On the Dynamical and Physical State of the “Diffuse Ionized Medium” in Nearby Spiral Galaxies

Open Access
Publisher Website
Google Scholar
Abstract
We report the initial results from a program to study the morphology, physical state, and kinematics of the "diffuse ionized medium" (DIM) in a sample of the nearest and brightest late-type galaxies. For each of five galaxies (NGC 2403, M81, NGC 4395, M51, and M101), we have analyzed deep narrowband Hα images of the entire star-forming disk and long-slit spectra of the inner (~10 kpc) disk with a resolution of 40-75 km s-1. We find that the DIM covers most of the star-forming disk and is morphologically related to the presence of high surface brightness gas (the giant H II regions). The DIM and the giant H II regions differ systematically in their physical and dynamical state. The DIM is characterized by enhanced emission in the low-ionization forbidden lines ([O I], [N II], and [S II]), and even the high-ionization [O III] λ5007 line is moderately strong in the DIM in at least three cases. This last result contrasts with upper limits on the [O III] surface brightness in the local DIM of our own Galaxy (the "Reynolds Layer"). We directly verify the inference made by Lehnert and Heckman that the DIM contributes significantly to the spatially integrated (global) emission-line ratios measured in late-type galaxies. We also find that the DIM is more disturbed kinematically than the gas in the giant H II regions. The deconvolved (intrinsic) widths of the Hα and [N II] λ6584 lines range from 30 to 100 km s-1 (FWHM) in the DIM compared to 20-50 km s-1 in the giant H II regions. The high-ionization gas in the DIM is more kinematically disturbed than the low-ionization gas: the [O III] λ5007 lines have intrinsic widths of 70-150 km s-1. The differing kinematics implies that the DIM is not a single monolithic phase of the ISM. Instead, it may consist of a "quiescent" DIM with a low ionization state and small scale height (few hundred parsec) and a "disturbed" DIM with a high ionization state and moderate scale height (0.5-1 kpc). We argue that the quiescent DIM is most likely photoionized by radiation from O stars leaking out of giant H II regions (although this requires fine-tuning the opacity of galactic disks to ionizing radiation). The disturbed DIM is most likely heated by the mechanical energy supplied by supernovae and stellar winds. Since the disturbed DIM accounts for only a minority (<20%) of the Hα emission in the regions we have studied, there is no fundamental energetics problem with this model, but it does require mechanically heated gas to have a high areal covering factor in the inner disk (which needs to be confirmed observationally). We find no clear discontinuity between the physical and dynamical properties of the giant H II regions and the quiescent DIM. The quiescent DIM is morphologically related to the giant H II regions and there is a smooth dependence of the emission-line ratios and emission-line widths on the surface brightness of the emission. Thus, we suggest that a unified approach to the study of the DIM and giant H II regions in star-forming galaxies will prove fruitful.
All Related Versions