Chandra and XMM-Newton observations of RS Ophiuchi in the 2006 outburst
M. Orio (INAF-Padova and U Wisconsin), T. Nelson (U Wisconsin and INAF), E. Leibowitz, J. Cassinelli (U Wisconsin), D. Prialnik, O. Yaron (U Tel Aviv), P. Mucciarelli (Univ Padova)
RS Ophiuchi outbursts:
• • • • • • • Recurrent nova. Very fast: Decay by 2 mag in 2-3 days. V~4.3 at max, V~12.5 at quiescence Outburst lasts less than a year. Similar shape of optical light curve but different time scales in each burst. Outbursts detected at all wavelengths from radio to X-ray. Interaction with giant wind makes it different from most novae (violently shocked material in Xrays and IR). Thought to be massive CO WD=> possible type Ia progenitor ???
•
Summary of X-ray observations:
• 4 RXTE observations in first 2 weeks. • SWIFT XRT observations every ~2 days in February, March and April. • 5 XMM-Newton observations. • 1 Chandra HETG spectra in February (DDT time) • 3 LETG spectra (March, April, June, DDT)+continued in the fall during decay.
The EPIC-pn spectra clearly show the exceptional evolution
February: The wind
Two weeks after outburst: rich emission line spectrum; resonance lines of He-l like and H-like ions of Fe, S, Si, Mg and Ne. Wide range of plasma temperatures: coolest transition O VIII Lyman alpha doublet at 19.97 A (emissivity peak at T=3MK; but also N VII in RGS ), hottest is Fe XXV 1s2p-1s2 resonance line (emissivity peak at T=60MK).
Lines are blue shifted with velocity inversely dependent on ionization state! (Only Fe XXV is red shifted). FWHM ~1500-2000 km/s, except Fe XII 3540 km/s.
Collisional ionization or photoionization? From (f+i)/r ~1 we infer a collisionally ionized plasma.
Thermal model with different components (APEC): difficulty to explain soft RGS lines, need to assume nova wind with O and N enriched abundances, so… the colliding winds are all ejected by the nova, no red giant material? Alternative model: bow shocks in a clumpy red giant wind, like O stars (Zeta Pup model).
Comparison between XMM-Newton spectra on February 26 and March 10: the nebular emission line spectrum is still present… but it cools… and the soft emission emerges.
~35 sec period detected only in this portion of light curve The soft X-ray light curve is very variable on short time scales all through March. This is the March 10 EPIC-pn lightcurve, at E < 0.4 keV. Variability less evident at 0.4-1 keV.
The lines in the soft spectrum change dramatically, ID possible only if v~8000 km/s.
The emerging supersoft X-ray source:
In the April RGS spectrum, pile up makes lower energy data unusable, longwards of 28 AA we have no usable data.
Model fitting: initial results
T~800,000 K at maximum, unvaried for 2 weeks Using Rauch model for V4743 Sgr, we find evidence of high N/C (typical of CNO ashes) => some evnvelope retained The luminosity increased in March, then remained constant until the end of April. A rapid decay followed at endApril and May
Spectrum on April 20
The spectrum after May 2007:
• Supersoft phase ended - emission line spectrum observed with almost no continuum. Not back to accretion phase yet, and not back to minimum in X-rays.
What have we learned so far?
Still working on wind spectra: collisionally ionized plasma with multiple components and nova abundances, or more complex models necessary? Atmospheric models indicate that the WD is very hot - around 800,000K. 35 s period discovery very important! May be spin period of WD, possibly related to collimated outflow from WD poles? Fast cooling time, hight T(WD) imply low envelope mass => massive WD. C/N high, data suggest C depletion => some burned material is not ejected but retained. WD is indeed very massive and compact. It seems to be a CO WD, as predicted in type Ia SN models.