Cas A point source spectrum

[Patrick Slane <slane@head.cfa.harvard.edu>]

Hi folks:

I've looked at the spectrum of the point source in Cas A using the full VLP 
data set. As expected, the spectrum shows the effects of pileup. There are no 
obvious features corresponding to atmospheric lines, though one could probably 
hide some quite easily in the regions where the fit residuals are large. Below 
Una has placed a postscript file on the VLP web site that contains a plot of the 
spectrum. The solid histogram is the best-fit blackbody model with a pileup 
correction taken into account. The parameters are: 
Temperature: $T_{bb}^\infty = (4.89 \pm 0.07) \times 10^6$~K 
Emitting area radius: $R_{bb}^\infty = 0.83 \pm 0.03$~km 
Column density: $N_H = (1.25 \pm 0.03) \times 10^{22}{\rm\ cm}^{-2}$ 
$\chi^2 = 493.7$ for 315 degrees of freedom)

The dashed histogram is the unpiled model spectrum.

The residuals are worst around the Si absorption edge. This is an artifact of 
pileup (because the effective area here is very low, and pileup puts some 
photons in these bins which get divided by a very small numbers; if not 
accounted for perfectly, the residuals are large...).

Models without the pileup correction don't fit at all. A power law doesn't 
give a good fit in either case. Models for a light atmosphere at B = 0, 10^12, 
and 10^13 G (nsa model in xspec - Pavlov et al.) produce worse fits than a 
blackbody, and a heavy element atmosphere (xspec model nsspec - Gaensicke et 
al.) is ruled out by the lack of features (unless something is hidden in the 
the area of the residuals). 
 
-- Pat