1

INTEG

RA

L's

Eart

h O

bse

rvati

on

The X-ray Universe 2008The X-ray Universe 2008Granada, 29 May 2008Granada, 29 May 2008

INTEGRAL Spectrum of the INTEGRAL Spectrum of the Cosmic X-ray Background from Cosmic X-ray Background from

Occultation by the EarthOccultation by the Earth

Marc Türler Marc Türler (ISDC, Geneva)(ISDC, Geneva)

(Collaborators:(Collaborators: M. Chernyakova, A. Neronov,M. Chernyakova, A. Neronov,N. Produit, R. Walter, T. CourvoisierN. Produit, R. Walter, T. Courvoisier))

2

INTEG

RA

L's

Eart

h O

bse

rvati

on

INTEGRAL's Earth Observation INTEGRAL's Earth Observation

● 4 similar public observations of ~8 hours at start of revol. 401 404, 405 & 406 in Jan-Feb 2006

● Earth moving through the FoV at ~fixed attitude

● Pointing includes the galactic plane close to the bulge (Norma arm)

● Earth diameter ~10°, but decreasing with time (+ rotation, etc.)

Churazov et al. (2007)

3

INTEG

RA

L's

Eart

h O

bse

rvati

on

Two different approachesTwo different approaches

• Churazov et al. (2007) use Monte-Carlo simulations to model the spectral shape of the Earth emission and its reflection of the CXB (HEAO-1 spectrum). They find a CXB normalization exceeding that of HEAO-1 by about 10%, but cannot constrain the CXB spectral shape.• Our approach uses the spatial distribution of the Earth emission, the galactic ridge and point sources to derive the CXB spectral shape.

Churazov et al. (2007)

4

INTEG

RA

L's

Eart

h O

bse

rvati

on

Sky component modelSky component model

● We model the spatial distribution of 4 components in the IBIS/ISGRI FoV modulated by the Earth occultation:

● The Galactic ridge emission

● All (>2) point sources

● An uniform CXB

● The Earth emission as a smooth transition between

● CXB reflection (at low E.)

● Atmospheric emission (due to cosmic rays)

5

INTEG

RA

L's

Eart

h O

bse

rvati

on

Instrumental components Instrumental components

Additional components are due to the instrument: an uniform (time variable) instrumental background, as well as the modulation by the instrumental responses:

IBIS exposure map IBIS efficiency map(Nomex) at 30 keV

6

INTEG

RA

L's

Eart

h O

bse

rvati

on

Model of detector lightcurvesModel of detector lightcurves

The convolution with the exposure and efficiency maps modifies the detector lightcurve profile:

7

INTEG

RA

L's

Eart

h O

bse

rvati

on

Fitting the model lightcurves Fitting the model lightcurves

Model detector lightcurves are generated for each EO and all spectral channels and are then fitted to the data.

8

INTEG

RA

L's

Eart

h O

bse

rvati

on

Using all 4 Earth observations Using all 4 Earth observations

● We do separate fits to all four EOs which cannot be averaged because of different:

– s/c attitude

– point sources (fixed count rate extracted for each EO)

– instr. background (same variations as seen in the SPI-ACS)

9

INTEG

RA

L's

Eart

h O

bse

rvati

on

Resulting count spectra Resulting count spectra

– To obtain smoother spectra for each EO, we do iterative fits with additional minimization of the dispersion of counts in adjacent energy bins (maximum entropy approach).

– The final count spectra are the average of all 4 EOs with uncertainties based on their dispersion.

10

INTEG

RA

L's

Eart

h O

bse

rvati

on

The resulting CXB spectrum The resulting CXB spectrum

● Unfolded spectra are obtained with XSpec using OSA 7.0 ISGRI ARF & RMF responses

11

INTEG

RA

L's

Eart

h O

bse

rvati

on

ConclusionsConclusions

– The obtained CXB spectrum in the 20-100 keV range is consistent with the HEAO-1 spectrum.

– We do not confirm the about 10% higher normalization found by Churazov et al. (2007) in agreement with the BeppoSAX results of Frontera et al. (2007).

The Galactic ridge emission is also very well constrained and is in qualitative agreement with Krivonos et al. (2007).

Krivonos et al. (2007)