Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://star.arm.ac.uk/preprints/2009/537.pdf Äàòà èçìåíåíèÿ: Wed Jan 21 14:35:15 2009 Äàòà èíäåêñèðîâàíèÿ: Tue Oct 2 02:51:27 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: ñîçâåçäèå ëåáåäÿ

Mon. Not. R. Astron. Soc. 000, 000­000 (0000)

Printed 20 January 2009

A (MN L TEX style file v2.2)

2XMMi J225036.9+573154 ­ a new eclipsing AM Her binary discovered using XMM-Newton
Gavin Ramsay1, Simon Rosen2, Pasi Hakala3, Thomas Barclay 1
2 3 4

1,4

Armagh Observatory, Col lege Hil l, Armagh, BT61 9D Department of Physics and Astronomy, University of Tuorla Observatory, University of Turku, V¨is¨l¨ntie a aa Mul lard Space Science Laboratory, University Col lege

G Leicester, University Road, Leicester LE1 7RH 20, FIN-21500 Piikki¨, Finland o London, Holmbury St. Mary, Dorking, Surrey, RH5 6NT

20 January 2009

ABSTRACT

We report the discovery of an eclipsing polar, 2XMMi J225036.9+573154, using XMMNewton. It was discovered by searching the light curves in the 2XMMi catalogue for ob jects showing X-ray variability. Its X-ray light curve shows a total eclipse of the white dwarf by the secondary star every 174 mins. An extended pre-eclipse absorption dip is observed in soft X-rays at =0.8­0.9, with evidence for a further dip in the soft X-ray light curve at 0.4. Further, X-rays are seen from all orbital phases (apart from the eclipse) which makes it unusual amongst eclipsing polars. We have identified the optical counterpart, which is faint (r=21), and shows a deep eclipse (>3.5 mag in white light). Its X-ray spectrum does not show a distinct soft X-ray component which is seen in many, but not all, polars. Its optical spectrum shows H in emission for a fraction of the orbital period. Key words: Stars: binary - close; novae - cataclysmic variables; individual: - 2XMMi J225036.9+573154; X-rays: binaries

1

INTRODUCTION

Cataclysmic Variables are accreting binary systems in which a white dwarf accretes material from a late typ e main sequence star through Roche lob e overflow. If the white dwarf has a significant magnetic field then the formation of an accretion disk can b e disrupted or prevented. For white dwarfs with field strengths greater than 10 MG, the accretion stream gets channelled onto the magnetic p oles where X-rays are emitted from the p ost-shock region. The magnetic field also forces the spin p eriod of the white dwarf to synchronise with the binary orbital p eriod. These accreting binaries are called AM Her binaries or p olars, since their optical emission is strongly p olarised. The study of p olars was transformed with the launch of the X-ray satellite ROSAT in 1990. Prior to this, around 17 systems were known. ROSAT led directly to the discovery of around 30 new systems (eg Beuermann & Burwitz 1995). It was exp ected that XMM-Newton, launched in 1999, would lead to the discovery of many more such systems. Surprisingly, comparatively few have so far b een discovered. The 2XMM catalogue (Watson et al 2009) gives a description of serendipitous X-ray sources discovered using the EPIC wide-field instruments on b oard XMM-Newton. This was followed by the release of the 2XMMi incremental catalogue which has 17 p ercent more discrete sources than the
c 0000 RAS

2XMM catalogue. Moreover, each source is accompanied by source sp ecific light curve and sp ectral products. In this pap er we rep ort the discovery of an eclipsing p olar, 2XMMi J225036.9+573154, which was found as a result of searching the 2XMMi catalogue for sources which showed variability in their X-ray light curve.

