Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.stsci.edu/~kgordon/papers/PS_files/uwcen_col.ps.gz Äàòà èçìåíåíèÿ: Thu Apr 24 02:49:35 2003 Äàòà èíäåêñèðîâàíèÿ: Tue May 27 07:55:14 2008 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: ï ï ï ï ð ï ð ï

Draft version March 22, 1999
Preprint typeset using L A T E X style emulateapj
THE EVER CHANGING CIRCUMSTELLAR NEBULA AROUND UW CENTAURI
Geoffrey C. Clayton 1 , F. Kerber 2 , Karl D. Gordon 1 , Warrick A. Lawson 3 , Michael J.
Wolff 4 , D.L. Pollacco 5 and E. Furlan 6
Draft version March 22, 1999
ABSTRACT
We present new images of the reflection nebula surrounding the R Coronae Borealis Star, UW Cen.
This nebula, first detected in 1990, has changed its appearance significantly. At the estimated distance
of UW Cen, this nebula is approximately 0.6 ly in radius so the nebula cannot have physically altered in
only 8 years. Instead, the morphology of the nebula appears to change as different parts are illuminated
by light from the central star modulated by shifting thick dust clouds near its surface. These dust clouds
form and dissipate at irregular intervals causing the well­known declines in the R Coronae Borealis
(RCB) stars. In this way, the central star acts like a lighthouse shining through holes in the dust clouds
and lighting up different portions of the nebula. The existence of this nebula provides clues to the
evolutionary history of RCB stars possibly linking them to the Planetary Nebulae and the final helium
shell flash stars.
Subject headings: circumstellar matter --- stars: individual (UW Cen) --- stars: variables: other
1. INTRODUCTION
R Coronae Borealis (RCB) stars are a small group of
hydrogen­deficient carbon­rich supergiants which undergo
spectacular declines (\DeltaV up to 8 magnitudes) at irregular
intervals apparently tied to their pulsation cycles (Clay­
ton 1996). A cloud of carbon­rich dust forms; temporarily
eclipses the photosphere, revealing a rich ``chromospheric''
emission spectrum; then disperses, allowing the star to re­
turn to maximum light. The dust is thought to form in
patches or puffs, not as a complete shell. Only when the
dust condenses in the line of sight will a deep decline occur,
although clouds may form during every pulsation cycle
somewhere over the surface of the star. Thus, the events
are irregular and unpredictable. Extensive ground­ and
space­based observations of R CrB, RY Sgr and V854 Cen
have pointed to a unified empirical model of the RCB de­
cline (see Clayton 1996 and references therein). A close
connection between pulsational phase and the time of dust
formation---seen in RY Sgr and V854 Cen---implies that
condensation occurs near the star (Pugach 1977; Lawson
et al. 1992, 1999). Observed timescales for radiative ac­
celeration of the dust, eclipse of the chromospheric region,
and dispersal of the dust add further support to the ``near­
star'' model.
There are two major evolutionary models for the origin
of RCB stars: the double degenerate and the final helium
shell flash (Iben, Tutukov, & Yungelson 1996). The former
involves the merger of two white dwarfs, and in the lat­
ter a white dwarf/evolved Planetary Nebula (PN) central
star is blown up to supergiant size by a final helium shell
flash. In the final flash model, there is a close relation­
ship between RCB stars and PN. The connection between
RCB stars and PN has recently become stronger, since the
central stars of three old PN (Sakurai's Object, V605 Aql
and FG Sge; Duerbeck & Bennetti 1996; Clayton & De
Marco 1997; Gonzalez et al. 1998; Kerber 1998; Kerber et
al. 1999) have had observed outbursts that transformed
them from hot evolved central stars into cool giants with
the spectral and dust formation properties of an RCB star.
This establishes a possible connection between RCB stars,
central stars of PN and the final flash scenario. On the
other hand, there is evidence that after the final flash a
central star acquires RCB characteristics only for a short
time: V605 Aql was an RCB star in 1921, but by 1986 it
appeared to have a much hotter spectrum T eff ¸ 50,000
K (Clayton & De Marco 1997). If most final flash stars
evolve so rapidly, this scenario for the formation of RCB
stars may not be able to yield the number of observed RCB
stars. Moreover, the recent claim that white dwarf merg­
ing times might not be as long as previously thought, has
made the double degenerate scenario a valid and appealing
alternative to the final flash scenario for the formation of
RCB stars (Iben et al. 1996).
