2. Observations and data analysis
Results from the the Medium-Energy Concentrator Spectrometer (MECS; 1.8-10 keV; Boella et al. 1997) and the Phoswich Detection System (PDS; 15-300 keV; Frontera et al. 1997) on-board BeppoSAX are presented.
We have used all the 4 observation of IC443 available in the public BeppoSAX archive; Table 1 lists their coordinates and exposure times. Good data were selected from intervals when the elevation angle above the Earth's limb was and when the instrument configurations were nominal, using the SAXDAS 2.0.0 data analysis package.
Table 1. Public BeppoSAX Observations of IC443
Fig. 1 shows the MECS mosaics of IC443 in different bands. The images are vignetting and exposure corrected. The 2.0-4.0 MECS images show an emission morphology similar to the GIS broad-band maps presented by K97. On the other hand, the hard MECS image above 4 keV is very different. There is weak diffuse extended emission in the north of the remnant, and two strong compact sources, Src A and B. Table 2 reviews the properties of the two sources. In Fig. 1, we also marked the weak extended hard X-ray emission with C. Both the Src A and B were observed by K97; Src A corresponds to the hard source named "HXF", while Src B corresponds to the hard extended emission called "The Ridge" (also reobserved by Olbert et al. 2000). In the hard X-ray image in Fig. 1 we have also reported the contour of the molecular hydrogen line emissivity observed by Burton et al. (1988), which shows that the hard X-ray sources are closely correlated with it.
Table 2. Compact sources detected in the hard MECS images.
In Fig. 2 we report the 4.0-10.5 keV profile of Src A and Src B compared with the MECS PSF profile in the same energy range. The PSF profile was derived from a BeppoSAX observation of the very bright point source Cyg X-1. Src A seems to be extended, while Src B is consistent with the expected PSF. The latter was reported as an elongated ridge by K97, but we note that the ASCA GIS pointing direction is such that the location of Src B is seen at large off-axis angle, which can account for most of source extension observed by ASCA. Instead, the BeppoSAX Src A and B are placed only at and off-axis, where the instrumental distorsion is at its minimum.
We have also investigated if Src A and Src B have soft X-ray counterpart using archive ROSAT observations. Fig. 4 is a PSPC 0.2-2.0 keV image of the region containing MECS Src A and B, with the MECS 4.0-10.5 keV contours of Fig. 1 (right panel) overimposed. Src A has an HRI counterpart quoted by K97 and the PSPC WGACAT source 1WGA J0617.1+2221 with a rate of cnt s-1 is located inside the BeppoSAX error circle. Src B has a weaker PSPC counterpart indicated by the arrow in Fig. 4, located north of the MECS source centroid and not reported in other PSPC catalogs. We have verified that the extended hard X-ray emission we have labelled with C in Fig. 1 has a thermal origin and it represents the weak hard tail of the bright soft X-ray nebula visible in the PSPC image.
We have performed spectral fits of the MECS spectra of Src A and B using the MEKAL optically thin thermal plasma model modified by the interstellar absorption assuming the standard abundances. Spectra were extracted from circular regions with radii of and for Src A and B, respectively. We have used the IC443H observation for Src A and IC443E observation for Src B. MECS spectra have been rebinned to have at least 20 counts per channel. The MECS background was collected in an annulus between and for Src A, while we have used the standard background for Src B which is immersed in the thermal emission and a proper background region could not be found around it.
A thermal origin for the spectrum of Src A is strongly rejected (), while a power law model nicely fit the data (Table 3). The spectrum of Src B cannot be reproduced by a power law or by a thermal emission model alone ( 61/33 and 78/33 respectively) but requires a combination of the two, reported in Table 3.
Table 3. Summary of MECS spectral fitting results.
We have also analyzed the PDS spectra. Background was subtracted using the OFF collimator positions; we have verified that no other contaminating sources are present in the background positions. Spectra were collected using the Variable Rise Time Threshold mode and have been rebinned following a logarithmic scheme suggested by the PDS hardware group to not undersample the instrument spectral resolution. Table 4 reports the observed countrates and flux in different energy ranges. A joint fit of Src A MECS spectrum and IC443H PDS spectrum (shown in Fig. 3) yields best-fit results similar to the ones reported in Table 3 and a PDS/MECS normalization ratio of . This value is consistent with the expected value of 1, also considering that the expected constribution of Src B in the PDS spectrum of IC443 could be up to . As for Src B, the joint fit yields a PDS/MECS ratio of , we argued that the PDS spectra of IC443E is heavily contaminated by the brighter Src A. In fact, using the best-fit model of Src A and considering the triangular spatial response of the PDS collimator, we predict a Src A contribution of cnt s-1 in the PDS spectra of Src B between 15 and 30 keV, consistent with the observed rate (Table 4). Therefore, a proper analysis of the PDS spectra of Src B cannot be done.
Table 4. Observed background subtracted count-rates and unabsorbed fluxes in different energy ranges, MECS for keV and PDS for keV.
© European Southern Observatory (ESO) 2000
Online publication: October 24, 2000