2 2.1

XMM-Newton OBSERVATIONS The 2XMMi catalogue

The 2XMMi catalogue has associated sp ectra and light curves that are automatically extracted by the XMMNewton Science Survey Centre pip eline processing software (Watson et al 2001) for sources with more than 500 counts in the EPIC detectors. An assessment of variability in the in2 dividual light curves is made by determining of the data ab out the mean, and then computing the consequent probability of the constant (null) hyp othesis. Those light curves for which this probability is < 10-5 are deemed variable. Sources which were p ossibly compromised by further data quality issues were removed. An initial search of the catalogue found around 400 sources which passed these criteria. The light curves of these sources were visually insp ected for p eriodic b ehaviour. One source, 2XMMi J225036.9+573154


2
0.12 0.10 EPIC Ct/s 0.08 0.06 0.04 0.02 0.00 0.04 0.03 EPIC Ct/s 0.02 0.01 0.00 0.08 EPIC Ct/s 0.06 0.04 0.02 0.00 0.04 0.03 EPIC Ct/s 0.02 0.01 0.00 0 1 Mag 2 3 4 0.0 0.5 1.0 Phase 1.5 2.0

0.2-10keV

0.2-1keV

2-10keV
Figure 2. The power spectrum of the combined EPIC (pn plus MOS) 0.2­10keV light curve.

4-10keV

Optical

The data were processed using XMM-Newton SAS v8.0.1 (released Oct 2008). Only X-ray events which were graded as PATTERN=0-4 and FLAG=0 were used. Events were extracted from a circular ap erture with 10 radius centred on the source, with background events b eing extracted from source free areas on the same chip as the source. The background data were scaled to give the same area as the source extraction area and subtracted from the source area. (We estimate that the nearby source XMM J225037.9+573127 contributes around 1.5 p ercent of the flux b elow 2keV, and a negligible amount at energies ab ove 4keV). To ensure that the sp ectra were correctly flux calibrated we produced detector sp ectral resp onse files and ancillary files using the SAS tasks rmfgen and arfgen resp ectively.

Figure 1. The light curves of XMM J2250+5731 folded on a period of 174.2 mins and To =2454124.245888 (TT). From the top we show the combined EPIC pn plus EPIC MOS light curve in the 0.2­10keV energy band; the 0.2­1.0keV energy band; the 2­10keV energy band; the 4­10keV energy band (binned into 100 bins) and in the lower panel the white light data obtained using the NOT (the data have been folded but not binned).

3

THE X-RAY LIGHT CURVES

(hereafter XMM J2250+5731), was found which showed a characteristic rep eating shap e on a p eriod of 174 min (Figure 1).

2.2

Pointed observations

XMM J2250+5731 was found in the field of G107.5-1.5 which was observed on 23rd Jan 2007. The EPIC detectors were each configured in full window mode and used the medium filter. The field was observed for a total of 32.9 ksec in the EPIC pn detector and 34.5 ksec in b oth EPIC MOS detectors. The source was just outside the field of view of the Optical Monitor. Since the source was towards the edge of the EPIC detectors, and there was a nearby (28 ) X-ray source, XMM J225037.9+573127, which app ears to b e an active late-typ e star, we extracted the data from the XMMNewton archive and re-extracted the X-ray light curves and sp ectra of XMM J2250+5731.

We extracted light curves of XMM J2250+5731 in the 0.2­ 10keV, 0.2­1.0keV, 2­10keV and 4­10keV energy bands from the EPIC pn, EPIC MOS1 and EPIC MOS2 detectors using the method describ ed ab ove. We then obtained a combined light curve for each energy band by adding the separate light curves. Each light curve shows a distinctive sharp drop in intensity every 174 min. This is due to the secondary star eclipsing the accretion region(s) on the white dwarf and represents the binary orbital p eriod. The observation covers 3 eclipses. We used the standard Lomb-Scargle p ower sp ectrum analysis to search for p eriods in the data (Figure 2). The error on the p eriod was then determined using a b ootstrap approach incorp orating the generation of synthetic light curves. We find that the p eriod is 0.1210±0.0018 days (=174.2±2.6 mins). We folded the light curve in each of the 4 energy bands on this p eriod and show these light curves in Figure 1. We have phased the data so that the eclipse, which is total in each energy band, defines =0.0. XMM J2250+5731 is relatively faint in X-rays, reaching a p eak of 0.08 ct/s in the combined EPIC 0.2­10keV light curve, although this count rate has not b een corrected for the source b eing far off-axis. Prior to the eclipse, there is a marked decrease in soft
c 0000 RAS, MNRAS 000, 000­000


