Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://xmm.vilspa.esa.es/docs/documents/CAL-SRN-0238-1-0.ps.gz Äàòà èçìåíåíèÿ: Wed Sep 19 19:21:12 2007 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 22:50:04 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: crab nebula

XMM­Newton CCF Release Note
XMM­CCF­REL­238
An improved model of the RGS e#ective area based on the
build­up of carbon contamination
A.M.T. Pollock
August 23, 2007
1 CCF components
Name of CCF VALDATE List of Blocks changed XSCS flag
RGS1 EFFAREACORR 0005 2001­01­01T00:00:00 EFFAREACORR NO
RGS2 EFFAREACORR 0005 2001­01­01T00:00:00 EFFAREACORR NO
2 Changes
Last year saw the introduction of empirical RGS e#ective area corrections. The original corrections
were based on a set of observations of Mkn421 whose spectrum was assumed to be a simple power­
law subject to constant galactic absorption. Observations at di#erent times, though of variable
intensity and slope, were used to make a polynomial model of the wavelength­dependent changes
in e#ective area, particularly the decrease at long wavelengths that also became clear at the same
time. A final comparison with the Crab nebula allowed final adjustments for power­law slope and
overall normalisation.
Since then, by comparing the changes observed in the RGS spectra of objects of constant spectrum,
it has been shown that the time­variable component of the e#ective area can be well modelled by a
linearly increasing layer of carbon contamination as shown in Fig. 1. This has been combined with
a constant polynomial correction and the more detailed model of the Crab spectrum described by
Kaastra et al. in a submitted paper that includes additional corrections for dust scattering, pile­up
and spatial variations of the spectral index.
1

XMM­Newton CCF Release XMM­CCF­REL­238 Page: 2
Figure 1: The thickness of the layer of carbon contamination implied by di#erences in the RGS
fluxed spectra of the neutron star RXJ1856­3754 and the Vela Pulsar Wind Nebula, both of which
are e#ectively constant. The linear increase now accounts for the time­variable part of the model of
the RGS e#ective area.

XMM­Newton CCF Release XMM­CCF­REL­238 Page: 3
3 Scientific Impact of this Update
This release should contribute to the continuing e#ort to reconcile the RGS and EPIC instruments.
4 Estimated Scientific Quality
This CCF allows broadband fluxes to be measured at the few percent level but, perhaps more
importantly, provides a robust extrapolation into the future of the evolution of the RGS e#ective
area.
5 Test procedures & results
During their construction, many new RGS RMFs have been calculated with these new CCFs in­
cluding their application to the 11 observations of RXJ1856­3754 throughout the mission shown in
Fig. 2. The normalisation of the first order spectra of RGS1 and RGS2 is shown for the following
XSPEC model fit jointly to all the first order RGS data only, with the following best­fit parameters
with their formal errors.
TBabs nH = 6.1 â 10 18 cm -2 fixed
bbody kT = 62.09 ± 0.07 eV
bbody norm = 3.162 ± 0.029 â 10 -4
C­statistic = 57633.7 for 54522 PHA bins
The contamination history shown in Fig 1 is only sensitive to di#erences in the detected spectrum
of RXJ1856­3754 . The good agreement between the RGS black­body temperature and the cor­
responding independent values of EPIC and Chandra demonstrates the reliability of the overall
e#ective area model for which the polynomial correction is particularly important in a spectrum as
soft as this.
6 Expected Updates
Routine update are expected on an annual basis as the predictions of the linearly­increasing con­
tamination can be tested.

XMM­Newton CCF Release XMM­CCF­REL­238 Page: 4
Figure 2: The relative normalisation of RXJ1856­3754 RGS1 (red) and RGS2 (blue) 1st order spectra
throughout the mission. All the available data are shown, including the final two observations in
rev 1330, which has data problems, and rev 1335, which was an o#set pointing.