Twinkling little stars blink at XMM-Newton

XMM-Newton has obtained new X-ray data from a particularly interesting stellar system, the sixtuple star Castor. The observation was obtained as part of the Performance Verification phase and permits a deep check of all X-ray detectors. With this data set, XMM-Newton has passed challenging tests on high-resolution and time-resolved imaging, time-resolved spectroscopy, and spatially resolved spectroscopy in one strike.

Castor, the brightest star in the constellation of Gemini The Twins, is an inconspicuous star of spectral type A; it actually consists of a pair of A-type stars in an orbit with a period of 467 yrs, with an apparent separation of presently 3.9 arcseconds, just resolvable by an amateur optical telescope. A bit more than an arcminute south of this pair, the well-studied M-type dwarf star, YY Gem, has attracted astronomers for its frequent flare outbursts. To complicate matters, each of the three stars is a binary system by itself, each with very intriguing and different properties. The brighter of the A stars is accompanied by an invisible dwarf star in a highly eccentric orbit, while the other A star hosts a similar companion but apparently in a circular orbit. YY Gem, finally, is made up of two identical M dwarfs that orbit each other in a matter of only 19 hours. As we look at this system almost exactly from the side, every 9.5 hours one of the stars is almost completely eclipsed by its companion.

But the Castor star family reveals its real splendor in the X-rays. XMM-Newton observed the whole system on April 25, during about 1 full day, using all three EPIC cameras and both Reflection Grating Spectrometers (RGS). While it has been known that both YY Gem and some source among the two A stars are bright X-ray objects, XMM-Newton tells conclusive stories about who does what in this system. The EPIC MOS image, with pixel sizes of 1.1 arcseconds, clearly reveals two X-ray sources at the position where the two Castor A-type stars are supposed to be. This resolves a long-standing debate on whether both stars (or rather their respective low-mass companions) are magnetically active. Previous ROSAT observations addressed this problem and suggested the optically brighter system to be the likely X-ray source, although some ambiguity remained. The least ambiguous signal actually came from radio astronomy. Very Large Array observations conducted by one of the XMM-Newton/RGS Team members in 1994 conclusively showed that both stars are radio sources. The observing radio team predicted, on theoretical grounds, that both components will show up as X-ray sources. XMM-Newton has nicely confirmed their model in one shot.

Deep science is now being performed with the two RGS. The RGS system convincingly demonstrates that it can capture two high-resolution spectra for the price of one. The separation between the Castor system (both X-ray sources combined) and YY Gem nicely places two fully separated high-resolution spectra onto the RGS CCD strip. The observation profited from careful roll angle placement to avoid overlap between the spectra.

The RGS CCD detector image shows, in the middle part, the well-exposed line spectrum of the YY Gem binary, while the upper spectrum stems from the somewhat weaker Castor system. The multitude of spectral lines are now being used to explain the temperature structure and clarify abundance anomalies in the atmospheres.

The two extracted first-order spectra appear strikingly different. The ratio between the line fluxes indicates either different dominant temperatures or different elemental composition (e.g., the abundances of Ne and Fe) of the Castor and YY Gem sources. Precise modeling is currently underway, and requires up-to-date calibration parameters that are now becoming available.

Given the resolving power of the RGS of up to about 800, spectral line analysis will allow, after the conclusion of the present calibration campaign, analysis of motions of the order of 100 km/s in this system. The unprecedented effective area of the RGS is of prime importance to obtain time-resolved spectroscopy during X-ray flares.

Indeed, movies of the EPIC field of view covering the observing time reveal that all three X-ray sources are `blinking', indicating that all of them are flaring stars. Flares are giant energy releases that evolve in minutes to hours as a consequence of magnetic instabilities in the hot outer stellar atmospheres. While well studied on the Sun and other stars, the frequency of flaring on Castor is quite surprising. There is no period during the 90 ksec of observing time when the emission is at a constant level, perhaps indicating that much of the observed X-ray radiation stems from such eruptions. Some of the most exciting science will come from a time-dependent analysis of spectra taken during the flaring episodes. An early spectral analysis clearly shows that the plasma considerably heats up during these episodes.

YY Gem also nicely shows X-ray eclipses every 9.5 hours. Somewhat less than half of the X-ray light disappears for approximately one hour. Detailed geometric modelling of the eclipse dips will provide unique information on the size of the coronae.

Bert Brinkman of SRON/The Netherlands is the Principal Investigator of the Reflection Grating Spectrometer. Co-Investigator institutes include Columbia University/USA, MSSL/UK, and PSI/Switzerland. Martin Turner of Leicester University in the UK is the Principal Investigator of EPIC. Co-Investigator institutes are: IFC Milan, MPE Garching, CEA Saclay, AI Tuebingen, Birmingham University,IAS Orsay, CESR Toulouse, ITESRE Bologna, OA Palermo.

The YY Gem and Castor results from XMM-Newton and from radio observatories are further being analyzed by a team under the lead of Manuel Güdel and Marc Audard from PSI/Switzerland as part of the RGS Performance Verification Phase.

Figures:

1. Extraction of EPIC MOS1 image of part of the observation. The lower, brighter stellar image is from YY Gem. The upper image shows the two flaring Castor components.

2. For comparison, a deep radio observation of the Castor system reveals that all three sources are also active at radio wavelengths. This emission is thought to be produced by high-energy electrons in magnetic fields; they seem to be related to the heating mechanism that produces the detected X-rays.

3. RGS spectra shown as a function of dispersion angle (x axis) and CCD pixel energy (y axis). The first order spectrum shows up in the brighter, lower bright bow, second order is fully separated in the upper bow.

4. Extracted 1st order spectra of YY Gem (top) and Castor AB (bottom). Note some differences in the line flux ratios.

5. Image of the detector plane of RGS1. The spectra are dispered across 9 CCDs in horizontal direction. The lower spectrum is from YY Gem, the upper one from the Castor binary.