2XMMi J225036.9+573154 ­ a new AM Her binary
X-ray photons over the phase range 0.7­1.0, compared to those at higher energies. This phenomenon has b een seen in other p olars (eg Watson et al 1989) and occurs when the accretion stream obscures our view of the hot accretion region located in the upp er hemisphere on the white dwarf. Compared to V2301 Oph (for instance, Ramsay & Cropp er 2007), the `pre-eclipse' dip seen in XMM J2250+5731 is more extended suggesting that material gets lifted out of the orbital plane over a wider range in azimuth. At softer energies (<1keV) there is also a dip in the light curve centered at 0.4 and with a duration of 0.1­ 0.2 cycles. At higher energies, there is no obvious broad dip at these orbital phases although there are a couple of bins b etween =0.4­0.5 which are consistent with zero counts. However, since other bins with negligible flux are also seen at different phases this may just b e due to low counting statistics. This dip could either b e due to a second dip caused by an accretion stream or it could b e due to the rotation of the accretion regions rotating into and out of view as the white dwarf rotates. We will discuss this further in §7.
NH Fluxo Fluxo Fluxo Fluxu Fluxu Fluxu 2 5+0.3 -0.4 8+0.4 -0.4 1+0.4 -0.3 1+0.9 -1.0 4+1.0 -0.9 8+0.9 -0.8 3.4+2.2 â 1020 cm-2 -1.8 10-13 erg s-1 cm-2 10-13 erg s-1 cm-2 10-13 erg s-1 cm-2 10-13 erg s-1 cm-2 10-13 erg s-1 cm-2 10-13 erg s-1 cm-2 1.12 (34 dof )

3

EPIC EPIC EPIC EPIC EPIC EPIC

pn mos1 mos2 pn mos1 mos2

3. 2. 2. 8. 6. 4.

â â â â â â

Table 1. The spectral fit to the EPIC pn, MOS1 and MOS2 spectra extracted from =0.05­0.7. Fluxo refers to the observed flux measured over the 0.2­10keV energy band and Fluxu refers to the unabsorbed bolometric flux.

4

X-RAY SPECTRAL FITS

We extracted sp ectra from each EPIC detector in the manner describ ed in §2. Initially we extracted sp ectra using all the available data. However, since the light curves (cf Figure 1) imply the presence of a pre-eclipse dip, we then extracted sp ectra from the phase interval which was not strongly affected by absorption, ie 0.05­0.7. We also exclude the phase interval =0.38­0.5 which could also b e affected by absorption (§3). In p olars, X-rays are generated in a p ost-shock region at some height ab ove the photosphere of the white dwarf. Since the X-ray sp ectrum of XMM J2250+5731 has a relatively low signal to noise compared to many p olars previously studied using XMM-Newton (eg Ramsay & Cropp er 2004), we used a simple single temp erature thermal bremsstrahlung emission model rather than a more complex (and more physical) stratified cooling flow model (eg Cropp er et al 1998, 1999). We used the XSPEC package (Arnaud 1996) to fit the Xray sp ectra. We fitted all three EPIC sp ectra simultaneously and tied the sp ectral parameters apart from the normalisation parameters. We used the tbabs absorption model (the Tubingen­Boulder absorption ISM model, Wilms, Allen & ¨ McCray 2000), a single temp erature thermal bremsstrahlung comp onent with temp erature fixed at kT = 20keV. We added a Gaussian comp onent to account for any emission b etween 6.4­6.8keV. The sp ectra along with the b est fit (2 =1.12) are shown in Figure 3. We show the sp ectral pa rameters, the observed and unabsorb ed b olometric fluxes in Table 1. Due to the low signal to noise of the sp ectra, the equivalent width of the Fe K emission line features was p oorly constrained. In many p olars, a strong soft X-ray comp onent (kTbb 40eV) is seen in their X-ray sp ectra (eg Ramsay et al 1994, Beuermann & Burwitz 1995). This is due to the hard X-rays irradiating the photosphere of the white dwarf which are then re-emitted as soft X-rays. The standard accretion model predicts that Lsof t /Lhard 0.5, where Lsof t and Lhard are the luminosities of the soft and hard X-ray
c 0000 RAS, MNRAS 000, 000­000

