Figure 5.3: INTEGRAL/JEM-X significance mosaic images of the Ophiuchus cluster region in the 3-5, 5-7, 7-10 and 10-18 keV bands. The source in the bottom of the image is the Cataclysmic Variable V2400 Oph.
it can fully extract the flux of the source.
5.3 Morphological analysis
We used the mosaics extracted from both ISGRI and JEM-X data to study the morphol-ogy of the source and its energy dependence. Figure 5.4 shows the full ISGRI and JEM-X images in the 20-40 keV and 3-18 keV bands. For comparison with the PSF of the instru-ments, the insets show the image of the point source V2400 Oph, located 50 arcmin from the cluster, re-normalized to the same brightness. In the JEM-X image, we can see that the source is clearly extended. However, the extension of the source is not obvious in the ISGRI image.
Since we detect only the central regions of the cluster, a conventional fit of the JEM-X data with a spherical beta-model is not possible, but fitting the JEM-X image with a Gaussian model,
I(r) =Aexp
−ln(2)r2 R2
, (5.1)
whereR is the half-width at half-maximum (HWHM) of the source, one can get an indi-cation on the size of the detected region. Besides, the model described in Eq. 5.1 is very similar to the spherical beta-model for r ≤ R, so using this model we can estimate the core radius Rc of the cluster. The best fit to the JEM-X image with the Gaussian model gives R = 3.6±0.1 arcmin. This value is significantly larger than the PSF of JEM-X (1.8 arcmin HWHM), which confirms that the source is extended. TheR parameter cor-responds to the superposition of the intrinsic radial profile of the source with the PSF of the instrument,
R2=R2source+HW HMJEM-X2 . (5.2)
Subtracting the contribution of the JEM-X PSF, we find that the intrinsic radial profile of the source is Rsource = 3.1±0.1 arcmin. This value is consistent with the core radius
Figure 5.4: INTEGRAL images of the Ophiuchus cluster in the 20-40 keV band from ISGRI (left) and in the 3-18 keV band from JEM-X (right). For comparison, the insets show the image of a known point source in the same field, the Cataclysmic Variable V2400 Oph.
obtained from higher-statistics ASCA data (Rc = 3.2 arcmin, Watanabe et al. (2001)).
Figure 5.5 shows the radial profile of the source compared to the PSF of the instrument, as well as the total apparent radial profile from JEM-X data. We can see that the data are very well represented by our model.
Since the source is clearly detected and spatially-resolved in several energy bands, it is pos-sible to investigate the temperature variations within the detected area. Indeed, if we as-sume that the emission below 20 keV is completely dominated by thermal bremsstrahlung and that the abundance is constant throughout the cluster, the 3-7/7-18 keV hardness ra-tio depends only on the temperature of the gas in a specific region. In order to convert the JEM-X 3-7/7-18 keV hardness ratio to the gas temperature, we simulated JEM-X spectra for different input temperatures and an abundance fixed to 0.49 (Mohr et al. 1999), and computed the dependence of the 3-7/7-18 keV JEM-X hardness ratio as a function of the gas temperature. In the center of the cluster, we found a hardness ratioHR= 0.78±0.09, which corresponds to a temperature kT = 9.1+1.40.9 . Figure 5.6 shows the spatial devia-tions of the hardness ratio from the central value, in units of σ. No deviations from the central value are found above 0.7σ, which corresponds to temperature variations of ± 2 keV. This result is consistent with theASCA results (see Fig. 5.1). Indeed, even though much hotter temperatures are found in the outer regions of the cluster, the core shows an almost iso-thermal profile, which is consistent with our results. This analysis shows that a single-temperature bremsstrahlung model should describe well the total JEM-X spectrum.
On the other hand, with a core radius Rc ∼ 3.2 arcmin, the source should be almost point-like for ISGRI. Renaud et al. (2006b) presented the results of ISGRI data simula-tions assuming a uniform disk. For a disk of an angular size of 5 arcmin, they evaluate that the relative error on the reconstructed flux assuming that the source is point-like is
∼5%. Since the cluster has a peaked radial profile instead of uniform disk, the relative
5.3. Morphological analysis 83
Figure 5.5: Deconvolved radial profile of the Ophiuchus cluster extracted from JEM-X data (dashed blue line), compared to the PSF of the instrument (dashed green). The actual data (purple bars) and the apparent profile (red line) are also shown.
Figure 5.6: Deviations of the JEM-X 3-7/7-18 keV hardness ratio from the central value, in units of σ. No significant temperature variations are found in the detected region.
Figure 5.7: Chandra image of the core of the Ophiuchus cluster. The circles represent the ISGRI best fit position in the 20-24 (green), 24-30 (black) and 30-40 keV (white) bands with 90% error radius. The black cross shows the position of the maximum emission in the cluster on the central cD galaxy, while white crosses show the position of the weak point sources in the cluster.
error on the flux extracted by the standard OSA spectral extraction tool is < 5%. For this reason, it is safe to treat the source as point-like for ISGRI.
Since the source is not resolved spatially, in principle we cannot exclude the possibility that a hard point source in the cluster (e.g. a very absorbed AGN) has a significant influence on the hard X-ray spectrum. Nevertheless, to study the energy dependence of the source we can compute the best-fit position of the source in several energy bands and compare the results to the low-energy morphology. For comparison, publicChandra data on the cluster with very high angular resolution (<1′′) are available, which allows us to distinguish the point sources from the diffuse emission. Figure 5.7 shows the 1-10 keV Chandra image of the cluster with 90% error circles for the position of the source in the 20-24, 24-30 and 30-40 keV bands. The black cross shows the position of the central cD galaxy, close to the maximum of the low-energy emission. On the other hand, the white crosses indicate the weak point sources detected byChandra. One can clearly see that the best-fit ISGRI positions are well-centered on the central cD galaxy, and is not consistent with any of the weak point sources. This indicates that the emission detected by ISGRI is truly of diffuse origin, and that significant contamination of the high-energy spectrum by point sources is unlikely.