So the presence or absence of nebulae around RCB stars
is of great interest. An old PN that is no longer ionized
could still be seen around a cool RCB star in starlight re­
flected from dust. An example of such a reflection nebula
is the nebulosity observed around UW Cen (Pollacco et
al. 1991). This is the only cool RCB star known to have
a visible nebula 1 . However, an extended shell­like struc­
ture around R CrB has been detected at 100 ¯m (Walker
1 Department of Physics and Astronomy, Louisiana State University, Baton Rouge, LA 70803; Electronic mail: gclayton,
gordon@fenway.phys.lsu.edu
2 ST­ECF, Karl­Schwarzschild­Strasse 2, D­85748 Garching bei M¨unchen, Germany; Electronic mail: fkerber@eso.org
3 School of Physics, University College, University of New South Wales,
Australian Defence Force Academy, Canberra ACT 2600, Australia; Electronic mail: wal@ph.adfa.edu.au
4 Space Science Institute, Suite 294, 1234 Innovation Dr., Boulder, CO 80303­7814; Electronic mail: wolff@colorado.edu
5 Isaac Newton Group, Santa Cruz de La Palma, Tenerife 38780, Canary Islands, Spain; Electronic mail: dlp@ing.iac.es
6 Institut f¨ur Astronomie, Universit¨at Innsbruck, Technikerstr. 25, Innsbruck, Austria Electronic mail: elise.furlan@uibk.ac.at
1 V348 Sgr, a hot RCB star has an ionized nebula surrounding it (Pollacco, Tadhunter, & Hill 1990)
1

2 UW Cen
1986).
2. OBSERVATIONS
The lightcurve of UW Cen has been provided by
the American Association of Variable Star Observers
(AAVSO, Mattei 1999, personal communication). The
two observational epochs discussed here occurred during
very deep declines of UW Cen when the star was ¸6 mag
below maximum (V max ¸9.5). Figure 1a shows the V­
band image of UW Cen obtained on 1990 May 16/17 with
the ESO New Technology Telescope when the star was at
V=16.5. The description of instrument setup and reduc­
tion is contained in Pollacco et al. (1991). The image
scale is 0.3 00 pixel \Gamma1 and the seeing was ¸1 00 . We have
obtained new BVR­band images using the 1 m and 2.5
m telescopes at Las Campanas Observatory. At the 1m
Swope telescope, we used the SITE #1, chip (2048 x 2048
pixels) which gave a field of view of almost 24 0 at a scale of
0: 00 69/pixel. At the 2.5 m du Pont telescope, we employed
the TEK 5 chip (2048 x 2048 pixels) with 24 ¯m pixels re­
sulting in a scale of 0: 00 26/pixel. In all observations, two or
three exposures of 100 to 600 s duration, depending on the
filter used, were taken and co­added during reduction. The
intensities in the 1m telescope images were calibrated using
observations of the PG1323­086 field which were obtained
before and after the UW Cen images (Landolt 1992). The
2.5 m images were then calibrated using the 1 m data.
Only the V­band image from the 2.5 m telescope is shown
in Figure 1b. It was obtained on 12 February 1998 when
UW Cen was at V=14.4 and the seeing was better than
1 00 .
3. MODELING THE NEBULA
Figure 1 shows how different the UW Cen nebula ap­
pears in images taken just 8 years apart. The nebula is
very tenuous with a maximum surface brightness of ¸21­
22 mag/ut 00
. In 1990, the nebulosity is seen predominantly
to the west of the star with clumps to the northeast, east
and southwest. In 1998, there is shell­like nebulosity cov­
ering about 45 ffi to the north of the star and very little
elsewhere. The nebula has a diameter of approximately
15 00 centered on the star. The best estimate of the distance
to UW Cen is 5.5 kpc (Lawson et al. 1990). The physical
radius of the nebulosity is then about 6 x 10 17 cm or 0.6
ly. In order for material to flow from the star to the edge
of the visible nebula in only 8 years, an outflow velocity of
¸24,000 km s \Gamma1 would be necessary. Such a large velocity
has never been seen in an RCB star. Outflow velocities
of a few hundred kilometers per second are typically seen
in the circumstellar gas during declines (Clayton 1996). If
this is a final flash shell, then 30­40 km s \Gamma1 is a more rea­
sonable velocity (Guerrero & Manchado 1996). Therefore,
the apparent changes seen between 1990 and 1998 were
not caused by an actual change in the distribution of dust
clouds in the nebula. The likely explanation is that the
dust nebula around UW Cen is being illuminated through
shifting holes in a changing distribution of newly formed
dust near the star. Since both images were obtained dur­
ing deep declines we can surmise that dust had recently
formed around UW Cen in both instances. To investigate
this idea, we have produced a simple model of UW Cen
and its circumstellar dust.