comp onents resp ectively (Lamb & Masters 1979, King & Lasota 1979). Although the energy balance in p olars was a source of great debate for many years, Ramsay & Cropp er (2004) showed that the ma jority of p olars in a high accretion state have X-ray sp ectra which are in good agreement with the standard model. To see if such a soft X-ray comp onent could b e `hidden' by the moderate level of absorption (cf Table 1) we added a blackb ody with a range of different temp eratures. We fixed its normalisation so that the implied ratio, Lsof t /Lhard 0.5. Since the soft X-rays are optically thick, and hence the intrinsic soft X-ray luminosity is viewing angle dep endant, we assumed a viewing angle of 45 for argument. If we just consider the X-ray data, we find a blackb ody with temp erature less than kT <20eV can easily b e hidden. We were fortunate in b eing able to obtain a short observation of the field of XMM J2250+5731 using Swift on the 3rd and 4th Dec 2008. Observations using the UV Optical Telescop e (Roming et al 2005) were made using the UVW2 filter (p eak effective wavelength 2120°). XMM J2250+5731 A was not detected, and we estimated a 3 upp er limit of 2.4 â 10-17 erg s-1 cm-2 °-1 . If we assume a blackb ody A of different temp eratures and with a normalisation such that Lsof t /Lhard 0.5, we find that a blackb ody of kT 5­20eV can b e present and not detected in the near UV or soft X-ray energy ranges. (Although a handful of X-ray events were detected near the source p osition of XMM J2250+5731, they were too low to derive any meaningful information). The unabsorb ed b olometric flux implies an X-ray luminosity of 8 â 1029 d200 erg/s, where d200 is the distance 1 1 in units of 100 p c. Ramsay & Cropp er (2004) found that the mean b olometric luminosity in their sample of p olars observed in a high state using XMM-Newton was 2 â 1032 erg/s. In the next section we find that XMM J2250+5731 shows a range in optical brightness over the longer term and therefore a range of accretion states (a general characteristic of p olars). Assuming that XMM J2250+5731 was observed in a high accretion state at the ep och of the XMM-Newton observations we find that in order that XMM J2250+5731 has an X-ray luminosity consistent with other p olars in a high state it must lie at a distance of 1.5­2.0 kp c. With Galactic co-ordinates of l = 107.2 and b = -1.6 , this places XMM J2250+5731 close to the Perseus spiral arm (Xu et al 2005).


4

normalized counts/sec/keV sign(d-m)*2

-2

0

2

4

10-3

0.01

0.5

1 2 channel energy (keV)

5

Figure 3. Upper Panel: The EPIC spectra extracted from the bright phase interval along with the best fits (the upper most spectrum is that derived from the EPIC pn, while the lower ones are derived from the EPIC MOS 1 and 2 detectors). Lower Panel: the residuals to the best fit are shown in units of 2 . Figure 4. The finding chart of XMM J2250+5731 extracted from a g band image taken with the INT WFC on 6th Nov 2008.

