Oxygen Absorption in M87: Evidence for a Warm+Hot ISM

Abstract

We present a re-analysis of the ROSAT PSPC data within the central 100 kpc of M87 to search for intrinsic oxygen absorption similar to that recently measured in several galaxies and groups. Using a spatial-spectral deprojection analysis we find the strongest evidence to date for intrinsic oxygen absorption in the hot gas of a galaxy, group, or cluster. Single-phase plasma models modified by intervening Galactic absorption cannot fit the 0.2-2.2 keV PSPC data as they under-predict the 0.2-0.4 keV region and over-predict the 0.5-0.8 keV region where the emission and absorption residuals are obvious upon visual inspection of the spectral fits. Since the excess emission between 0.2-0.4 keV rules out intrinsic absorption from cold gas or dust, the most reasonable model for the excess emission and absorption features is warm, collisionally ionized gas with a temperature of ~10^6 K. Simple multiphase models (cooling flow, two phases) modified by both intervening Galactic absorption and by a single oxygen edge provide good fits and yield temperatures and Fe abundances of the hot gas that agree with previous determinations by ASCA and SAX. The multiphase models of M87 inferred from the PSPC can account for the excess EUV emission observed with EUVE and the excess X-ray absorption inferred from EINSTEIN and ASCA data above 0.5 keV. Although the total mass of the warm gas implied by the oxygen absorption is consistent with the matter deposited by a cooling flow, the suppression of the mass deposition rate and the distortion of the X-ray isophotes in the region where the radio emission is most pronounced suggest some feedback effect from the AGN on the cooling gas. (Abridged)