The lack of robust constraints on the radial temperature profiles of hot gas contained within galaxy clusters is probably the major uncertainty in the present determination of dynamical properties of clusters, which hinders clusters from acting as an ideal laboratory of testing theories of formation and evolution of structures in the universe including a direct estimate of the cosmic mass density parameter by combining the baryon fraction measurement and the Big Bang Nucleosynthesis. Indeed, previous studies have arrived at conflicting results regarding the radial temperature gradients in clusters. By analyzing 30 clusters observed with ASCA, Markevitch et al. (1998) claimed a significant temperature decline with radius quantified by a polytropic index of 1.2-1.3 on the average. However, subsequent studies have soon raised doubt about the ubiquity and steepness of these temperature decline: Irwin, Bregman & Evrard (1999) carried out an analysis of the color profiles of the same clusters used by Markevitch et al. (1998) but found an essentially flat temperature profile. Applying the spectral-imaging deconvolution method to a large sample of 106 ASCA clusters, White (2000) has showed that 90 percent of the temperature profiles are actually consistent with isothermality. Further argument against the nonisothermality of intracluster gas has been put forward recently by Irwin & Bregman (2000), who reported the detection of a flat and even increasing temperature profile out to of the viral radius for a sample of 11 clusters observed with BeppoSAX.
Theoretically, it deserves an investigation into the possibility of deriving the radial temperature profiles of intracluster gas from the well-motivated physical mechanisms, incorporated with the X-ray imaging observations. This may provide a valuable clue to resolving the above temperature profile discrepancy. There are two well-established facts on which we can rely today: (1)The gravitational potential of a cluster is dominated by the dark matter distribution which can be described by the so-called universal density profile, as suggested by a number of high-resolution simulations (Navarro, Frenk & White 1995 and hereafter NFW), although the innermost slope is still under debate. (2)The azimuthally-averaged X-ray surface brightness of a cluster is reliably measurable out to several or even times as large as the X-ray core radius, for which a good approximation is provided by the conventional model (Cavaliere & Fusco-Femiano 1976). These two facts, along with the hydrostatic equilibrium hypothesis and a reasonable choice of the boundary conditions, permit a unique determination of the gas temperature profile (Wu & Chiueh 2000). On the other hand, a comparison of the theoretically expected temperature profile with the result from the X-ray spectroscopic measurement constitutes a critical test for the validity of the NFW profile and the hydrostatic equilibrium in clusters.
In this Letter , we will attempt for the first time to derive the temperature profiles of 3 well-defined clusters with good X-ray imaging observations extending to relatively large radii, based on the method developed by Wu & Chiueh (2000). Our derived temperature profiles will be compared with the recent results of 11 clusters observed with BeppoSAX (Irwin & Bregman 2000). We will examine the possible similarity in the gas temperature profiles as a result of the underlying structural regularity (e.g. Neumann & Arnaud 1999). The implication of our results for the reported temperature profile discrepancy will be discussed. Throughout the Letter we assume km s-1 Mpc-1 and a flat cosmological model with and .
© European Southern Observatory (ESO) 2000
Online publication: August 23, 2000