A&A 461, 943-947 (2007)
DOI: 10.1051/0004-6361:20066110
L. Schmidtobreick1 - C. Tappert2 - H. Horst1,3 - I. Saviane1 - C. Lidman1
1 - European Southern Observatory, Casilla 19001, Santiago 19, Chile
2 -
Departamento de Astronomía y Astrofísica,
Pontificia Universidad Católica, Casilla 306, Santiago 22, Chile
3 -
Zentrum für Astronomie, ITA, Universität Heidelberg,
Albert-Überle-Str. 2, 69120 Heidelberg, Germany
Received 26 July 2006 / Accepted 10 October 2006
Abstract
Aims. TV Ret was classified as a cataclysmic variable due to an outburst observed in 1977. We intended to confirm this classification and derive some basic properties of the system.
Methods. Low resolution optical spectra were obtained for a spectral classification of the object.
Results. We find that the object is not a cataclysmic variable but an emission line galaxy with a redshift z=0.0964. An R-image taken in very good seeing conditions shows that the object is extended.
Conclusions. We show that TV Ret is a blue dwarf galaxy, probably compact, with an absolute magnitude of
MB = -17.5, a metallicity of 0.12 solar, and an average temperature of
K. The line ratios place it among the H II galaxies, although close to the border of the Seyfert 2s. The outburst, which was observed in 1977, could thus be explained by a supernova explosion. However, with an absolute magnitude around MB = -21, it was an extremely bright one.
Key words: stars: dwarf novae - stars: individual: TV Ret - galaxies: dwarf - galaxies: starburst - galaxies: active
We performed spectroscopic observations in order to confirm this classification. Surprisingly, the object turned out to be a narrow emission-line galaxy of nearly 0.1 redshift. This class of galaxies consists of two different types: starburst or H II-galaxies and the narrow-emission line AGNs, which are either Seyfert 2 galaxies or LINERs.
In the first case, the emission lines origin in gas that is photo-ionised by young, hot OB stars present in the star-forming regions. In the case of AGNs the emission lines are formed in the Narrow Line Region which is comprised of clouds ionised by radiation from the central engine of the galaxy. Several empirical methods have been developed to distinguish between these two ionisation mechanisms, mainly by comparing the line-ratios in the spectrum (see e.g. Baldwin et al. 1981; Veilleux & Osterbrock 1987; Dessauges-Zavadsky et al. 2000).
We attempt such a distinction and also derive the physical parameters of the galaxy such as size, luminosity, metallicity, and gas temperature and density.
The observations were performed on 2003-09-26 and on 2004-11-19 using EFOSC2 (Buzzoni et al. 1984) at the 3.6 m telescope at La Silla, Chile. In Table 1 the observational parameters are summarised.
The standard reduction of the data was performed using IRAF. The bias was subtracted and the data were divided by a flat field, which was normalised by fitting Chebyshev functions of high order. The spectra were optimally extracted (Horne 1986). Wavelength calibration yielded a final resolution of 1.1 nm FWHM for the 2003 data and 1.2 nm FWHM for the 2004 data. Flux calibration was performed only for the 2003 spectrum, using the spectrophotometric standard LTT 7379 which was observed with an airmass difference of 0.05.
For all further analysis, if not especially indicated, the MIDAS package and self-written routines were used.
In Fig. 1 the flux-calibrated spectrum observed in 2003
is plotted. The spectrum is clearly dominated by strong emission lines.
The continuum matches the spectral energy distribution (SED) of a late A or
early F type
star; the corresponding temperature is
K.
In Fig. 2,
the normalised spectrum from 2004 is plotted, which shows the same
emission lines as the 2003 spectrum and an additional line at 408.9 nm,
which was outside the spectral range of the 2003 data.
Both spectra show clearly that the object is
not a cataclysmic variable, as the typical emission lines for this kind
of object are not present at their rest wavelengths.
Instead, we find several strong emission lines,
which turned out to be redshifted Balmer lines as well as
O I, O III, N II, S II, and Ne III. The properties of these lines
are given in Table 2.
We averaged the individual shifts of these
lines to find the redshift of the object as
z = 0.0964(2).
The redshift, which indicates an extragalactic object, as well as the
strength of the ionised lines, are best interpreted if we assume the
object to be an emission line galaxy,
which, however, is unresolved in previous images.
In Fig. 3, a arcmin R-image is plotted,
obtained under good seeing conditions of 0.6 arcsec.
In this image, the
object appears extended and slightly elongated. The size can be estimated to
about
arcsec.
