Radio Images of a Large Stellar Coronal Loop on Algol
Total Page:16
File Type:pdf, Size:1020Kb
Radio Images of A Large Stellar Coronal Loop on Algol W. M. Peterson*, R. L. Mutel*, M. Gudel**,¨ W. M. Goss† *Department of Physics and Astronomy, University of Iowa, Van Allen Hall, Iowa City, Iowa 52240, USA **ETH Zurich, Institute of Astronomy, 8093 Zurich, Switzerland †National Radio Astronomy Observatory, Pete V. Domenici Science Operations Center, 1003 Lopezville Road, Socorro, New Mexico 87801, USA The close binary Algol contains a radio-bright K subgiant star in a very close (0.062 AU), rapid (2.86 day) orbit with a main sequence B8 star. Since the rotation periods of the two stars are tidally locked to the orbital period, the consequent rapid rotation drives a robust magnetic dynamo. A large body of evidence points to the existence of an extended, complex coronal magnetosphere originating at the cooler K subgiant1–4. The detailed morphology of the subgiant’s corona and its possible interaction with its companion are unknown, though theory predicts that the coronal plasma should be confined in magnetic loop structure5, as seen on the Sun. Here we report multi-epoch radio imaging of the Algol system, in which we see a large, persistent coronal loop approximately one subgiant diameter in height, whose base is straddling the subgiant and whose apex is oriented toward the B star. This strongly suggests that a persistent asymmetric magnetic field structure is aligned between the two stars. The loop is larger than anticipated theoretically6, 7, but the size may be a result of a magnetic interaction between the two stars. 1 We made six twelve-hour observations of Algol during a period from 6 April to 17 August 2008 with the High Sensitivity Array (HSA), a global very long baseline interferometer array. For these observations, the HSA consisted of the ten 25-meter very long baseline array (VLBA) antennas, the 100-meter Green Bank Telescope, the 100-meter Effelsberg radiotelescope, and the Very Large Array in phased array mode. We observed at a frequency of 15 GHz, a crucial choice since it provided a resolving beam size (0.4 mas x 0.6 mas) which was half the size of most previous VLBI observations2, 8–12 and much smaller than the projected orbital separation (2.3 mas). This made it possible to to discern the detailed radio morphology of the system and its association with the individual components. The angular motion of the less-massive KIV subgiant (here denoted as Algol B) during a twelve-hour period is significant compared with the resolving beam, particularly for eclipse ob- servations when the star moves by as much as 2 milliarcseconds, crossing several beamwidths. To compensate for this, we applied a small, time-based position correction to the data to cancel out the motion of Algol B so that the phase fitting algorithm tracked the K star center position. We determined a global astrometric solution for proper motion, parallax, and fiducial reference position by combining the centroid position from all six epochs with twelve additional radio posi- tions from previous VLBI observations2, 9, 13, 14 and unpublished data from the VLBA archive. All observations serendipitously used the same phase reference source J0313+412. We determined the proper motion, parallax and fiducial position of the barycenter of the Algol using orbital pa- rameters well-determined by spectroscopy15 and optical interferometry16. Fig. 1 shows that the radio centroid closely tracks the position of Algol B, confirming the previous result of Lestrade 2 et al.14. This is also supported by X-ray eclipse observations in which strong flares have been occulted during secondary eclipse17, 18. More importantly, by explicitly assuming that there is no systematic offset (constant at all epochs) between Algol B and the radio centroid position, the radio centroid represents the position of Algol B in the International Celestial Reference Frame (ICRF) to ±0.38 mas accuracy. This is significantly more accurate than recent astrometric surveys of radio star positions13, 19 and may be the most accurate determination of a stellar position ever made. For all epochs, the radio morphology consisted of either an elongated double structure or an entire loop, with angular scale comparable to the diameter of Algol B. We interpret these images as compelling evidence for a single large coronal loop structure with feet fixed on the polar regions of Algol B and its apex co-moving with and approximately oriented along the line of sight to the main sequence companion (Algol A). We note that the term ’single loop’ is resolution-dependent: At higher angular resolution, the physical configuration may have a more complicated topology e.g., filamentary structures, such as suggested for solar loops20. In order to support the large coronal loop interpretation, we modeled the emission that would