Документ взят из кэша поисковой машины. Адрес оригинального документа : http://xmm.vilspa.esa.es/docs/documents/CAL-SRN-0254-1-3.ps.gz
Дата изменения: Fri Jul 17 17:19:09 2009
Дата индексирования: Mon Oct 1 22:56:29 2012
Кодировка:
Поисковые слова: п п п п п п п п п п п п п п п п п п п п п п п п п п п п п п п п п п п п п п

XMMNewton CCF Release Note
XMMCCFREL254
EPIC MOS screening thresholds
C. M. H. Chen, R. D. Saxton
15 July 2009
1 CCF components
Name of CCF VALDATE EVALDATE Blocks changed XSCS flag
EMOS1 SCREENTHRESH 0001.CCF 20000101 SCREEN THRESH NO
EMOS2 SCREENTHRESH 0001.CCF 20000101 SCREEN THRESH NO
2 Changes
These files give the screening thresholds which should be used when determining the
lowenergy noise properties fo the MOS CCDs.
For each CCD, the file gives the softband minimum and maximum, the hardband min
imum and maximum and the rateband minimum and maximum energies in keV. In
addition it gives the threshold for classifying each MOS CCD as noisy. These thresholds
are expressed as selectlibstyle string expressions containing two variables---the hardness
ratio H, i.e., the ratio of hardband to softband event rates, and the event rate R in the
rate band. Both H and R are to be measured within the CORNER region as specified in
a block of the same name in the CCFs EMOSn LINCOORD nnnn.CCF only.
According to a presentation by Kip Kuntz during the XMM Background Working Group
meeting in 3/2009, the characteristics of the lowenergy noise is changing with time. It is
thus expected that the thresholds will have to be revised from time to time.
3 Scientific Impact of this Update
This CCF is read by the new task emtaglenoise beginning in SAS 9 as part of the procedure
to identify CCDs su#ering from lowenergy noise in an observation, based on an algorithm
described in Kuntz and Snowden (2008) A&A 478, 575--596. As such, it provides a new
1

functionality in SAS. Placing all numeric quantities and event selection criteria in a CCF
allows these thresholds to be modified in the future as the noise characteristics change.
4 Estimated Scientific Quality
Kuntz and Snowden (2008) have showed how, using these thresholds, one can e#ectively
identify if CCDs 4 and 5 of MOS 1 and CCDs 2 and 5 of MOS 2 su#er from lowenergy noise
in all observations publicly available as of 1st April, 2006 (which numbered over 2000).
The current specification of the screening thresholds is taken verbatim from Table 1 of
Kuntz and Snowden (2008).
Figure 9 of the same paper displays graphically how they were chosen. For MOS 1 CCDs 4
and 5 and MOS 2 CCD 2, the thresholds unambiguously delineate noisy observations from
normal ones on the (H,R)plane. For MOS 2 CCD 5, the normal and noisy observations
blend together, and the noise strength varies smoothly between theh normal and noisy
states. Kuntz and Snowden's chosen division point is `where the amplitude of the low
energy excess is comparable to the uncertainty in the mean spectra.' They have detected
no noisy observation in the remaining CCDs.
5 Test procedures and results
The SAS task emtaglenoise reads this new CCF via CAL calls:
calServer.state()>ccd()>set(_ccdnr);
ScreenThresholdServer *st = 0; st = calServer.getAtom(st);
_minEnergy[i] = (int)( 1000 * st>ratebandmin() );
_maxEnergy[i] = (int)( 1000 * st>ratebandmax() );
_minEnergy[i] = (int)( 1000 * st>softbandmin() );
_maxEnergy[i] = (int)( 1000 * st>softbandmax() );
_minEnergy[i] = (int)( 1000 * st>hardbandmin() );
_maxEnergy[i] = (int)( 1000 * st>hardbandmax() );
const string & thresh = st>threshold();
Using the test harness of emtaglenoise, we have shown that the SAS task successfully
reads in and makes use of data in this CCF and subsequently produces the expected
lowenergy noise classification on Fedora and Solaris, in highmemory mode.
2