Документ взят из кэша поисковой машины. Адрес оригинального документа : http://xmm.vilspa.esa.es/calibration/documentation/epic_cal_meetings/200106/ferrando_Pile-up.pdf Дата изменения: Wed May 14 13:33:59 2014 Дата индексирования: Wed Apr 13 13:32:28 2016 Кодировка:

Spectral and spatial deformations on piled-up sources

Spectral and spatial deformations on piled-up sources
"a one case study, of MOS data"

EPIC Cal-Ops Meeting - Leicester University, June 7, 2001

Philippe Ferrando, CEA/ SAp


Spectral and spatial deformations on piled-up sources

rationale

Motivations : · A number of sources are with high flux, generating pile-up · Standard spectral and spatial analysis is no more valid, · Large corrections to get absolute flux, spectral shape, spatial distribution...

Plan for this study : · Step 1 : measure pile-up effects on spectra and spatial distribution ... done · Step 2 : check against theoretical predictions... · Step 3 : provide "standard analysis" route for piled-up sources... started future

EPIC Cal-Ops Meeting - Leicester University, June 7, 2001

Philippe Ferrando, CEA/ SAp


Spectral and spatial deformations on piled-up sources

The source

GRS 1758-258 MOS 1 & 2 in small window PN in large window Count rate ~ 20 c/s

EPIC Cal-Ops Meeting - Leicester University, June 7, 2001

Philippe Ferrando, CEA/ SAp


Spectral and spatial deformations on piled-up sources

Central CCDs

MOS 1 21.87 ± 0.03 c/s (stable) (~ 6.6 counts/frame)

MOS 2 21.65 ± 0.03 c/s (stable) (~ 6.6 counts/frame)

EPIC Cal-Ops Meeting - Leicester University, June 7, 2001

Philippe Ferrando, CEA/ SAp


Spectral and spatial deformations on piled-up sources

Ring definitions

· Spectra and counts studied in concentric rings with following characteristics, with counts being for pattern 0 events : R_in arcsec 0 3 6 12 24 R_out arcsec 3 6 12 24 44 Physical pixels 23 70 280 1122 3531 counts /frame 0.73 1.04 1.42 1.03 0.64 c/frame /pixel 3.12e-2 1.48e-2 5.07e-3 9.15e-4 1.80e-4

· Look for spectral differences in different rings · All study performed with pattern 0 events only, since pattern 0 the less sensitive to pile-up, especially regarding spectral effects

EPIC Cal-Ops Meeting - Leicester University, June 7, 2001

Philippe Ferrando, CEA/ SAp


Spectral and spatial deformations on piled-up sources

Encircled energy refer ence

Results will be compared to encircled energy data from file XRT1_XENCIREN_0000.CCF (merci Jean) This file show a strong energy dependence in the encircled energy

EPIC Cal-Ops Meeting - Leicester University, June 7, 2001

Philippe Ferrando, CEA/ SAp


Spectral and spatial deformations on piled-up sources

Background

EPIC Cal-Ops Meeting - Leicester University, June 7, 2001

Philippe Ferrando, CEA/ SAp


Spectral and spatial deformations on piled-up sources

Raw spectra comparison

EPIC Cal-Ops Meeting - Leicester University, June 7, 2001

Philippe Ferrando, CEA/ SAp


Spectral and spatial deformations on piled-up sources

Simple fit inner core

EPIC Cal-Ops Meeting - Leicester University, June 7, 2001

Philippe Ferrando, CEA/ SAp


Spectral and spatial deformations on piled-up sources

Simple fit outer ring

EPIC Cal-Ops Meeting - Leicester University, June 7, 2001

Philippe Ferrando, CEA/ SAp


Spectral and spatial deformations on piled-up sources

Ratio spectrum inner core to reference

EPIC Cal-Ops Meeting - Leicester University, June 7, 2001

Philippe Ferrando, CEA/ SAp


Spectral and spatial deformations on piled-up sources

Ratio spectrum 3-6 arcsec to reference

EPIC Cal-Ops Meeting - Leicester University, June 7, 2001

Philippe Ferrando, CEA/ SAp


Spectral and spatial deformations on piled-up sources

Ratio spectrum 6-12 arcsec to reference

EPIC Cal-Ops Meeting - Leicester University, June 7, 2001

Philippe Ferrando, CEA/ SAp


Spectral and spatial deformations on piled-up sources

Ratio spectrum 12-24 arcsec to reference

EPIC Cal-Ops Meeting - Leicester University, June 7, 2001

Philippe Ferrando, CEA/ SAp


Spectral and spatial deformations on piled-up sources

Model of flux loss 1/2

· Flux loss and pile-up was modelled using Jean Ballet pile-up program, together with the 1-D PSF calculated from the encircled energy CCF file and the pattern distribution at 1.5 keV · Incoming flux was adjusted so that modelled observed rate equals the actual value of ~ 6 / frame

EPIC Cal-Ops Meeting - Leicester University, June 7, 2001

Philippe Ferrando, CEA/ SAp


Spectral and spatial deformations on piled-up sources

Model of flux loss 2/2

Numerical results from model are : · Incident flux of 8.0 cts/frame needed to get 6.2 cts/frame detected, i.e. that the average flux loss is ~ 22 %for all patterns · Pile-up and flux loss dependence for mono-pixels can be large : Ring (arcsec) central pixel 0.0 - 3.3 3.3 - 5.5 5.5 - 12.1 12.1 - 24.2 24.2 - 44.0 piled-up fraction 16.9 % 3.4 % 1.7 % 0.39 % 0.067 % 0.0074 % flux-loss 98.8 % 65.2 % 34.4 % 9.2 % 1.6 % 0.2 %

EPIC Cal-Ops Meeting - Leicester University, June 7, 2001

Philippe Ferrando, CEA/ SAp


Spectral and spatial deformations on piled-up sources

Flux loss correction and data

EPIC Cal-Ops Meeting - Leicester University, June 7, 2001

Philippe Ferrando, CEA/ SAp


Spectral and spatial deformations on piled-up sources

Conclusions

Conclusions
· Spectral distorsion strong even when pattern 0 selected · The energy dependence of the encircled energy CCF function does not seem to fit these data - A flatter energy dependence in the wings would do better... · The first rough attempt to use Jean's model looks promising, at least for flux loss corrections. · Pile-up is essentially a local effect, and a good 2-D PSF description is needed to go further in modelling. · Work need to be redone on more sources...

EPIC Cal-Ops Meeting - Leicester University, June 7, 2001

Philippe Ferrando, CEA/ SAp