Study of the X-Ray Background Spectrum and Its Large-Scale Fluctuation with ASCA

We studied the energy spectrum and the large-scale fluctuation of the X-ray background with the ASCA GIS instrument based on the ASCA Medium Sensitivity Survey and Large Sky Survey observations. A total of 91 fields with Galactic latitude |b| > 10^∘ were selected with a sky coverage of 50 deg² and 4.2 Ms of exposure. For each field, non-X-ray events were carefully subtracted and sources brighter than ~ 2× 10⁻¹³ \,erg \,cm⁻² \,s⁻¹ (2–10 keV) were eliminated. Spectral fits with a single power-law model for the individual 0.7–10 keV spectra showed a significant excess below ~ 2 \,keV, which could be expressed by an additional thermal model with kT 0.4 \,keV or a steep power-law model with a photon index of Γˢoft 6. The 0.5–2 keV intensities of the soft thermal component varied significantly from field to field by 1\,σ= 52⁺⁴₋₅%, and showed a maximum toward the Galactic Center. This component is considered to be entirely Galactic. As for the hard power-law component, an average photon index of 91 fields was obtained to be Γʰard = 1.412± 0.007± 0.025 and the average 2–10 keV intensity was calculated as FXʰard = (6.38± 0.04± 0.64)× 10⁻⁸ \,erg \,cm⁻² \,s⁻¹ \,sr⁻¹ (1\,σ statistical and systematic errors). The Galactic component is marginally detected in the hard band. The 2–10 keV intensities show a 1\,σ deviation of 6.49+0.56-0.61%, while deviation due to the reproducibility of the particle background is 3.2%. The observed deviation can be explained by the Poisson noise of the source count in the f.o.v. (~ 0.5 \,deg²), even assuming a single logN -- logS relation on the whole sky. Based on the observed fluctuation and the absolute intensity, an acceptable region of the logN -- logS relation was derived, showing a consistent feature with the recent Chandra and XMM-Newton results. The fluctuation of the spectral index was also examined; it implied a large amount of hard sources and a substantial variation in the intrinsic source spectra (ΓS 1.1± 1.0).