Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://xmm.vilspa.esa.es/docs/documents/CAL-TN-0063.ps.gz Äàòà èçìåíåíèÿ: Wed Feb 23 01:03:04 2011 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 20:54:52 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: http www.badastronomy.com bad tv foxapollo.html

XMM­SOC­CAL­TN­0063
RGS blazar spectra and prospects for an E#ective Area correction
A.M.T. Pollock
10 October 2004
The class of blazar AGN exemplified by Mkn421 is widely thought to have simple featureless X­ray
continua so that when interstellar absorption is taken into account the RGS flux is simply
f # # # #-2 exp (-# # NH )
where # is the PhotonIndex beloved of XSPEC; # # is the photoelectric cross­section; and NH is the
absorbing column density. The best available estimate of f # is the so­called ``fluxed spectrum'' f RGS
#
produced by the SAS task rgsfluxer. If everything is working as expected in the RGS and in the
universe, the simple transform
g # (#, NH ) = # 2-# exp (+# # NH )f RGS
#
.
= constant = K
where .
= is taken here to mean ``should equal''. I've calculated this transform for the 8 suitable blazars
listed in the table for which RGS data are available hoping, but not expecting, to find a flat line. The
rgsfluxer spectra used combined RGS1 and RGS2 and order 1 and order 2 in all observations to
produce a single spectrum for each blazar. With a suitable choice of #, NH , K for each blazar all the
curves g # (#, NH )/K should line up. I've made choices of # and K mainly by trial­and­error and NH
guided by NED's value of EB-V and popular values used in the literature. Before transformation,
the rgsfluxer spectra look quite di#erent; afterwards they look pretty much the same as shown,
for example, in Fig. 1. This shape of the transformed spectrum ­ reasonably flat above about the
oxygen edge and decreasing at long wavelengths ­ is roughly the same as type of disagreement often
found between RGS and EPIC in the cross­calibration work underway at Villafranca. It's also a shape
seen in nearly all such comparisons between di#erent blazars. The series of figures following Fig. 2
shows the comparisons between the same transformed Mkn421 spectrum and the other 7 objects. The
Mkn421 red line is the same in each figure. A similar comparison between spectra of Mkn421 from
individual observations also works quite well except perhaps for the faintest spectra.
The simplest interpretation of the (remarkable) agreement between the transformed
spectra of several di#erent objects at di#erent redshifts is that it shows what the RGS
is doing and that this could and should provide the basis for robust corrections to RGS
data.
1

XMM­SOC­CAL­TN­0063 2
z EB-V # NH K
10 20 cm -2
Mkn421 0.03002 0.015 2.10 1.34 1.
PKS2155­304 0.116 0.022 2.55 0.93 4.
Mkn501 0.01554 0.019 2.12 1.20 7.
1H1219+301 0.13 0.024 2.30 2.00 7.
PKS0548­322 0.069 0.035 1.90 2.36 20.
H1426+428 0.129 0.012 1.84 1.36 20.
Mkn180 0.04528 0.013 2.10 0.90 20.
1H0414+009 0.287 0.119 2.10 5.00 40.
Table 1: Eight X­BL blazars observed by RGS that are supposed to have simple power­law spectra
subject to interstellar absorption.

XMM­SOC­CAL­TN­0063 3
Figure 1: rgsfluxer spectra of PKS2155­304 and H1426+428 before and after transformation..

XMM­SOC­CAL­TN­0063 4
Figure 2: Transformed rgsfluxer spectra of Mkn421 (red) and PKS2155­304 and Mkn421 (red) and
1H1219+301..

XMM­SOC­CAL­TN­0063 5
Figure 3: Transformed rgsfluxer spectra of Mkn421 (red) and Mkn501 and Mkn421 (red) and
H1426+428.

XMM­SOC­CAL­TN­0063 6
Figure 4: Transformed rgsfluxer spectra of Mkn421 (red) and PKS0548­322 and Mkn421 (red) and
Mkn180.

XMM­SOC­CAL­TN­0063 7
Figure 5: Transformed rgsfluxer spectra of Mkn421 (red) and 1H0414+009.