next up previous
Next: Fits to the 0.6-10 Up: ASCA observations of the Previous: Extraction of the ASCA

Spectral analysis

In all fits, cosmic abundances for the fitted column density N tex2html_wrap_inline496 and the photoelectric cross sections of Morrison & McCammon (1983) were used. A fit was retained acceptable if the corresponding tex2html_wrap_inline494 probability was equal or larger than 5%. The improvement in the tex2html_wrap_inline494 obtained by adding additional fit parameters was estimated with the F-test, with a threshold for a significant improvement P tex2html_wrap_inline502 =95%.

A single power law model absorbed by the Galactic column density (N tex2html_wrap_inline504 cm tex2html_wrap_inline348 ; Murphy et al. 1997) along the line of sight provides a poor fit to the ASCA data of 3C445, with a reduced tex2html_wrap_inline494 of tex2html_wrap_inline510 for 421 degrees of freedom (dofs), which is rejected at > 99% confidence. This fit yields a negative photon index tex2html_wrap_inline514 . Figure 1 shows the residuals of the single power law model for the SIS and GIS detectors separately. Deviation from the model in the form of excess flux below 1.5 keV and the Fe line at tex2html_wrap_inline356 6 keV are apparent.

The modelling of the Fe line depends critically on the parameterization of the underlying continuum. We thus concentrated first on finding an adequate description of the latter, and restricted our spectral fits to the energy range 0.6-10 keV excluding the 5-7 keV region. The fits to the continuum are discussed in § 3.1, with the Fe line fits following in § 3.2. The results of the spectral fits are summarized in Table 2, where all quoted uncertainties on the parameters are 90% for 1 parameter of interest ( tex2html_wrap_inline518 =2.7). All quantities in Table 2 are in the observer's rest-frame.





Scott Koch
Fri Nov 14 14:50:41 EST 1997