The ASCA satellite [13] comprises two pairs of detectors, the Solid-state Imaging Spectrometer (SIS: [14,15]) and the Gas Imaging Spectrometer (GIS: [16]). The large field-of-view of the GIS makes it the better instrument for looking at the extended structure of Cygnus-A while the GIS' superior efficiency at higher energies makes it most sensitive to the harder spectral component. On the other hand, the higher spectral and spatial resolution of the SIS and its better efficiency at lower energies makes it the instrument of choice for examining the central cooling flow region of the cluster.
What follows is a description of the results pertaining to the AGN obtained from two pointings at Cygnus-A made in 1993. The full account of the data reduction and analysis will be published elsewhere (Arnaud et al., in preparation).
Spectra of the core of the cluster were extracted by using a circular region of radius 3' for the SIS and 5' for the GIS. The GIS has a significantly worse spatial resolution than the SIS so a larger region was used for the former to ensure as much as possible of the core emission was captured. Models were fitted to the spectra above 4 keV in order to minimize the effects of emission from the cooling flow. There was no adequate fit of a single component thermal model to the data, the high temperature required was incompatible with the observed relative strengths of H-like and He-like Fe-Kalpha. Adding a power-law with its own absorbing column gave an acceptable fit. The best fit model is shown in figure 1 superimposed on the data from the second observation.
Figure 1: GIS and SIS spectra with best fit thermal plus absorbed
power-law
Figure 2: Confidence contours for power-law index and absorbing
column from fit to GIS and SIS spectra
The confidence contours for power-law index and absorbing column are shown in figure 2. This figure also shows the Ginga result, the much smaller confidence region is due to the much larger collecting area and higher energy spectral cut-off of Ginga. There is no evidence for changes in either column or power-law index between the Ginga observation and the two ASCA observations.
This result narrows down the location of the hard X-ray component to within 5 arcminutes of the Cygnus-A radio core. This is a slight improvement over Ginga, which localised this component to +/-6 arcmin of the core. To improve on this an image was constructed using the SIS data by extracting only those X-rays with energies above 6.5 keV, the energy at which the power-law component starts to dominate the thermal component. Figure 3 shows the surface brightness profile, generated by summing counts in concentric annuli, from this image. For comparison, a surface brightness profile of an image generated in the 1.5-4.5 keV band is also shown along with the telescope point spread function. These last two profiles were normalized to match the 6.5-9.5 keV profile in the central bin. Figure 3 demonstrates that the higher energy X-ray emission from Cygnus-A is indeed from a point source and that point source is coincident with the radio core of Cygnus-A. The main uncertainty in the X-ray position of the point source is from systematics in the satellite attitude solution, which amount to +/- 20 arcsec.
Figure 3: Surface-brightness profiles from SIS0. The crosses with
circles are for 1.5-4.5 keV, the simple crosses are for
6.5-9.5 keV, and the stepped line is the telescope point
response
There is no evidence for a redshifted neutral iron line and the 90% confidence upper limit on the equivalent width relative to the power-law is 100 eV. Fluourescent iron emission is expected because although both line and continuum will be absorbed by the torus the line photons will be absorbed less since they are produced throughout the torus [17,18]. The line equivalent width increases with column through the torus but also depends on the detailed geometry. The upper limit on the equivalent width of 100 eV does allow us to conclude that the absorbing column is less than 10^24 cm^-2, in agreement with the observed continuum shape, however more detailed analysis will have to await further radiative transfer calculations (Pier, in preparation).