1. Marcus Kirsch Science Operations & Data Systems Division Research & Scientific Support Department Page 1 XMM-Newton MOS Timing Mode calibration Marcus G. F. Kirsch, B. Altieri, R. Saxton

2. Marcus Kirsch Science Operations & Data Systems Division Research & Scientific Support Department Page 2 XMM-Newton EPIC menu energy calibration flux calibration MOS Timing versus pn

3. Marcus Kirsch Science Operations & Data Systems Division Research & Scientific Support Department Page 3 XMM-Newton EPIC Situation last cal meeting black: LW red: Timing trend is similar for both MOSs Timing mode is overcorrected by up to 1.5 %

4. Marcus Kirsch Science Operations & Data Systems Division Research & Scientific Support Department Page 4 XMM-Newton EPIC reason for over correction CTI correction is performed for events with y-values from 0-1024 (just a counter with no meaning in that mode) with the mean of ~ 500 CTI correction should however be performed for y=300 for ALL events Now implemented in SAS 6.0

5. Marcus Kirsch Science Operations & Data Systems Division Research & Scientific Support Department Page 5 XMM-Newton EPIC 1E0102-SAS 6.0 0.4-2.0 keV: Timing mode agrees with LW mode within +/- 0.5 %.

6. Marcus Kirsch Science Operations & Data Systems Division Research & Scientific Support Department Page 6 XMM-Newton EPIC SS433: SAS 6.0 M1 Timing Medium M2 FF Medium M1 Timing Medium M2 FF Medium pn SW Medium pn(Timing) > MOS ? pn(Timing) = pn(FF)± 0.3 % at 6 keV ~ +0.3 %

7. Marcus Kirsch Science Operations & Data Systems Division Research & Scientific Support Department Page 7 XMM-Newton EPIC Cas-A analysis 2002 pn < Chandra < MOS Rational: –CAS-A is very bright –MOS CTI model does not take precursors into account  CTI over correction contradictory to SS433 ???

8. Marcus Kirsch Science Operations & Data Systems Division Research & Scientific Support Department Page 8 XMM-Newton EPIC Flux: Calibration source MOS1 flux in Timing mode at Mn seems to be 10 % lower

9. Marcus Kirsch Science Operations & Data Systems Division Research & Scientific Support Department Page 9 XMM-Newton EPIC Flux: NGC 5548 (rev 290) MOS Timing Medium M2 SW Medium pn SW Medium  Fit:const*(model_linked) –M1: 1.0 –M2: 1.02 –PN: 1.16  below 0.6 keV: – MOSs agree – but less flux than pn  above 8 keV problem in MOS Timing mode

10. Marcus Kirsch Science Operations & Data Systems Division Research & Scientific Support Department Page 10 XMM-Newton EPIC Flux: SS433 (rev 610) pn Timing Medium M1 Timing Medium M2 FF Medium Medium  Fit const*(model_linked) –pn: 1.0 –M1: 0.94 –M2: 0.52 (pileup)  below 0.6 keV: problem in MOS Timing mode  above 8 keV problem in MOS Timing mode

11. Marcus Kirsch Science Operations & Data Systems Division Research & Scientific Support Department Page 11 XMM-Newton EPIC Flux: HER-X1 (rev 207)  pn:1  MOS: 0.9  flux in 0.315- 10.0 keV – pn: 1.66E-09 erg cm -2 s -1 – MOS: 1.55E-09 erg cm -2 s -1 – MOS/pn: 0.93 pn Timing MOS1 Timing Fit with linked parameters and constFit with free parameters and const=1

12. Marcus Kirsch Science Operations & Data Systems Division Research & Scientific Support Department Page 12 XMM-Newton EPIC Summary Energy calibration: MOS timing mode energy calibration agrees within ± 0.3 % with MOS-LW data at ~ 6-7 keV –pn Timing mode > MOS Timing mode by 0.3 % (SS433) –pn Timing mode > pn FF mode by 0.3 % (Cas-A) –MOS FF mode should agree there with MOS Timing mode if source is not extended and bright For bright extended sources MOS would need a rate dependent CTI model ! Rev 832: Cas-A observation in –pn: FF and extFF –MOS: Timing and LW Flux calibration: Discrepancy of up to 10-15 % at energies below 1keV Contradictory results for different targets and observations Cas-A data for all EPIC modes + other SNR data will allow overall energy calibration analysis Is the MOS timing mode not stable in flux ?