Документ взят из кэша поисковой машины. Адрес оригинального документа : http://xmm.vilspa.esa.es/docs/documents/CAL-TN-0084.pdf Дата изменения: Tue Aug 14 12:39:08 2012 Дата индексирования: Mon Oct 1 20:49:12 2012 Кодировка: Поисковые слова: universe

XMM-Newton Calibration Technical Note
XMM-SOC-CAL-TN-0084 Evaluation of pile-up thresholds in the EPIC-MOS camera
Matteo Guainazzi (ESA-ESAC, Vil lafranca del Castil lo, Spain), Martin Stuhlinger (ESA-ESAC, Vil lafranca del Castil lo, Spain) August 14, 2012

History
Version 1.0 Date August 14, 2012 Editor M.Guainazzi Note First public version

1


XMM-Newton Technical Note

XMM-SOC-CAL-TN-0084

Page:

2

1

Scop e

This rep ort aims at giving an assessment of the count rate thresholds, ab ove which pile-up can affect the quality of scientific observations with the EPIC-MOS cameras (Turner et al. 2001, A&A, 365, L27). Pile-up is defined as the reconstruction of indep endent events as a single one during a read-out cycle. The registered events are erroneously interpreted as one single event, whose energy is the sum of the individual incoming event energies. Pile-up distorts the sp ectral shap e is three main ways, not necessarily exclusive:

1. by "hardening" the observed sp ectrum with resp ect to the intrinsic source (b ecause events are shifter to higher energies) 2. by suppressing flux, if the sum of the incoming event energies is higher than the on b oard energy rejection threshold, the corresp onding events are lost 3. by "joining" separate mono-pixel events into a single multi-pixel event ("pattern migration")

Pile-up can b e mitigated in affected observation by extracting sp ectra of single events only (see, e.g., Molendi & Sembay, XMM-SOC-CAL-TN-0036, available at: http://xmm2.esac.esa.int/docs/documents/CAL-TN-0036-1-0.ps.gz). A software library to correct sp ectra affected by pile-up is under implementation in the SAS. The reference pile-up thresholds are rep orted in Tab. 3 (Sect. 3.3.2) of the XMM-Newton Users Handb ook (Piconcelli et al. 2012; available at: http://xmm.esac.esa.int/external/xmm user support/documentation/uhb/index.html). For what said ab ove, it is of paramount imp ortance that XMM-Newton prop osers take them into account when they choose the optimal instrumental configuration for their observation.

2
2.1

Metho dology
The samples

The results of this study make use of the Cross-calibration database (XCAL hereafter), a compilation of over 250 observation p erformed over the mission and routinely used for calibration and cross-calibration purp oses (Stuhlinger et al. 2010, XMM-SOC-CAL-TN-0052, available at: http://xmm2.esac.esa.int/docs/documents/CAL-TN-0052.ps.gz). They are mainly bright sources with astrophysically well defined (alb eit not necessarily fully understood) astrophysical models. They have b een mainly observed in "windowed" (i.e. Large or Small Window) EPIC instrumental modes to mitigate pile-up. Nonetheless many of them are still affected by pile-up. This makes of the XCAL an ideal sample for our study.


XMM-Newton Technical Note

XMM-SOC-CAL-TN-0084

Page:

3

Table 1: Best-fit parameters of the Pf versus R function for different modes Mode MOS1-FF MOS1-SW MOS1-LW MOS2-FF MOS2-SW MOS2-LW a 1.52 ± 2.74 ± 1.99 ± 1.48 ± 3.01 ± 1.83 ± b ± ± ± ± ± ± c 0.0 ± 0.2 -0.34 ± 0.13 0.0 ± 0.2 -0.05 ± 0.16 -0.50 ± 0.13 0.0 ± 0.4

- - - - - -

0.08 0.09 0.15 0.07 0.09 0.15

0.6 1.55 0.8 0.49 2.00 0.4

0.2 0.13 0.2 0.16 0.13 0.4

2.2

Pile-up estimation

The largest fraction of events affected by pile-up in a given XCAL sp ectral extraction region is estimated as follows: the PSF-weighted mean of the image obtained by dividing an image of diagonal events (PATTERN = 25­29; selectlib expression: (FLAG & 0x1)!=0 && (FLAG & 0x766ba000)==0) by an image extracted from bona fide X-ray events (PATTERNs from 0 to 12, b oth included). The image is smoothed with a 2-pixels side b oxcar function to avoid the maximum intensity p eak b eing significantly affected by pixel-to-pixel fluctuations. The statistical errors are calculating by propagating the Poissonian error on each pixel of the diagonal image using the Gehrels' prescription: C = (C + 0.75)1/2 + 1.0 (Gehrels, 1986, ApJ, 303, 336), where C is the numb er of counts in a pixel.

3

Results

In Fig. 1 we present the pile-up fraction as a function of the full band energy band (0.2-12 keV) count rate, R. The dotted lines represent the b est-fit with a the function: Pf = 10a+
b log(R)+c log(R)2

where a, b and c are fit parameters. In Tab. 1 we rep ort the values of these parameters for the main MOS instrumental modes.

4

Comparison with the existing estimates

In Fig. 1 we compare the pile-up fraction as defined in Sect. 2.2 with the prediction by the software used during the evaluation of user prop osals (already known as PHSTools: dashed lines), as well as in the XMM-Newton User's Handb ook (UH). While there is a good agreement b etween the predicted and calculated level of pile-up fr Small Window, the pile-up level is underestimated by ab out 50% in Full Frame and large Window. Up dated thresholds will b e published in the UH Release for AO-13.


XMM-Newton Technical Note

XMM-SOC-CAL-TN-0084

Page:

4

Figure 1: MOS pile-up fraction as a function of the full band count rate for the MOS1 (top panel) and
the MOS2 (bottom panel) cameras. The instrumental modes are colour coded. The dotted lines represent the best-fit of the data (details in text). Arrows indicate upper limits. The dashed vertical lines indicate the nominal thresholds according to the XMM-Newton User Hand-Book. Dashed lines indicate the current model predictions as implemented in the PHSTools (silently called by XMM-Newton proposers when the run the technical evaluation of their proposals).