We employed the DIRTY radiative transfer model (Gor­
don, Witt, & Clayton 1999). This model uses Monte Carlo
techniques to calculate the radiative transfer through ar­
bitrary distributions of dust viewed from any angle. The
inputs to DIRTY are the stellar distribution, in this case
one star, the dust distribution, and the dust grain prop­
erties. We have modeled the dust distribution in the UW
Cen nebula using two concentric shells. The model inner
shell lies near the star representing the new dust clouds
which form during a decline episode. This shell is opti­
cally thick, Ü V =3, but contains a number of holes through
which starlight shines to illuminate the outer shell. The
model outer shell is a 1 00 thick homogeneous shell of dust
located at a distance of 6: 00 5 to 7: 00 5 from the star. In be­
tween the two shells, there is a uniform distribution of dust
having a density which is a factor of ten below that of the
outer shell. The optical depth of the nebula not including
the inner optically thick shell is small, Ü V =0.19. The dust
albedo and scattering phase function are computed assum­
ing amorphous carbon grains similar to those deduced for
V348 Sgr (Hecht et al. 1998). The size distribution of the
dust grains is a modified gamma function over the interval
0.03 ­ 0.10 ¯m. The resulting dust albedo in the V­band
is 0.025 and the Henyey­Greenstein scattering phase func­
tion asymmetry is 0.14. Very good fits to the two images
were produced by varying the number, size and optical
depth of the holes in the inner shell.
Our model images are shown in Figure 2. Only a few
holes in the inner shell are necessary to fit the observed
nebula. The covering factors of the inner shell in these
models are 87% and 89% for the 1990 and 1998 images, re­
spectively. These numbers include a 5% contribution from
photons passing through the Ü V =3 inner shell. However,
the covering factor of the hemisphere facing away from the
line of sight is not strongly constrained in this model. It is
possible to uncover a significant proportion of the back of
the star, ¸20% of the surface area of the whole star, with­
out seeing significant variation in the model images. The
dust behind the star contributes less to the scattered light
because of the slight asymmetry in the scattering phase
function. The model used here was chosen for simplicity.
It is not unique. The observations could also be modeled
using a clumpy outer shell and a smaller covering factor
for the star. More images obtained during declines of UW
Cen are needed to fully map the nebula. When its true
morphology is determined, the UW Cen nebula can be
compared to the final flash stars. In A30, for example, the
PN is symmetric while its final flash shell is very clumpy
(Guerrero & Manchado 1986).
4. DISCUSSION
The B­V colors of the UW Cen nebula are bluer than
the star near the star and then progressively redder toward
the outer edge. These colors are typical of those observed
in other reflection nebulae. Our model implies that the
total optical depth from the star through the visible neb­
ula is ¸0.2. Assuming the amorphous carbon grains used
for our model, the dust mass is ¸6 x 10 \Gamma4 M fi . If we
assume a normal interstellar gas­to­dust ratio and solar
hydrogen abundance for the nebula, then the total mass
is ¸ 0:2 M fi (Bohlin, Savage, & Drake 1978). UW Cen is
extremely hydrogen deficient and nothing is known about
the abundances in the nebula so this estimate in very un­

Clayton et al. 3
certain. The mass of the final flash shell of A30 has been
estimated to be ¸0.1 M fi . The large IRAS shell around
R CrB has an estimated mass of ¸0.3 M fi (Gillett et al.
1986).
Comparisons of the direct emission and re­radiation
from circumstellar dust in the IR give clues to the cov­
ering factor of new dust clouds forming around RCB stars
(Clayton 1996 and references therein). Figure 3 shows the
flux distribution for UW Cen. The photometry of UW
Cen were obtained from Jones et al. (1989), Kilkenny &
Whittet (1984) and the IRAS Point Source Catalog. The
optical and near­IR photometry were obtained when the
star was at, or near, maximum light. These fluxes were
dereddened assuming a small optical extinction (A V ú 0.2
mag; Lawson et al. 1990) and using the typical diffuse in­
terstellar extinction law (Rieke & Lebofsky 1985; Cardelli,
Clayton & Mathis 1989).
