On Disk Reflection and Cross-Calibration

ATel #1966; J. M. Miller (University of Michigan)
on 12 Mar 2009; 16:18 UT
Credential Certification: Jon Miller (jonmm@umich.edu)

Tweet

RXTE is an exceptionally well-calibrated mission. Over the 3-25 keV bandpass, power-law models for emission from the Crab nebula reveal deviations of 0.6% or less. The Xe L3 edge at 4.8 keV is sometimes apparent in data, but this is easily modeled with a simple edge. Owing to its calibration and its longevity - over 12 years of operation, RXTE has overlapped with ASCA, BeppoSAX, Chandra, XMM-Newton, INTEGRAL, Swift, and Suzaku - RXTE serves as an effective standard against which the spectral response of new missions can be compared.

Disk reflection spectra are a useful test of spectral response in that both low energy (a disk line and photoelectric absorption) and high energy features (the Compton back-scattering hump peaking at 20-30 keV) are expected. RXTE has facilitated the detection of broad-band disk reflection spectra in virtually every black hole candidate that it has observed. In the case of GX 339-4, for instance, simultaneous RXTE spectra helped to define the continuum and evaluate the robustness of disk reflection features in one long Chandra observation (Miller et al. 2004 ApJ 601 450), two long XMM-Newton observations (Miller et al. 2004 ApJ 606 L131, Miller et al. 2006 ApJ 653 525), and recent observations with Swift (Tomsick et al. 2008 ApJ 680 593).

Suzaku is well-matched to RXTE by virtue of its broad spectral bandpass. It is worth examining, then, how simultaneous RXTE and Suzaku observations compare. Suzaku observed GX 339-4 once during its last outburst (on 2007 Feb. 12), finding strong disk reflection features (Miller et al. 2008 ApJ 679 L113). RXTE observed GX 339-4 at the same time. Simple disk blackbody plus power-law fits to the Suzaku XIS and HXD spectra give power-law indices of Gamma = 2.2, 2.3, and 2.4 (XIS0, XIS1, HXD); simultaneous fits to RXTE PCA and HEXTE-B spectra find a power-law index of Gamma = 2.4. This degree of agreement is excellent, given the substantial differences between the cameras.

The figure linked HERE shows the data/model ratio that results when the Suzaku and RXTE spectra are fit with the simple disk blackbody ("diskbb") plus power-law model. The 4-7 and 15-45 keV ranges were ignored in fitting the spectra in order to evaluate the reflection spectrum in a model-independent way. (The low resolution of the PCA required a Gaussian to be fit to the iron line in the RXTE spectrum; its normalization was set to zero to form the ratio.) The iron emission line and Compton back-scattering hump are consistent, both in magnitude and shape. Asymmetry in the iron line is only clearly detected with Suzaku, owing to its superior spectral resolution.

The fact that the shape of the hard continuum and disk reflection spectra agree signals that the broad spectral response of Suzaku is well-calibrated. It also signals that bright transients can be observed with the appropriate Suzaku camera modes without severe distortions from photon pile-up. Similar strong agreement has been obtained from observations of Cygnus X-1 (Nowak 2008, 7th Microquasar Workshop Proceedings, arxiv:0810.1519).

Finally, it must be noted that even when the same value of N_H is required for the XIS0, XIS1, and the PCA cameras, different inner disk temperatures (kT = 0.83, 0.76, 1.02 keV) and flux normalizations (K = 740, 810, 460) are found. Difficulties of this kind are common when comparing simultaneous observations; it is standard practice to let a constant factor float between detectors to account for flux offsets (see, e.g., Kubota et al. 2005 ApJ 630 1062). The results briefly noted here indicate that systematic errors on inner disk radii derived using simple models can be as high as 30% due to differing detector flux zero-points.