X-ray observations of polars

Polars or AM Her systems are accreting binary systems in which material transfers from a dwarf secondary star onto a magnetic (B~10-200MG) white dwarf through Roche lobe overflow. For polars in a high accretion state, the accretion flow forms a strong shock at some height above the photosphere of the white dwarf. The maximum temperature in the post-shock flow is set by the mass of the white dwarf. For a 0.7Msun white dwarf the shock temperature is ~30keV, with the temperature decreasing as the gas settles onto the white dwarf. Some fraction of the hard X-rays intercept the photosphere of the white dwarf, are thermalised and then re-radiated as soft X-rays or in the extreme UV. Polars, are ideal targets with which to investigate the accretion process in detail.

We have undertaken a survey of polars using XMM-Newton. This survey contains observations of 37 polars -- more than half of all known systems. Our survey has led to various papers on individual systems, which concentrated on phase resolved spectroscopy, as well as papers which have looked at the energy balance of these systems and those systems which were observed in low accretion states. Copies of our papers can be found in our pre-print web page.

X-ray light curves

The X-ray light curves of polars allow us to determine the system geometry of these objects. Further, by comparing the soft and hard X-ray light curves we can obtain insight into the location of the accretion stream - which preferentially absorbs softer X-rays. One interesting example is the eclipsing system EP Dra. In this system, the main accretion comes into and out of view as the binary system rotates on its 105 min binary orbital period.

Prior to these X-ray observations, we obtained optical data of EP Dra using the Superconducting Tunneling Junction (STJ) array on the WHT . From that data we predicted that the accretion stream attaches onto a wide range of magnetic field lines and that this would result in soft X-rays being preferentially absorbed in the phase interval before the eclipse (Bridge et al 2002) . The X-ray observations (above) confirm this prediction, showing that in this system, soft X-rays start to be obscured shortly after the start of the bright phase. Further, there is a dense core of material in the accretion stream. See XMM-Newton observations of the eclipsing polar EP Dra (Ramsay et al 2004) for details

X-ray spectra

It is important for understanding the accretion physics to clarify the relationship between the different emission, absorption and reflection components in these systems, and their dependence on the bolometric luminosity. At MSSL we have therefore begun the process of modelling the X-ray emission from mCVs in much more detail. This has entailed the construction of optically thin bremsstrahlung and line spectra from the multi-temperature cooling flow of material as it settles onto the surface of the white dwarf. The temperature structure takes into account the additional cooling at each point resulting from the cyclotron emission from electrons spiralling in the intense magnetic field. In addition, the contribution of the Compton reflection from the surface of the white dwarf is included. The two spectra in the figure below indicate the difference between the emitted flux for a typical 40 MegaGauss magnetic field and one with zero field.

We show below the difference in the fits to the XMM-Newton RGS spectra of the intermediate polar EX Hya using our 'stratified accretion column' model and those models which are usually used to fit X-ray spectra. Our stratified model provides a significantly better fit than that obtained by using a 3-temperature thermal plasma model.

The X-ray energy balance

Using the most physical model for the emission region is very important when determining the ratio of the reprocessed component and the shocked component. In the EXOSAT and ROSAT eras a significant number of polars were found to show a soft/hard X-ray ratio much greater than that expected from the standard accretion shock model. This was known as the `soft X-ray excess'. We have made an snapshot survey of polars using XMM-Newton and determined their soft/hard ratios. We find that less than one in five of systems show a significant soft X-ray excess, while the rest show ratios consistent with that predicted by the standard model. We have investigated the discrepancy between this and the previous investigations by re-examining all the available ROSAT PSPC pointed observations of polars using more recent calibrations than in the original studies. We find that these data show an energy balance ratio which is broadly consistent with that of our XMM-Newton results. We conclude that the previous studies were affected by the data being less well calibrated. We discuss which physical mechanisms might give rise to a high soft X-ray excess and whether systems with high ratios show more variation in soft X-rays. Surprisingly, we find that 6 out of 21 systems found in a high accretion state did not show a distinct soft X-ray component. Two systems showed one pole with such a component and one which did not. Based on the ratio of the observed soft X-ray to UV flux measurements (which were obtained simultaneously using the Optical Monitor) we suggest that this is because the reprocessed component in these systems is cool enough to have moved out of the soft X-ray band and into the EUV or UV band. See The energy balance of polars revisited (Ramsay & Cropper 2004) for details

Those systems in a low accretion state

Off the 37 polars which were observed as part of our XMM-Newton survey, we found that 16 of these systems were in a low, or much reduced, accretion state. Of those, 6 were not detected in X-rays. This suggests that in any survey of polars, around half will be in a low accretion state. We tested if there was a bias towards certain orbital periods: this is not the case. Of the 10 systems which were detected at low, but significant rates in X-rays, 8 showed significant variability in their X-ray light curves. This implies that non-uniform accretion still takes place during low accretion epochs. The bolometric luminosity of these systems is ~10^30 ergs/s, two orders of magnitude less than for systems in a high accretion state. The X-ray spectra show no evidence of a distinct soft X-ray component. However, the X-ray and UV data imply that such a low temperature component exists: its temperature is low enough for its flux distribution to move outside the bandpass of the X-ray instruments. See XMM-Newton observations of polars in low accretion states (Ramsay et al 2004) for details

Gavin Ramsay (Aug 2004)