5

OPTICAL PHOTOMETRY

To locate the optical counterpart of XMM J2250+5731 we obtained optical photometry using ALFOSC on the Nordic Optical Telescop e (NOT) located on La Palma during 28th Sept 2008. Each exp osure was in `white light' and 15 sec in length, with another 5 sec of readout time, resulting in 2.9 h of data in total. Each source in the field (Figure 4) was searched for variability. One source showed a clear eclipse lasting for 12 mins and a depth >3.5 mag (Figure 1). This is the optical counterpart to XMM J2250+5731 and its co ordinates are 2000 = 22h 50m 36.97s , 2000 = +57 31 54.2 (which is within 0.8 of the X-ray p osition). We searched the IPHAS catalogue (Drew et al 2005) which surveyed the northern galactic plane in r, i,H filters to determine if the optical counterpart of XMM J2250+5731 was detected in this survey. We find that IPHAS gives r = 20.32 ± 0.05, i = 20.1 ± 0.2, H = 19.69 ± 0.09 for XMM J2250+5731. All sources within a 30 radius of the X-ray p osition were extracted. XMM J2250+5731 is at the extreme blue end in the (r - i) distribution and consistent with the location of the CVs found in IPHAS data in the (r - i), (r - H ) colour-colour plane (cf fig1 of Corradi et al 2008). We also obtained U, g , R images of XMM J2250+5731 using the Wide Field Camera on the Isaac Newton Telescop e on 6th Nov 2008: Figure 4 shows the g band image of the immediate field. Using standard star observations taken immediately b efore these observations we find that U = 21.75±0.11, g = 21.16 ± 0.05 and r = 21.53±0.05. Compared to other stars in the field, it is clearly blue and app ears to b e more than 1 mag fainter than found at the ep och of the IPHAS p ointings. This is not unexp ected since p olars are known to show different accretion states.

6th Oct 2008. We used the R300B and R158R gratings giving a sp ectral resolution of 2.5° and 5° resp ectively. A A The seeing was 0.8 and the slit was set to match the seeing. We took 16 sp ectra in b oth the red and blue arms. With an out of eclipse brightness of r 21, each individual sp ectrum was of low signal to noise. Moreover, in the blue arm, there was electronic noise in the images, the pattern of which varied from image to image. This coupled with the low signal to noise of the sp ectra prevented us from extracting any useful information from the blue arm. In the red arm, we were able to extract a sp ectrum from each image. For 9 sequential sp ectra we were able to detect H in emission. We show the mean of these sp ectra in Figure 5. For the remaining 7 sp ectra for which we did not detect H in emission we attribute this to the fact that the observations occurred during the phase interval of the pre-eclipse absorption dip or that the accretion stream was presenting a small surface area at those phase intervals. (Given the error on the orbital p eriod, §3, the phasing of the WHT sp ectra using the NOT photometric observations as a marker of the phasing is uncertain by approximately one orbital cycle).

7 7.1

DISCUSSION The X-ray light curve

6

OPTICAL SPECTROSCOPY

We obtained sp ectra of XMM J2250+5731 using the 4.2m William Herschel Telescop e and the Intermediate disp ersion Sp ectrograph and Imaging System (ISIS) on La Palma on

In p olars, it is thought that the magnetic axis of the white dwarf is tilted towards the secondary star, but shifted a few 10's of degrees ahead in azimuth (as the binary rotates) of the line of center joining the two stars (eg Cropp er 1988). It is therefore the accretion region in the upp er hemisphere which is obscured by the accretion flow during the preeclipse absorption dip. Eclipsing p olars have long b een the target of dedicated X-ray observations. Many of these p olars show a distinct bright and faint phase as the accretion region rotates into
c 0000 RAS, MNRAS 000, 000­000


2XMMi J225036.9+573154 ­ a new AM Her binary
350 300 250 Counts 200 150 100 50 0 6000

5

6200

6400 6600 Wavelength (A)

6800

7000

Figure 5. In 9 our of our 16 spectra taken of XMM J2250+5731 using the WHT and ISIS in Oct 2008, we detected the H emission line. This figure shows the mean of these 9 spectra. ISIS and the WHT on 6th Oct 2008. The flux scale is arbitrary.