For all further calculations, we use the standard flat cosmology with
and
(Spergel et al. 2003).
From
z = 0.0964(2)
we derive the angular size distance
Mpc. Thus, assuming that we see most of it,
the size of the galaxy is
kpc, which matches the size of a rather small
dwarf galaxy.
Table 1: All observations obtained for this investigation listed with their date, the telescope/instrument combination, the used grism and slit width, and the exposure time.
![]() |
Figure 1: The flux-calibrated spectrum of TV Ret taken in September 2003. The upper plot shows the full intensity range; in the lower plot the intensity range has been decreased to better display the weak emission lines and the slope of the continuum. |
Open with DEXTER |
![]() |
Figure 2: The spectrum of TV Ret taken in November 2004 with the continuum normalised to unity. The upper plot shows the full intensity range; in the lower plot the intensity range has been decreased to better display the weak emission lines. |
Open with DEXTER |
Table 2: Observed wavelengths, equivalent widths, and FWHM of all identified emission lines in the 2002 and 2003 spectrum of TV Ret. For 2003 the line flux is also given. Note that the uncertainty of the line flux describes the uncertainty of the relative flux in the line and does not include the photometric error. The line flux has been dereddened using the reddening laws for H II regions and for AGNs (see text for details).
The magnitude of TV Ret is given as B=20.5 (Kinman et al. 1991).
The absolute B-magnitude can be derived using
![]() |
(1) |
A more detailed classification of the galaxy can be done by analysing the emission lines. Apart from the features at 478.5 nm and 737 nm, which are clearly blends of [O I] and [SII 673.1]/[SII 671.7] respectively, the lines are not resolved within our resolution of 1.1 nm, and therefore are narrow-emission lines. Thus, TV Ret could be either an H II galaxy or a narrow-line AGN. We use the flux-ratio of the different emission lines to find the mechanism by which the emission lines are produced following the method described by Veilleux & Osterbrock (1987).
To have a comparable data set, we also determined the reddening in the
same way as these authors, using the
H/H
flux-ratio and the Whitford reddening curve parameterised
by Miller & Mathews (1972). We measure
.
If we assume
a starburst galaxy with a recombination value
,
the reddening
is
E(B-V) = 0.13. If we assume an AGN with
,
we find a rather low value of
E(B-V) = 0.05. For both cases, the dereddened flux values are listed
in Table 2. Although the dereddened flux values are different,
the logarithms of the line ratios used for the classification are similar
(see Table 3). This is expected, as the line ratios were chosen
for being insensitive to the reddening, i.e. are close in wavelength.
We compared the line ratios of the forbidden lines (see Table 3)
with the values given by Veilleux & Osterbrock
(1987), i.e. their Figs. 1-6. In all these diagnostic diagrams,
TV Ret lies close to the
border between AGNs and H II region-like objects, which is mainly due
to the high value of
.
In
versus
and
versus
,
it lies on the side of the H II region-like objects, while for
versus
it lies on the side of the AGNs.
![]() |
Figure 3:
The R-band acquisition image shows the
![]() |
Open with DEXTER |
Comparing the line ratios with the diagnostic diagrams of Dessauges-Zavadsky (2000) yields similar conclusions. We can definitely exclude that TV Ret is a LINER, but it is on the border between the Seyfert 2 and the starburst galaxies, although the latter appears slightly more likely.
Table 3:
The logarithm of the line ratios is given for
the two reddening laws, starburst (SB) and AGN,
as derived from
.
The forbidden line fluxes correspond to the following transitions:
O III: 500.7 nm, N II: 658.3 nm, S II: 671.6 + 673.1 nm, and
O I: 630.0 nm.
Table 4: Physical parameters of the HII galaxy.
For a first indication, we compared the line ratios with the models of Ferland & Netzer (1983) and Evans & Dopita (1985), which are overplotted on the diagnostic diagrams of Veilleux & Osterbrock (1987), and find a metallicity of 0.1 times solar and an ionisation temperature of 45 000 K.
A more sophisticated computation of the abundances have been done
following the so-called direct method
(e.g. Osterbrock 1989). The [OIII] region electron
temperature
,
electron density
,
and abundances
were computed using tasks within the NEBULAR package of IRAF.
The temperature was computed using the [OIII] lines ratio
(4959+5007)/4363,
and the density using the [SII] line ratio 6716/6731. Using these
values for
and
,
the ionic abundances
were computed with the IONIC task, with central wavelengths,
line ratios and errors taken from Table 2.
