Rays from Polars

The majority of the presently known polars was discovered in the ROSAT All-Sky-Survey by their soft X-ray and EUV emission (see Table 12.1). The dominance of soft X-ray emission is a property of many, but by no means of all polars [2,69,70]. The Chandra LETG spectrum of AM Her in Fig. 12.6 [8, and private communication by V. Burwitz] shows the presence of a strong soft X-ray component in its high state. The soft X-rays disappear in the low state, probably because the temperature drops and the emission retreats from the accessible window [2, see also Fig. 12.7, below]. In the high state, the hard thermal component consists of bremsstrahlung with emission lines of the abundant elements superimposed (e.g., OVII at 22 A). Most of the bolometric flux, however, is in an optically thick component that can be approximated by a blackbody component with a dominant temperature near krbb ~ 29 eV and no obvious emission or absorption lines. The blackbody approximation is reasonable as shown also by the EUVE spectra of several polars [53]. Although most polars were discovered as soft ROSAT sources, the sample of hard RASS sources contains additional polars [77], and data from XMM-Newton suggested that a larger fraction of polars than considered earlier is not dominated by soft X-rays [69,70]. In addition, very low accretion rate polars were discovered in the Hamburg quasar and the Sloan surveys by their cyclotron emission. They were not seen as soft X-ray sources in the RASS and are convincingly interpreted as magnetic pre-CVs, which

Fig. 12.5 Detail of the Chandra HETG spectrum of EX Hya in the vicinity of the FeXXII 3 ^ 2 lines, binned to 0.005 A (from [57])

wavelength (A)

10-1

10-3

Chandra LETG

OVII

ftli t

PQ Ge rfrf m

Energy (keV)

Fig. 12.6 Spectral energy distributions of AM Her and PG Gem based on Chandra LETG spectra binned into 100 bins per decade in energy. The model for the hard X-ray bremsstrahlung component is included (dashes lines) [8, and private communication V. Burwitz]

have synchronized already but transfer matter at a low rate from the wind of the secondary via the magnetic field to the white dwarf and have not yet started Roche lobe overflow [75].

The early model of a white dwarf atmosphere heated by bremsstrahlung from the postshock plasma suggested a ratio of blackbody vs. bremsstrahlung luminosities of about unity [36,45], substantially less than observed in AM Her. Much of the emission from the heated atmosphere was, furthermore, found to appear in the FUV rather than in soft X-rays [17, 39]. Remedy comes, as noted in Sect. 12.3, from the concept of buried shocks [2,15,44]. Shock burying (Fig. 12.1) occurs only for high mass flow densities m and the emerging overall spectral energy distribution depends, therefore, strongly on the m-spectrum incident on the white dwarf. Figure 12.7 shows representative spectra calculated for the parameters of AM Her, B = 14 MG, an angle of 0 = 60° against the field direction, and m from 10~3 to 100gcm~2s_1, each for an emitting area of 1015 cm2 at 100pc distance [2]. They are based on radiation-hydrodynamic calculations [13] and approximately account for absorption due to burying and for reprocessing. The reprocessed radiation from irradiation by low-m tall shocks appears in the UV rather than in the soft X-ray regime. The blackbody assumption used here is a crude approximation and a more correct treatment of irradiated atmospheres is presented in [39]. We find that the observed overall spectral flux distribution of AM Her can be synthesized from model spectra as shown in Fig. 12.7, with the lowest m being responsible for the

Energy (keV)

Fig. 12.7 Representative spectra for B = 14 MG, an angle & = 60° against the field direction, and mass flow densities m = 10~3,10~2,10"1,1,10, and 100g cm 2s 1 (from bottom to top), each calculated for an emitting area of 1015 cm2 at 100 pc distance. Dashed curves are the spectra emitted by the postshock plasma, solid curves the emergent spectra that include absorption and the reprocessed component from the white dwarf photosphere (from [2])

Energy (keV)

Fig. 12.7 Representative spectra for B = 14 MG, an angle & = 60° against the field direction, and mass flow densities m = 10~3,10~2,10"1,1,10, and 100g cm 2s 1 (from bottom to top), each calculated for an emitting area of 1015 cm2 at 100 pc distance. Dashed curves are the spectra emitted by the postshock plasma, solid curves the emergent spectra that include absorption and the reprocessed component from the white dwarf photosphere (from [2])

observed cyclotron emission, m values around 1 g cm~2s_1 producing the observed bremsstrahlung, and the highest m being responsible for the soft X-rays via thermal-ization, though with substantial overlap. In conclusion, a large soft excess indicates high m values and buried shocks, a weaker soft X-ray flux generally lower m. The m-spectrum of AM Her in its high state rises toward high m values, while it is rather flat in the low state [2]. These differences arise from either a gating mechanism in the magnetosphere or are imprinted on the stream already in the L1 point. Near the white dwarf, a high field strength acts as a bottle neck and increases m, which explains the generally softer X-ray emission of high-field polars. On the other end of the m-scale, the magnetic pre-CVs seem to be characterized not only by a low total accretion rate M, but also by a low mass flow density m and, hence, a comparatively large area of the accretion spot as defined by the field lines that connect to the secondary star.

Further information on the structure of the X-ray emission region in polars can be gathered by timing studies (light curves) and by X-ray spectroscopic studies. Light curves in the kiloelectronvolt regime are consistent with the notion that the observed bremsstrahlung originates from tall shocks, i.e., those subcolumns which are not buried. The soft X-ray light curves at photon energies E < 0.5 keV provide information on the structure of the hot photosphere. In the simplest notion, the hot spots in the white dwarf photosphere are modeled as blackbody-emitting surface elements. That this concept is too simple was shown already by the EXOSAT soft X-ray light curve of AM Her in its "reversed" mode, in which it emitted soft X-rays from a second, normally inactive pole. The observed light curve of the self-eclipsing source was found to be box-like with superimposed fluctuations, suggesting that the source was about as tall as wide and raised above the photosphere by some 107 cm [29]. Light curves of bright polars obtained with EUVE and especially with the ROSAT PSPC provided better statistics and suggested mounds rising to a height of a few percent of the white dwarf radius [76,81]. Such mounds can be explained by local heating of the photosphere surrounding individual matter-loaded flux tubes [47] and by splashes by which the photosphere reacts to the termination of a high-m filament or "blob"-like a compressed spring.

X-ray spectroscopy has reached a level that allows detailed diagnostics of the optically thin hard X-ray emitting post-shock plasma. What was said above for IPs applies for polars, too. It is already possible to perform radial velocity studies of the emitting ions in the postshock region [8] and one may envisage to sample the velocity structure of the flow as a function of temperature using the variation in excitation as the flow settles.

0 0

Post a comment