­28.0
­27.5
­27.0
­26.5
­26.0
­25.5
­25.0
­24.5
­24.0
log
Flux
Density
(W
m
--2
Hz
--1
)
12.0
12.5
13.0
13.5
14.0
14.5
15.0
15.5
log Frequency (Hz)
6000 K
650 K
UBVR I J H K L M N Q
12 25 60 100
Fig. 3.--- The flux distribution of UW Cen.
The stellar (optical and near­IR) flux and dust nebula
flux (mid­ and near­IR wavelengths) were fitted by two
Planck functions with temperatures of 6000 \Sigma 500 K and
650 \Sigma 50 K, respectively. Planck functions tend to under­
estimate the T eff of the star by 500--1000 K. UW Cen is
an F­type RCB star like R CrB and RY Sgr, which has
T eff of 7000 \Sigma 500 K (Lawson et al. 1990). Wavelengths
longer than 12 ¯m probably show the effect of bright IR
cirrus towards UW Cen. The cirrus contribution can be
fitted with a 100 K Planck function. The luminosity ratio
F IR /F ? was calculated for UW Cen, where F IR is the inte­
grated flux of the Planck fit to the IR excess and F ? is that
for the star. For UW Cen, we derive a ratio of F IR /F ? ¸
0.3. This is comparable to values given for UW Cen and
other RCB stars based on long­term IR photometry (Feast
et al. 1997). They find an apparent upper limit of F IR /F ?
¸ 0.5 for RCB stars. Since the albedo for carbon dust
is so low, this ratio is a measure of the covering factor
of the star. So typically, no more than 50% of the stel­
lar radiation is absorbed by the circumstellar clouds and
re­radiated in the IR. No significant change is measured
in the IR flux during an RCB decline indicating that the
covering factor does not increase by a large amount during
any dust formation episode.
Several studies of UV extinction due to the RCB circum­
stellar dust have estimated the covering factor (f) of the
dust clouds by fitting the extinction and scattering in the
dust. For the cool RCB stars, R CrB and RY Sgr, Hecht et
al. (1984) find f ¸ !0.5. These estimates are consistent with
the visible/IR result and imply an upper limit of f=0.5 for
RCB stars. However, the results of our modeling of the
UW Cen nebula imply a very high covering factor. This
result is softened by the fact that the model is somewhat
insensitive to the value of f on the far side of the star and
that there are only data from two declines. The covering
factor undoubtedly varies from decline to decline.
If UW Cen is a final flash object then it will be an old
PN central star and should be surrounded by the now neu­
tral PN as well as the final flash shell. Seven PN in the
Galaxy, A30, V605 Aql, A78, Sakurai's object, FG Sge,
IRAS 18333--2357 and IRAS 15154--5258 are hydrogen de­
ficient and have central stars which have experienced a
final helium shell flash (e.g., Guerrero & Manchado 1996;
Clayton & De Marco 1997, Jacoby, De Marco & Sawyer
1998; Gonzalez et al. 1998). In each of these objects,
the old PN surrounds a smaller final flash shell. Pre­
dicted time scales indicate that the final flash will happen
¸9000­18000 yr after the first ejection depending on mass
(Bl¨ocker 1995, Iben et al. 1983) and that the star will
rejoin the AGB a few hundred years after the final flash.
If we assume the star becomes an RCB shortly afterwards
(as indicated by Sakurai's object), and stays that way for
¸3000 to 10 000 yr (Iben et al. 1996) and the dust is ex­
panding at 35 km s \Gamma1 , then we would expect the old PN
for UW Cen to have a diameter of 24 00 to 50 00 , its final flash
shell to have a diameter of 8 00 to 26 00 and RCB dust inside
that. These diameters agree well with the measured value
of 15 00 for the UW Cen nebula. Since only one nebula has
been detected around UW Cen, it could be either an old
PN or the final flash shell. Alternatively, the nebula could
be a wind­blown bubble in the ISM. At 20 km s \Gamma1 , wind­
blown dust could form the observed nebula in 9,500 yr. A
link between RCB stars and PN, if it can be established,
will be an important step forward in understanding the
evolution of post­AGB stars. More observations are nec­
essary to both map out the nebula of UW Cen and to
resolve the disagreement between the large dust covering
factor predicted by the model and the smaller values in­
ferred from UV and IR data.
We thank the referee, Joel Kastner, for many useful sug­
gestions. We are grateful to Dr. Janet Mattei of the
AAVSO for providing the lightcurve data for UW Cen.
We thank Dr. O. De Marco for many helpful discussions.
