A&A 486, 529-531 (2008)
DOI: 10.1051/0004-6361:200809843
C. Contreras - C. Tappert
Departamento de Astronomía y Astrofísica, Pontificia Universidad Católica, Casilla 306, Santiago 22, Chile
Received 25 March 2008 / Accepted 12 May 2008
Abstract
Aims. NSV 13983 is catalogued as a dwarf nova based on a reported outburst from 2005. The system has not yet been studied spectroscopically. We attempt to confirm its nature as a dwarf nova and determine its orbital period.
Methods. We derive the orbital period by using time-resolved spectroscopic data to measure radial velocities.
Results. The average spectrum shows evidence that the system is a dwarf nova in quiescence. The radial velocity curves derived from measurements of the spectral lines H
and H
,
show a clear modulation with a period of 2.76 h. This places NSV 13983 below the upper edge of the gap in the period distribution of cataclysmic variables, implying that it is the 14th dwarf nova in the gap.
Key words: stars: dwarf novae - stars: individual: NSV 13983
The evolutionary status of CVs is believed to be strongly related to the orbital period, which therefore represents one of the most important system parameters. The period distribution of CVs consequently qualifies as an observational testbed for theoretical models of CV evolution. One of its characteristic features is the apparent lack of systems with periods ranging between 2-3 h. This so-called period gap represents a division between high-mass-transfer, long-period systems, that are mainly driven by magnetic braking, and CVs of short orbital periods and low-mass-transfer rates (Verbunt & Zwaan 1981; Kraft et al. 1962).
NSV 13983 was first observed to be a potential variable star at Bamberg Observatory by Friedrich & Schoffel (1971), and measured to have a photographic magnitude of 11. The system is classified as a dwarf nova in the Downes et al. (2005) catalogue based on an outburst reported during an AAVSO discussion in April 2005. We present the first spectroscopic study of NSV 13983, and derive its orbital period based on the radial-velocity variation in the system emission lines.
We obtained time-resolved spectroscopic data on July 4, 2006 at CTIO using the R-C spectrograph mounted on the Blanco 4 m telescope. The KPGL3 grating was used in combination with a 1.0'' slit yielding a wavelength range of 3590-7290
and a spectral resolution of 4.2 Å.
Twenty-four exposures were taken covering 3.08 h of observations. Individual exposure times were 240 s and 300 s. In addition, HeNeAr lamp spectra were taken every 5 object exposures for wavelength calibration purposes. No observations were acquired to complete a flux calibration of the data. Bias correction, flatfielding, and wavelength calibration of spectra were performed using the standard IRAF routines.
![]() |
Figure 1:
Average spectrum of NSV 13983. The large artifact at ![]() ![]() |
Open with DEXTER |
The average spectrum of NSV 13983 is presented in Fig. 1 and the properties of its prominent features are given in Table 1. The spectrum is dominated by strong emission lines of the Balmer series, clearly visible from H
through to H
,
and also by He I lines. We can furthermore identify Ca II (H and K emission) and Fe II
5169 lines. We do not observe highly-ionized lines like He II nor any lines that could be associated with the secondary star. The strength of the emission lines and the absence of HeII, imply that the system is a non-magnetic low-mass-transfer system, i.e. a dwarf nova. The absence of clear signatures of the secondary star reflects the faintness of this component and indicates a short orbital period.
Table 1: Properties of the most prominent features in the average spectrum of NSV 13983. Column 1 represents the observed wavelength determined by a Gaussian fit, Col. 2 provides the equivalent width, Cols. 3 and 4 the line identification and the rest wavelength, respectively. A colon marks uncertain values.
To determine the orbital period for this system, the radial velocity variations of the H
and H
lines were determined by fitting a single Gaussian profile to each emission line using the ``k'' routine in the splot package of IRAF. The measured radial velocities had to be corrected for two important effects: first, for the motion with respect to the local standard of rest, which was calculated using the rvcorrect routine in the astutil package of IRAF, and second, for the instrumental flexure between two calibration exposures by measuring the position of the night sky emission line at 6300.304 Å and subtracting the variation with respect to its rest wavelength.
Due to its superior signal-to-noise ratio, only radial velocities determined using the emission line of H
were analyzed to search for periodicity. This analysis was performed using the Scargle algorithm (Scargle 1982) implemented in MIDAS. Due to the limited orbital coverage of only one night of observations, the periodogram yields one broad peak that is centered on f = 8.68 cyc d-1, corresponding to P = 2.76 h (Fig. 2).
![]() |
Figure 2:
Scargle periodgram for H![]() |
Open with DEXTER |
The radial velocities of both emission lines were folded by this period and fitted by the sine function
![]() |
(1) |
Table 2: Calculated fit parameters.
![]() |
Figure 3:
Radial velocities of H![]() ![]() |
Open with DEXTER |
The uncertainty in the period,
,
was estimated by using Eq. (4) of Larsson (1996),
![]() |
(2) |
The corresponding radial velocities and their fits are presented in Fig. 3. The H
plot shows a larger scatter of the data points compared to that of H
,
but still yields an acceptable fit with respect to P. We therefore identify
P = 2.76(27) h with the orbital period of NSV 13983. Nevertheless, because of the limited amount of data presented here, more observations would be desirable to be able to determine a more accurate constraint on this value.
We speculate that NSV 13983 has a low orbital inclination, due to the low semi-amplitude
K1 = 46 km s-1 of its radial velocity. In contrast, Szkody et al. (2004) measured
K1 = 110 km s-1 for the radial velocity of H
in SDSS J170213.26+322954.1, also a dwarf nova in the gap. This system has a high orbital inclination of
(Littlefair et al. 2006).
The derived spectroscopic period places NSV 13983 in the upper third of the period gap (Fig. 4), where we have assumed the limits on the period gap to be P = 3.18 h (upper edge) and P = 2.15 h (lower edge) given by Knigge (2006).
![]() |
Figure 4: Close-up of the period distribution of CVs about the period gap, based on version 7.9 of the Ritter & Kolb (2003) catalogue, excluding objects of uncertain period (i.e., those marked with a colon). Two distributions are shown: all CV subtypes (light grey) and dwarf novae only (dark grey). The dashed lines mark the limits of the period gap defined by Knigge (2006). The bin size is 0.02 units. The arrow marks the position of NSV 13983, which is not included in the histograms. |
Open with DEXTER |
Schmidtobreick & Tappert (2006) listed 11 dwarf novae of periods that reside in the gap. One more system, V1006 Cyg, is added in version 7.9 of the Ritter & Kolb (2003) catalogue. Dillon et al. (2008) found a period of 2.38 h for the possible dwarf nova SDSS J124426.26+613514.6. This makes NSV 13983 the 14th dwarf nova to be found in the period gap and the one with the fourth longest orbital period in the gap.
Acknowledgements
We would like to thank the referee for helpful comments.