2002ApJ...574..663M

We present deep Chandra spectral imaging of the dwarf starburst galaxy NGC 1569. The unprecedented spatial resolution allows us to spatially identify the components of the integrated X-ray spectrum. Fitted spectral models require an intrinsic absorption component and higher metal abundances than previous studies indicated. Our results provide the first direct evidence for metal-enriched winds from dwarf starburst galaxies. We identify 14 X-ray point sources in NGC 1569. Most have properties consistent with those of high-mass X-ray binaries, but one is a steep-spectrum radio source that is probably a supernova remnant. The X-ray luminosity of NGC 1569 is dominated by diffuse, thermal emission from the disk (0.7 keV) and bipolar halo (0.3 keV). Photoelectric absorption from the inclined H I disk hardens the X-ray spectrum on the northern side of the disk relative to the southern side. Requiring the fitted absorption column to match the H I column measured at 21 cm implies that the metallicity of the H I disk is significantly less than solar but greater than 0.1 Z_☉. Hence, much of the H I is enriched to levels comparable to the metallicity of the H II regions [O/H=0.2(O/H)_☉]. The X-ray color variations in the halo are inconsistent with a free-streaming wind and probably reveal the location of shocks created by the interaction of the wind with a gaseous halo. The X-ray spectrum of the diffuse gas presents strong emission lines from α-process elements. Fitted models require α-element abundances greater than 0.25 Z_α,☉ and ratios of α-elements to iron 2-4 times higher than the solar ratio. The best fit to the spectrum is obtained with solar mass fractions for the α-elements, 1.0 Z_α,☉, but a degeneracy between the metallicity and the spectral normalization prevents us from deriving an upper limit on the wind metallicity from the X-ray spectrum alone. We argue, however, that abundances larger than 2.0 Z_α,☉ pose awkward implications for the dynamical evolution of the wind based on our knowledge of the starburst properties. For consistency with our best-fitting abundances, the mass of interstellar gas entrained in the wind must be about 9 times the mass of stellar ejecta in the wind. Most of the oxygen carried by the wind comes from the stellar ejecta rather than entrained interstellar gas. The estimated mass of oxygen in the hot wind, 34,000 M_☉, is similar to the oxygen yield of the current starburst. Apparently the wind carries nearly all the metals ejected by the starburst. These metals appear destined to contribute to the enrichment of the intergalactic medium. Much of the nucleosynthesis in NGC 1569 must have occurred during less violent periods of star formation, however, because our measurements imply that the neutral gas disk holds at least 5 times more oxygen than wind.