EPIC flux comparisonfrom 2XMM sources

S. Mateos, R. Saxton, S. Sembay & A. Read

Sources Used• Point-like sources from 2XMM detected in 2+ cameras

• > 200 counts in each camera

• Off-axis angle 0-12 arcmins

• F2-10 > 6E-12 have been excluded to avoid pile-up effects

Count rate – flux conversion

Count rates converted to fluxes using energy conversion factors (ECF) which are based on a spectral model of an absorbed power-law with

NH=3E20, slope=1.7

ECFs calculated using the detector matrices:

MOS: On-axis RMF for revolution 375 + on-axis ARF

PN: Latest, canned, on-axis, full-frame RMF for single andsingle+double events + on-axis ARF

Count rates found with - MOS: pattern=0-12, PN: 0.2-0.5 keV, pattern=0; 0.5-12 keV, pattern=0-4

PN v MOS-1: Band 3 (1-2 keV)

PN v MOS-1: Band 1 (0.2-0.5)

PN v MOS-1: Band 2 (0.5-1)

PN v MOS-1: Band 4 (2-4.5)

PN v MOS-1: Band 5 (4.5-12)

PN v MOS-1: Flux comparison

PN v MOS-2: Flux comparison

MOS-1 v MOS-2: Flux comparison

Flux Ratios (%)

Energy (keV) (m1-pn)/m1 (m2-pn)/m2 (m2-m1)/m1

0.2 - 0.5 2.7±0.6 0.9±0.4 -1.3±0.4

0.5 - 1.0 8.4±0.1 8.4±0.2 0.8±0.2

1.0 - 2.0 8.8±0.2 9.4±0.2 0.3±0.2

2.0 - 4.5 7.3±0.2 6.7±0.2 -0.8±0.2

4.5 - 12.0 12.5±0.4 9.0±0.4 -3.7±0.3The Kirsch relation: mos = k * pn where k is an energy independent constant, ~1.05 – 1.08

CAL-TN-0052-5 (Stuhlinger et al. 2008)

First Results‡ MOS cameras agree to better than 4% at all energies.

‡ PN has a ~constant offset from MOS cameras of 7-9% from 0.5-4.5 keV

‡ PN / MOS agreement much better (<3%) in 0.2-0.5 keV band

‡ PN / MOS agreement worse at high energies at least for MOS-1 (12.5%)

Low-Energy difference

Why so good ??

Is the use of a single RMF ok ?

Reminder:

MOS flux conversion uses RMF for on-axis (i.e. on patch) at Rev 0375.

PN: Uses on-axis (Y=9) RMF

These approximations will mainly effect low energies.

PN v MOS-1: Change with time

PN v MOS-2: Change with time

MOS-1 v MOS-2: Change with time

PN v MOS-1: Off-axis angle

Iufh

Yth

Tj

Tyj

e

PN v MOS-2: Off-axis angle

F

F

F

Low-Energy summary

• Ignoring sources which fall on the MOS patches, i.e. using Θ = 2 – 12 arcmins we get:

(m1-pn)/m1 = 10 - 12%

(m2-pn)/m2 = 2 - 7%

Time variability makes these numbers unreliable but m2/pn looks to be less than ~8%

PN v MOS-1: Flux comparison

?

High-Energy difference

(m1-pn) / m1=12.5% Why so high ??

Is the Kirsch relation wrong ??

PN v MOS-1: Off-axis angle

PN v MOS-2: Off-axis angle

MOS-1 v MOS-2: Off-axis angle

What depends on off-axis angle?

• Vignetting (all cameras)

• RGS obscuration (MOS)

• PSF (all cameras)

Azimuthal-angle dependent

Azimuthal-angle dependent (MOS)

MOS PSF

A measure of the XMM PSFs, lighter colour means a sharper PSF.

MOS CCDs

A measure of the XMM PSFs, lighter colour means a sharper PSF.

1

2

3

4 1

2

3

4

PN v MOS-1: Azimuthal angle

PN v MOS-2: Azimuthal angle

Conclusions for MOS / PN

• MOS = PN * 1.08 from 0.5 – 4.5 keV

• With this analysis we can’t say what the relation is in the 0.2-0.5 keV band.

• At high energies there is an extra off-axis angle, azimuthal-angle dependent effect which increases the MOS excess. This aligns with the RGS dispersion direction and probably means that the RGS absorption needs recalibrating.