Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.eso.org/~ppadovan/survey/dxrbs_paper/node8.html
Дата изменения: Wed Dec 3 16:31:47 2003
Дата индексирования: Sat Dec 22 22:36:13 2007
Кодировка:

Identifications and Efficiency

In Figure 2, we show the spectra of the optical counterparts to 85 of 97 observed X-ray/radio sources. All spectra have been smoothed with Gaussians of width 3 pixels. Twelve spectra are not shown, either because (1) too few photons were observed to allow a reliable classification to be made (6 sources), (2) the optical survey plate did not contain enough information to distinguish whether the counterpart was a bright star near the radio position or a fainter extragalactic object at the radio position (5 objects), or (3) because a lower-quality radio position was used (1 object). In this last case (WGAJ1022.1+4126), the object is also very close to a third-magnitude star, SAO 43310, making spectroscopic observations difficult.

Positional information for all 85 sources for which we announce identifications herein are given in Table 3, including information from WGACAT, the PMN and Green Bank surveys, the NVSS and our ATCA survey. A number of sources were serendipitously observed by ROSAT on more than one occasion; for completeness, we give WGACAT positions for all observations of DXRBS sources.

Of the 85 newly identified sources, 59 are FSRQs, and 22 are BL Lacs. Hence, our technique is $\sim 95\%$ efficient at selecting blazars, where we define the efficiency as the fraction of objects which turn out to be blazars in a given survey after selection criteria have been applied. Three of 85 objects are radio galaxies, with CaII breaks stronger than typical BL Lacs (but see § 7.1); while one quasar, which was observed before we had information on its spectral index, turned out to have $\alpha_{\rm r} \gt 0.7$. Classifications, redshifts and observational details for these sources are given in Table 4. The 0.1-2.0 keV X-ray fluxes given in Table 4 are not corrected for Galactic absorption; however, the 1 keV X-ray fluxes given therein are unabsorbed. Note that both the 0.1-2.0 keV and 1 keV X-ray fluxes have been derived from ROSAT count rates using the observed hardness ratio and assuming Galactic $N_{\rm H}$. These numbers may change somewhat when a more thorough analysis of the X-ray spectrum is done (this is in progress). For objects observed more than once by ROSAT, we give in Table 4 the count rates and X-ray fluxes found for each observation. For objects which we classify as either radio galaxies or BL Lacs, we give the equivalent width of the strongest emission line and Ca break strength in Table 5; these values have been also been displayed in Figure 1.

Previous X-ray and radio surveys can claim efficiencies which are nearly comparable to ours. For example, the efficiency of the Slew Survey at identifying HBLs within the well-known $(\alpha_{\rm ox},\alpha_{\rm ro})$ box (Perlman et al. 1996a) is 80-90%, and the fraction of BL Lacs and FSRQs within the flat-spectrum subset of the 1 Jy survey is nearly 90% (260/298) for a dividing line at $\alpha_{\rm r} = 0.5$, but goes below 80% (284/364) for a dividing line at $\alpha_{\rm r} = 0.7$ (Stickel et al. 1994; see also § 7). However, each of these survey techniques were insensitive to large portions of the blazar population (§ 3). The combination of high sensitivity in both the X-ray and radio bands plus a two-band survey method gives DXRBS significant advantages over these previous survey methods.