .248 thick
· · · b96 PC1 B(Bbc) .905 .222 thick
LM 1–61 97 FOC R(Rs) 1.18 .289 thin
– 34 –
Table 2—Continued
Name SMP camera Morph.aθD opt. depth
[arcsec] [pc]
aR=Round, E=Elliptical, B=Bipolar, Q=Quadrupolar, Rs=Round with (inner) structures, Es=Elliptical with (inner)
structures, Einc=Elliptical incomplete, Bbc=Bipolar core
bType I PN
cpre– and post–COSTAR FOC images
dsee text, §3.2.
– 35 –
Table 3. Ellipticities, Expansion Velocities, and
Dynamical Times
Name q Vexp τdyn,yr
[km s−1] [103yr]
SMC
N 1 1.0 8.46 3.94
N 2 .92 17.7 4.98
N 4 .69 18.1 20.2
N 5 1.0 16.0 5.38
N 6 1.0 19.3a2.18
N 18 .80 13.2 27.7
L 305 .90 19.3 21.4
N 67 .71 27.9 13.6
L 343 .94 17.4 20.8
L 536 .80 29.3 16.5
LMC
SMP 2 1.0 5.44 9.01
N 78 1.0 13.9 · · ·
LM 2–5 1.0 14.2 6.98
N 97 1.0 27.0 5.11
N 24 1.0 11.5 · · ·
N 192 1.0 23.2 5.60
WS 12 .82 22.7 8.44
WS 16 .64 30.0 3.14
LM 1–27 .75 20.2 14.0
N 122 .50 43.2 1.14
N 52 .85 12.7 10.3
SMP 72 .64 · · · · · ·
N 60 1.0 15.9 · · ·
N 69 1.0 6.21 · · ·
N 215 .57 20.6 5.91
SMP 96 .34 33.5 3.25
LM 1–61 1.0 25.3 5.60
– 36 –
Table 3—Continued
Name q Vexp τdyn,yr
[km s−1] [103yr]
atwo components: 12.1 and 26.74 km s−1
Observed PSF 3: SN1987a field star (Apr 1992)
Model PSF 2
Observed PSF 4: BPM 16274 (Aug 1990)
Observed PSF 1: BPM 16274 (Aug 1990)
Deconvolution: PSF 1 Deconvolution: PSF 2
Deconvolution: PSF 3 Deconvolution: PSF 4
Model PSF matched Model PSF unmatched
Deconvolution: matched PSF Deconvolution: unmatched PSF
(a) SMC N2 (b) SMC N4
(c) SMC N5 (d) SMC N18
(e) SMC L305 (f) SMC N67
(g) SMC L343 (h) SMC L536
(i) LMC N97 (j) LMC N97, Post-COSTAR
(k) LMC N24 (l) LMC WS12
(m) LMC N192 (n) LMC N192, Post-COSTAR
(o) LMC LM1-27 (p) LMC N201 (Hβ)
(q) LMC N52 (r) LMC LM1-61