be produced by a theoretical coronal loop filled with energetic electrons emitting gyrosynchrotron radiation. Assuming a power-law distribution of electron energies, we determined the flux density at each point on the loop by solving for the specific intensity using approximate expressions21 for gyrosynchrotron emission and absorption coefficient. We also assumed a uniform electron density, since the scale height for free electrons on Algol B is several stellar radii. We were able to best match the observed emission patterns with a surface magnetic field strength on the order of 3 103 Gauss and electron densities of ∼ 103 cm−3, as shown for a sample epoch in Fig. 3. A large, quasi-stable, pole-oriented coronal loop structure anchored on Algol B appears to be the dominant feature of the magnetosphere of this star. This diverges from what we know of the Sun, where most flaring and coronal loop activity is seen in relatively compact magnetic structures close to the equatorial regions22. It is more reminiscent of global magnetospheres as seen on planets and some low-mass cool stars23. This large coronal loop, co-rotating and oriented toward the inner binary’s center of mass, may provide a natural explanation for a number of previous observations of coronal emission from Algol and perhaps other similar active binaries. For example, observations of Algol’s quiescent X-ray light curve over three consecutive binary orbits24 showed a repeating periodic pattern with a phase indicating that the hemisphere of Algol B facing Algol A is more active than the other. This is expected if the X-ray corona is confined to a large loop oriented toward Algol A. Periodic Doppler shifts in several X-ray lines were observed that were smaller than expected from a source located very near Algol B1, suggesting an asymmetric corona biased toward the system center of mass. In addition, the line widths were larger than expected from rotation and thermal broadening, suggesting a radially extended corona on the order of a stellar radius. A giant X-ray flare17 was occulted during secondary eclipse, allowing an accurate determination of its location, which was found to be just above the south pole of Algol B. This is also consistent with a coronal loop scenario in which the coronal feet are located near the poles of Algol B. Finally, the orientation of the coronal loop toward the center of mass may be a result of the magnetic topology in which a magnetically threaded accretion disk around Algol A25 interacts with Algol B’s magnetic field, causing reconnection and consequent electron acceleration toward the polar regions 4 at the feet of the coronal loop. There is some evidence for this scenario based on observations of periodic ’super humps’ in the long-term radio light curves4, 26. Coronal loop models with sizes several time the the active star’s radius have been proposed to explain coronal radio emission from active binary systems27, 28. If these structures are bright emission features of a dipolar magnetic structure, it suggests the presence of a globally operating magnetic dynamo. The similarities in non-flare radio and X-ray emission, including overall size24 and orientation toward Algol A1 suggest that both plasmas are driven by a common energy source, and may be largely co-spatial. A major future challenge is the construction of a self-consistent model to test whether both non-thermal radio and possibly thermal X-ray plasma may co-exist in large loops with feet tied to the poles of the active star. 1. Chung, S. M., Drake, J. J., Kashyap, V. L., Lin, L. W. & Ratzlaff, P. W. Doppler shifts and broadening and the structure of the X-ray emission from Algol. Astrophysical Journal 606, 1184–1195 (2004). 2. Mutel, R. L., Molnar, L. A., Waltman, E. B. & Ghigo, F. D. Radio emission from Algol. I. Coronal geometry and emission mechanisms determined from VLBA and Green Bank Inter- ferometer observations. Astrophysical Journal 507, 371–383 (1998). 3. Ness, J. U., Gudel,¨ M., Manuel, Schmitt, J. H. M. M., Audard, M. & Telleschi, A. On the sizes of stellar X-ray coronae. Astronomy and Astrophysics 427, 667–683 (2004). 4. Retter, A., Richards, M. T. & Wu, K. Evidence for superhumps in the radio light curve of 5 Algol and a new model for magnetic activity in Algol systems. Astrophysical Journal 621, 417–424 (2005). 5. Aschwanden, M. J., Poland, A. I. & Rabin, D. M. The new solar corona. Annual Review of Astronomy and Astrophysics 39, 175–210 (2001). 6. Mullan, D. J., Mathioudakis, M., Bloomfield, D. S. & Christian, D. J. A comparative study of flaring loops in active stars. Astrophysical Journal Supplement Series 164, 173–201 (2006). URL http://adsabs.harvard.edu/abs/2006ApJS..164..173M. 7. Rosner, R., Tucker, W. H. & Vaiana, G. S. Dynamics of the quies- cent solar corona. Astrophysical Journal 220, 643–645 (1978). URL http://adsabs.harvard.edu/abs/1978ApJ...220..643R.