view and out of view and many show a characteristic preeclipse absorption dip. In these systems there is no evidence for a second accretion p ole. One of the few eclipsing p olars to show emission throughout the binary phase is V2301 Oph (Ramsay & Cropp er 2007). We find that in the case of XMM J2250+5731, X-ray emission is also seen throughout the orbital phase. We attempted to invert the X-ray light curves and map the X-ray regions on the white dwarf using an approach similar to that of Cropp er & Horne (1994). However, b ecause of the relatively low signal to noise of the data we could not identify a unique solution. In §3 we noted the presence of a broad dip in soft X-rays at 0.4 which could b e attributed to either an accretion stream (since there is no similar feature at higher energies) or the rotation of the accretion region(s) as they come into and out of view. In the former case, the dip could b e due to a second accretion stream obscuring our line of sight to the accretion region located in the lower hemisphere of the white dwarf. To our knowledge this would make XMM J2250+5731 unique amongst p olars in showing two absorption dips. In the latter case, the change in the soft X-ray light curve could b e due to either the rotation of two accretion regions, located in opp osite hemispheres, or the rotation of one relatively large p olar region. (Our inversion maps showed that b oth scenarios could re-produce the soft X-ray light curves). The fact that soft X-rays emitted at the base of the accretion region are optically thick and hence viewing angle dep endant could account for the change in the soft X-ray flux. In contrast, the harder X-rays are optically thin and therefore not viewing angle dep endant. Optical p olarimetry data would b e able to confirm the presence of two accretion p oles. However, since XMM J2250+5731 is rather faint, this may prove challenging. 7.2 The energy balance

which was discovered serendipitously using XMM-Newton, does not show a soft X-ray comp onent. We have searched the literature for further observations of p olars observed using XMM-Newton in a high state: we find an additional 6 p olars. (We are aware of a numb er of observations of p olars in the high state which have b een carried out but have not as of yet b een published). V1309 Ori (Schwarz et al 2005), V1432 Aql (Rana et al 2005) and SDSS J075240.45+362823.2 (Homer et al 2005) all show distinct soft X-ray comp onents while SDSS J072910.68+365838.3 and SDSS J170053.30+400357.6 (Homer et al 2005) do not. In the case of SDSS J015543.4+002807.2 (Schmidt et al 2005) the existence of a soft comp onent is not required at a high significance and hence we define it as not having a soft X-ray comp onent. We therefore find that 10 out of 27 systems observed in a high state do not show a distinct soft X-ray comp onent. Ramsay & Cropp er (2007) suggested that if the temp erature of the re-processed X-rays was low enough, it would not b e observable using the XMM-Newton X-ray detectors. This view is also supp orted by the analysis carried out by Vogel et al (2008) on observations of 2XMMp J131223.4+173659. The reason for this could b e that the accretion flow covers a larger fraction of the photosphere of the white dwarf or that the mass accretion rate is lower than in systems which showed a soft comp onent (since 1/4 kTbb (M /f ), where M is the mass accretion rate and f is the fractional area over which accretion is occurring). There is no obvious reason as to why some p olars would have accretion occurring over a larger area than others: they share no common characteristics such as magnetic field strength or orbital p eriod. Indeed, as noted by Ramsay & Cropp er (2004) two systems (BY Cam and RX J2115­58) have one p ole which shows a soft comp onent and one p ole which does not. Further, three systems which have at least one p ole which does not show a soft comp onent are asynchronous systems. However, V1432 Aql which does show a soft comp onent is also an asynchronous p olar.

8

CONCLUSIONS

Ramsay & Cropp er (2004) presented the results of a snapshot survey of p olars observed in a high accretion state using XMM-Newton. They found that 7 out of 21 systems did not show a distinct soft X-ray comp onent. Vogel et al (2008) also rep ort that 2XMMp J131223.4+173659,
c 0000 RAS, MNRAS 000, 000­000

