Документ взят из кэша поисковой машины. Адрес оригинального документа : http://xmm.vilspa.esa.es/calibration/documentation/epic_cal_meetings/201504/Maitra_Cal_2015.pdf Дата изменения: Fri Jul 17 14:47:32 2015 Дата индексирования: Wed Apr 13 13:32:03 2016 Кодировка: Поисковые слова: comet

FLAGGING ANOMALOUS CCDS IN MOS:
IN CONTINUATION WITH THE PATH OF EMTAGLENOISE

Chandreyee Maitra 2015 EPIC calibrations and operating meeting ESAC, April 9th 2015
1


LOW ENERGY NOISE ON MOD CCDS
a.k.a anomalous state in CCDs (Kuntz & Snowden 2008)
!

Time variable component in the EPIC-MOS non cosmic particle background
Analysis of large extended sources difficult when
the noise is present
Interferes with the source detection in PPS and spurious
sources detected

!
2


LOW ENERGY NOISE ON MOD CCDS
Work done by Kuntz & Snowden while modeling the QPB

CHARACTERISTICS
Enhanced count rate at E < 1 keV and low hardness ratio.
Most affected MOS 2 CCD 5 and MOS 1 CCD 4 but found in other CCDs as well.
No correlation with radiation monitor or HK parameters found exact cause unknown.

3


MOS1 CCD 2
The CCDs in "anomalous state" quasi uniformly
brighter than its neighbors
Spatial distribution is fairly homogenous within &
outside the F.O.V
! Event times consistent with Poisson statistics
Occurs on CCD 2-7 : Plateau with break ~ 1-2 keV
One or more CCDs may be affected in an ObsID

!

MOS 2 CCD 5

Chen & Ballet

4


EMTAGLENOISE
Based on Kuntz & Snowden's method of identification

OBSID : out of FOV calculate "H": (2.5-5.0 keV)/(0.5-0.8 keV) calculate "R": (0.3-10.0 keV) Test if "H" and "R" meets "Anomalous" criteria

Method based on criteria applied to each CCD independently in an observation

5


TO INVESTIGATE MARKED CHANGES IN THE NOISE CHARACTERISTICS
Speculation that noise characteristics changed over time? Break energy shifted to higher energies.. Criteria which compares CCD within themselves in an ObsID : method independent of noise characteristics More robust, sensitive ?

IN CONTINUATION TO THE PATH:

6


Exercise performed on a set of observations which showed noise characteristics of two types in the same ObsID
0116340601 : 30 ks Observation MOS 2 CCD 5

MOS 2 CCD 2
7


ALGORITHM
OBSID : out of FOV calculate "cr[i]": counts/area/exposure check cr[i] > 2* (median[cr[k!=i]] +dcr[i]) calculate N=cr[i]/median[i]
Provide additional information like N, Break E to users Method based on criteria on inter CCD comparison: optimum energy band for procedure : starting 0.3 or 0.5 keV ? 8


COMPARISON OF THE TWO METHODS :OBSID 0116340601
0.5-0.8 KEV
CCD N O I S E _ F R AC ERROR H

(2.5-5.0 KEV)/(0.5-0.8 KEV)
ERROR

2 3 4 5 6 7

18.36 0.66 0.60 19.53 0.74 1.35

13.26% 22.80% 21.76% 13.07% 22.43% 20.50%

0.4 6.46 7.59 2.45 5.40 4.79

8% 22% 20% 4.5% 22% 18%
9


Kuntz & Snowden method more sensitive: smaller errors Our only advantage is assurance of triggering noisy CCDs with different spectral range
10


For all the observations provided which has noisy CCDs > 1 keV : MOS2 CCD 5 shows this effect reported in Kuntz & Snowden 2008 as `VERIFICATION HIGH STATE' All the ObsIDs Rev 25-30 ( < 42 as in Kuntz & Snowden 08) We have no evidence of recent ObsIDs with similar characteristics.
MOS 2 CCD 5 : Verification high state

11

Kuntz & Snowden 2008


SUMMARY
We have devised an alternative method to identify the presence of low energy noise in MOS CCDs. Our method is more robust in identifying noise over different spectral ranges but is less sensitive than the previous method. However at present we have no evidence of the presence of such characteristics of noise in the recent observations to implement it formally.

12


Thank you

13