2003ApJ...595..151B

Using new XMM and Chandra observations, we present an analysis of the metal abundances of the hot gas within a radius of 100 kpc of the bright nearby galaxy group NGC 5044. Motivated by the inconsistent abundance and temperature determinations obtained by different observers for X-ray groups, we provide a detailed investigation of the systematic errors on the derived abundances considering the effects of the temperature distribution, calibration, plasma codes, bandwidth, Galactic N_H, and background rate. The iron abundance (Z_Fe) drops from Z_Fe~1 Z_☉ within R~50 kpc to Z_Fe~0.4 Z_☉near R=100 kpc. This radial decline in Z_Feis highly significant: Z_Fe=1.09±0.04(statistical)±0.05+0.18(systematic) in solar units within R=48 kpc (5') compared to Z_Fe=0.44±0.02(statistical)±0.10+0.13(systematic) in solar units over R=48-96 kpc (5'-10'). There is no evidence that the radial profile of Z_Feflattens at large radius. The data rule out with high confidence a very subsolar value for Z_Fewithin R=48 kpc, confirming that previous claims of very subsolar central Z_Fevalues in NGC 5044 were primarily the result of the Fe bias, i.e., the incorrect assumption of spatially isothermal and single-phase gas when in fact temperature variations exist. Next to iron the data provide the best constraints on the silicon and sulfur abundances. Within R=48 kpc we obtain Z_Si/Z_Fe=0.83±0.02(statistical)±0.02+0.07(systematic) and Z_S/Z_Fe=0.54±0.02(statistical)±0.01+0.01(systematic) in solar units. These ratios (1) are consistent with their values at larger radii, (2) strongly favor convective deflagration models over delayed detonation models of Type Ia supernovae (SNe Ia), and (3) imply that SNe Ia have contributed 70%-80% of the iron mass within a 100 kpc radius of NGC 5044. This SN Ia fraction is also similar to that inferred for the Sun and therefore suggests a stellar initial mass function similar to that of the Milky Way. We mention that at the very center (R~2 kpc) the XMM and Chandra CCDs and the XMM Reflection Grating Spectrometer (RGS) show that the Fe and Si abundances drop to ~50% of their values at immediately larger radius analogously to that seen in some galaxy clusters observed with Chandra. We find the magnitude of this dip to be sensitive to assumptions in the spectral model, but if real it is difficult to reconcile with the expectation that metal enrichment from the stars in the central galaxy should result in a centrally peaked metal abundance profile in the hot gas.