It is a pleasure to thank Dr. M. Roth and the staff of Las
Campanas Observatory for their support. F.K and E.F
acknowledge a travel grant from the Austrian Ministry of
Science and the University of Innsbruck. Special thanks
to Londo Mollari. This project was supported by NASA
grant JPL 961526.

4 UW Cen
REFERENCES
Bl¨ocker, T. 1995, A&A, 299, 755
Bohlin, R.C., Savage, B.D., & Drake, J.F. 1978, ApJ, 224, 132
Cardelli, J.A., Clayton, G.C., & Mathis, J.S. 1989, ApJ, 345, 245
Clayton, G. C., & De Marco, O. 1997, AJ, 114, 2679.
Clayton, G.C. 1996, PASP, 108, 225
Duerbeck, H.W. & Benetti, S. 1996, ApJ, 468, L111.
Feast, M.W., Carter, B.S., Roberts, G., Marang, F., & Catchpole,
R.M. 1997, MNRAS, 285, 317
Gillett, F. C., Backman, D. E., Beichman, C., and Neugebauer, G.
1986, ApJ, 310, 842
Gonzalez, G., Lambert, D. L., Wallerstein, G., Rao, N. K., Smith,
V. V., & McCarthy, J. K. 1998, ApJS, 114, 133
Gordon, K. D., Witt, A. N., & Clayton, G. C. 1999, in preparation
Guerrero, M. A. & Manchado, A. 1996, ApJ, 472, 711
Hecht, J. H., Holm, A. V., Donn, B., & Wu, C. C. 1984. ApJ, 280,
228
Hecht, J.H., Clayton, G.C., Drilling, J.S. , & Jeffery, C.S. 1998, ApJ,
501, 813
Iben, I. Jr., Kaler, J. B., Truran, J. W., & Renzini, A. 1983, ApJ,
264, 605
Iben, I., Tutukov, A. V., & Yungelson, L. R. 1996, ApJ, 456, 750.
Jacoby, G. H., De Marco, O., & Sawyer, D.G. 1998, AJ, 116, 1367
Jones, K., van Wyk, F., Jeffrey, C.S., Marang, F., Shenton, M., Hill,
P., &Westerhuys, J. 1989, Circ. South African Astr. Observatory,
13, 39
Kerber F. 1998, Reviews in Modern Astronomy, ed. R.E. Schielicke,
Vol. 11, p. 161
Kerber F.,Blommaert, J.A.D.L., Kimeswenger, S., Groenewegen,
M.A.T., K¨aufl, H.U. 1999, ISO's View on the Universe, Paris,
eds.P. Cox & M. Kessler, in press
Kilkenny, D., & Whittet, D.C.B. 1984, MNRAS, 208, 25
Landolt, A.U. 1992, AJ, 104, 340
Lawson, W.A., Cottrell, P.L., Kilmartin, P.M., & Gilmore, A.C.
1990, MNRAS, 247, 91
Lawson, W. A., Cottrell, P. L., Gilmore, A. C., & Kilmartin, P. M.
1992, MNRAS, 256, 339
Lawson, W.A., Maldoni, M., Clayton, G.C., Valencic, L., Jones, A.,
Kilkenny, D., van Wyk, F., Roberts, G., & Marang, F. 1999, AJ,
in press
Pollacco, D.L., Tadhunter, C.N., & Hill, P.W. 1990, MNRAS, 245,
204
Pollacco, D.L., Hill, P.W., Houziaux, L., & Manfroid, J. 1991,
MNRAS, 248, 1p
Pugach, A. F. 1977, IBVS No. 1277
Rieke, G.H., & Lebofsky, M.J. 1985, ApJ, 288, 618
Walker, H.J., 1986, in Hunger, K., Sch¨onberner, D., Rao, N.K., eds,
IAU Coll. 87, Hydrogen Deficient Stars and Related Objects,
Reidel, Dordrecht, p. 407

Clayton et al. 5
Fig. 1.--- (left) The UW Cen nebula in 1990 (Pollacco et al. 1991). (right) The UW Cen Nebula in 1998. Both images
are plotted as log flux. The nebulosity ranges from 20.7 to 23.9 mag/ut 00
. The axes are RA offset (x) and declination
offset (y). The sources which appear in Figure 1 but not Figure 2 are field stars.
Fig. 2.--- Monte Carlo Dust Scattering Models. (left) The UW Cen nebula in 1990. (right) The UW Cen Nebula in 1998.
They are plotted as described in Figure 1