Recent research‎ > ‎

Seeing how the wind blows in an eclipsing nova

posted May 31, 2013, 7:03 AM by Jeremy Drake   [ updated Jun 13, 2013, 10:25 AM ]
After a nova explosion on a white dwarf star - see earlier posts on U Sco and V407 Cyg for descriptions of what a nova is - hydrogen-rich gas remains on the white dwarf surface undergoing steady nuclear fusion. This energy source is sufficiently strong that it lifts gas from the white dwarf purely by radiation pressure and pushes it outward at speeds of a few hundred to a few thousand km/s in a fairly steady but likely inhomogeneous radiatively-driven wind.  U Sco is a special nova because it repeats every ten years or so, and it is eclipsing so we can get a good idea of the system parameters.  The mass donor star, a slightly evolved G subgiant, eclipses the white dwarf companion every 1.23 days. 
Shortly after the outburst, the wind is so strong that its photosphere - the outer layers from which light can finally escape - lies
outside of the G star orbit.  As the nuclear burning proceeds, fuel is consumed and mass is driven off, the mass loss rate gradually declines and the photosphere recedes. As it shrinks it gets hotter according to Stefan's law, with temperature varying as the fourth root of the photospheric size for a fairly constant luminosity. By the time the photosphere is the size of the G star orbit and eclipses are seen again, the temperature is several hundred thousand degrees and it is known as a "supersoft X-ray source". At this point, we can use the length and depth of eclipses seen in soft X-rays to estimate the source size geometrically - the longer the eclipse lasts and the shallower it is, the bigger the X-ray source or photosphere. 

We used X-ray eclipse observations of U Sco made by the Suzaku X-ray satellite, combined with eclipses observed by the XMM-Newton satellite, to examine the rate of change of the photospheric size.  Since the size is directly related to the mass loss rate, we can see how the mass loss evolves in time. This is the first application of this direct geometrical technique to get at the mass loss evolution in a novae event. We found that the total mass lost to the radiatively-driven flow would have been about 10-6 - 10-7 solar masses. These estimates are very important for U Sco because it remains a contender to be a supernova type 1a progenitor if the white dwarf is gaining mass rather than gradually losing it through nova explosions.  We still do not know what objects end their lives as type 1a supernovae, yet they remain crucial cosmologically for probing and understanding dark energy. If U Sco is a mass gainer, the white dwarf will eventually explode as the increasing pressure causes its core temperature and density reach values able to ignite catastrophic carbon and oxygen fusion. This work was lead by postdoc Dai Takei and was published in the 2013 May 20 edition of the Astrophysical Journal Letters.