The total oxygen abundance
was computed as
,
i.e. neglecting
the usually small contribution from
(Skillman & Kennicutt
1993). In the case of sulfur, the abundance
was computed as
,
with the ionisation correction factor (ICF) computed as
.
This expression for the ICF was first proposed by Stasinska
(1978)
with
,
but we used the value
which reproduces
Garnett's (1989) photo-ionisation models better in the
range
.
Nitrogen abundances were computed
assuming
,
and then using the nitrogen
to oxygen ratio in the equation
.
In Table 4, the derived parameters are listed for TV Ret.
We find that the average metallicity is about 0.12 solar, in agreement
with the values derived from the model of Ferland & Netzer (1983).
The gas temperature derived from the [OIII] lines is
K,
the density derived from the
[S II] lines is about 240/cm3.
These values again confirm TV Ret as an H II galaxy.
If this was an outburst in a foreground object, by chance superimposed on the
galaxy,
we can estimate an upper limit for the brightness of this object.
With a
of our spectra, we would be able to see an
object about three times
fainter than the galaxy. Since no trace of such a forground object is
visible in the spectrum, it must be fainter than
22 mag.
In that case, the
amplitude of the outburst was at least 5.5 mag, leaving two
explanations, a dwarf nova or a nova outburst. In the case of a dwarf nova, the
object would be in quiescence now, close to 22 mag, but would show strong
Balmer emission lines.
No emission is found at the restframe wavelength of these lines, making
this possibility very unlikely. Note that we cannot refute the possibility
of a dwarf-nova super-outburst which would imply that the quiescent object
is even fainter, and the Balmer emission are no longer observable.
In the case of a nova, the outburst amplitude itself would have been larger,
so that the object is not visible in quiescence. However, even faint novae
have an absolute magnitude of MB = -7 (e.g. Della Valle & Livio
1995), yielding a lower limit of its
distance as 450 kpc and placing it far outside our Galaxy.
![]() |
Figure 4: The B-magnitude of TV Ret plotted against the heliocentric Julian Date. The data are taken from Table 5 of Kinman et al. (1991). The smaller plot is a zoom on the last 40 days, when the outburst happened. |
Open with DEXTER |
In the following, we assume that the outburst originates in the galaxy
and is not a chance transient from a foreground object. The
luminosity
of the outburst can then be estimated via
![]() |
(2) |
The variation on short time scales indicates a compact source, so an AGN seems more likely. However, they do not generally show such strong variations in the optical. While X-ray variability on short timescales is a common phenomenon in AGNs, optical intra-night variability of up to 10% is only found in luminous Quasars with MB < -24.5 (Gupta & Joshi 2005; Stalin et al. 2005). More extreme variabilities are only exhibited in Radio-loud AGN with ultra-relativistic jets and there is no indication that TV Ret is such an object. Radio and/or X-ray observations of this galaxy would help to decide on the presence of an AGN and thus also on the possible nature of the outburst.
There is a possible third explanation, if we assume that the accuracy of the measurements of Kinman et al. is not the 0.03 mag that they claim, but rather of the order of 0.3 mag. In that case, the short-term variation would not be real and for the outburst magnitude one would have to subtract the average magnitude values of outburst and quiescence magnitudes instead of the extreme values. This would lower the outburst amplitude by 0.4 mag, putting it closer to the possible range of a bright supernova. However, the method that the authors used is well-known and the quoted errors seem reasonable for it. Still, the fact that the variation is also present during the quiescence phase might indicate that it is a scatter in the measurements rather than a real variation. On the other hand, the scatter seems to be slightly larger during outburst, where one would actually expect the higher accuracy. We know of no photometric monitoring of TV Ret after 1977, which would be an important observation. If the short-term variation was confirmed by such measurements, this would be a strong indication for the presence of a compact source, i.e. an AGN. If no variation was detected, the probability would grow that the variation in 1977 is just the uncertainty in the measurements.
All in all, the reason for the outburst remains a mystery. Too many possible explanations exist, and none of them is really convincing. Unless the original measurements have rather high uncertainties, we can conclude that the outburst was some rare and unusual event that is not comparable with normal sources of variability.
We have no final conclusion concerning the observed outburst. If associated with the galaxy, it is about 1 mag brighter than normal supernovae of type Ia. One would thus prefer to explain the outburst via a foreground transient. However, the two possible explanations - dwarf nova or nova - have their drawbacks as well.
Acknowledgements
This research has made use of the Simbad database operated at CDS, Strasbourg, France. We thank the referee for valuable comments.