We have serendipitously discovered a faint p olar, XMM J2250+5731, with an orbital p eriod of 2.9 h, in the 2XMMi catalogue. We have identified the optical counterpart as a r 21 ob ject and it shows a deep eclipse in the optical and X-ray bands lasting 12 mins. At soft X-ray energies there is a distinctive drop in counts starting 0.3 cycles b efore the eclipse. This is due to the accretion stream obscuring the accretion region in the upp er hemisphere of the white dwarf. A second dip is seen in soft X-rays at 0.4 which could either b e due to obscuration of the accretion region by a second stream or due to the rotation of the accretion region(s) rotating into and out of view. Amongst eclipsing p olars, XMM J2250+5731 is unusual in that X-ray emission is visible over the whole of the binary orbital phase, apart from the eclipse. We have analysed the X-ray sp ectrum of XMM J2250+5731 and find no evidence for a distinct soft X-ray comp onent. Of the 27 p olars which have b een observed using XMM-Newton and found to b e in a high accretion state, 10 show no distinct soft X-ray comp onent. This is a surpris-


6
ingly high fraction. This together with the result that only a small fraction of p olars show a soft X-ray excess (Ramsay & Cropp er 2004), changes our whole p erception of p olars b eing strong soft X-ray sources. Further, it suggests that p olars with strong soft X-ray comp onents were preferentially discovered using EXOSAT.

9

ACKNOWLEDGEMENTS

Based on observations obtained with XMM-Newton, an ESA science mission with instruments and contributions directly funded by ESA Memb er States and NASA. We thank Gillian James for providing an initial reduction of the XMM-Newton data and Diana Hannikainen and Hanna Tokola for assisting with the NOT observations. Observations were made using the William Herschel Telescop e, the Isaac Newton Telescop e and the Nordic Optical Telescop e on La Palma. We gratefully acknowledge the supp ort of each of the observatories staff. We also thank Andrew Beardmore and other memb ers of the Swift team for scheduling observations of our target. Some of the data presented here have b een taken using ALFOSC, which is owned by the Instituto de Astrofisica de Andalucia (IAA) and op erated at the Nordic Optical Telescop e under agreement b etween IAA and the NBIfAFG of the Astronomical Observatory of Cop enhagen.

REFERENCES
Arnaud, K. A., 1996, Astronomical Data Analysis Software and Systems V, eds Jacoby, G., Barnes, J., p17, ASP Conf Series, 101 Beuermann, K., Burwitz, V., 1995, In ASP Conf Series, 85, 99 Corradi, R. L. M., et al, 2008, A&A, 480, 409 Cropper, M., 1988, MNRAS, 231, 597 Cropper, M., Horne, K., 1994, MNRAS, 267, 481 Cropper, M., Ramsay, G., Wu, K., 1998, MNRAS, 293, 222 Cropper, M., Wu, K., Ramsay, G., Kocabiyik, A.,1999, MNRAS, 306, 684 Drew, J., et al, 2005, MNRAS, 362, 753 Homer, L., et al, 2005, ApJ, 620, 929 King A. R., Lasota J. P., 1979, MNRAS, 188, 653 Lamb D. Q., Masters A. R., 1979, ApJ, 234, 117 Ramsay, G., Mason, K. O., Cropper, M., Watson, M. G., Clayton, K. L., 1994, MNRAS, 270 Ramsay, G., Cropper, M., 2004, MNRAS, 347, 497 Ramsay, G., Cropper, M., 2007, MNRAS, 379, 1207 Rana, V. R., Singh, K. P., Barrett, P. E., Buckley, D. A. H., 2005, ApJ, 625, 351 Roming, P. W. A., et al, 2005, Space Sci Rev, 120, 95 Schmidt, G. D., et al, 2005, ApJ, 620, 422 Schwarz, R., Reinsch, K., Beuermann, K., Burwitz, V., 2005, A&A, 442, 271 Vogel, J., Byckling, K., Schwope, A., Osborne, J. P., Schwarz, R., Watson, M. G., 2008, A&, 485, 787 Watson, M. G., King, A. R., Jones, M. H., Motch, C., 1989, MNRAS, 237, 299 Watson, M. G., et al, 2001, A&A, 365, L51 Watson, M. G., et al, 2009, A&A, 493, 339 Wilms, J., Allen, A., McCray, R., 2000, ApJ, 542, 914 Xu, Y., Reid, M. J., Zheng, X. W., Menten, K. M., 2006, 311, 54

c 0000 RAS, MNRAS 000, 000­000