The persistent, spectral variable pre-nova X-ray source in V2491 Cyg: Swift J194302.1+321913

ATel #1478; A. Ibarra, E. Kuulkers (ESA/ESAC, Spain), A. Beardmore, P. Evans (U. Leicester, UK), K. Mukai (NASA/GSFC & UMBC, USA), J.-U. Ness (Arizona State U., USA), M. Orio (INAF-Padova, Italy & U. of Wisconsin, USA), J. P. Osborne, K. L. Page (U. Leicester, UK), R. Saxton (ESA/ESAC, Spain), S. Starrfield (Arizona State U., USA), J. Tueller (NASA/GSFC, USA)
on 15 Apr 2008; 19:43 UT
Credential Certification: Erik Kuulkers (ekuulker@rssd.esa.int)

Tweet

We provide further details on the V2491 Cyg pre-nova X-ray source (ATel #1473). The full designation of the pre-nova XRT source is Swift J194302.1+321913.

Below we list all Swift/XRT pre-nova observations, and provide the background corrected count rates in two bands, and the 68% confidence Baysian rate errors (Kraft et al. 1991, ApJ 374, 344) because of the low numbers of counts. The source is variable, and changes dramatically in spectral shape between observations on a time scale down to 4 days. Especially noticeable is the soft spectrum on 2007 July 21.

Date	exposure	c/ks	c/ks
	time (ks)	0.3-1 keV	1-8 keV
2007-05-25	1.06	1.9 +1.9/-1.3	20 +5/-4
2007-06-10	0.97	11.8 +4.2/-3.6	31 +7/-6
2007-07-17	3.82	3.5 +1.3/-1.1	19 +3/-3
2007-07-21	4.87	27.9 +3.2/-3.2	9 +2/-2
2008-01-02	5.65	3.9 +1.0/-1.0	20 +2/-2

The last 3 observations provided sufficient counts to fit the spectra. We adopted N_H=1.5e21 atoms/cm^2 derived from E(B-V)=0.3 (IAUC # 8935 ) and obtained satisfactory fits with simple power laws (except in the case of the softest spectrum, which appears to be more complex than a power law). We find photon indices of -1.4+/-0.3, -4.8+/-0.5, and -1.2+/-0.3 for the observations performed on 2007 Jul 17, Jul 21 and 2008 Jan 2, respectively (90% errors are quoted). The 0.3-8 keV unabsorbed fluxes at the 3 epochs are (1.4, 4.4, 1.5)e-12 erg/cm^2/s, respectively. Except for the softest spectrum, the spectra are quite hard and more like those seen from magnetic rather than non-magnetic cataclysmic variables in quiescence.

Swift J194302.1+321913 has a position consistent with that of the faint (1.71+/-0.43 e-11 erg/cm^2/s, 14-195 keV) unidentified Swift/BAT source Swift J1942.8+3220. However, the BAT source has a position error (~6') which also includes the X-ray source Swift J194245.9+322411 at (J2000) RA=19h42m45.9s, Dec=+32d24'10.7", with 90% confidence error radius of 3.6". While the latter source has an absorbed AGN-like X-ray spectrum, it is not bright enough to account for the BAT detection.

The XMM-Newton slew-survey source XMMSL1 J194301.9+321911 (see ATel #1473) has a position error radius of 8" including systematics. This position is consistent with the pre-nova XRT source, but not with the AGN-like XRT source. The 1.2+/-0.6 c/s XMM-Newton slew source count rate would give ~170 c/ks in the XRT assuming the mean XRT spectrum (power-law index=-1.4, N_H=1.5e21 atoms/cm^2), somewhat higher than the observed XRT count rates. We note that the XMM-Newton slew source detection significance was less than 4 sigma.

The ROSAT All-Sky Survey Faint Source 1RXS J194259.9+321940 (see ATel #1473) had a PSPC count rate of 0.028+/-0.011 c/s. Assuming the mean XRT spectrum, this corresponds to ~27 XRT c/ks, consistent with the XRT observations. The position of this source is also not consistent with that of the AGN-like XRT source reported above.

Summarising, a variable and sometimes very soft X-ray source has been present at the position of V2491 Cyg from the ROSAT survey era (1990/91) to 3 months before the nova outburst. A harder spectral component would be suggested by the Swift/BAT detection, but the association with the pre-nova is not secure. This is only the 2nd nova (after V2487 Oph, Hernanz & Sala 2002, Science 298, 393) to have been seen in X-rays before eruption. Together with the optical spectral peculiarity (IAUC # 8935 ) and the very fast decline (ATel #1475), these observations mark out V2491 Cyg as an unusual nova, worthy of further observations.