HiDensity Distribution Driven by Supernovae: A Simulation Study

Restricted Access till July 2009. (3 year embargo). Open access version available at arXiv.orgWe model the complex distribution of atomic hydrogen (H I) in the interstellar medium (ISM) assuming that it is driven entirely by supernovae (SNe). We develop and assess two different models. In the first approach, the simulated volume is randomly populated with nonoverlapping voids of a range of sizes. This may relate to a snapshot distribution of SN-remnant voids, although somewhat artificially constrained by the nonoverlap criterion. In the second approach, a simplified time evolution (considering momentum conservation as the only governing constraint during interactions) is followed as SNe populate the space with the associated input mass and energy. We describe these simulations and present our results in the form of images of the mass and velocity distributions, and the associated power spectra. The latter are compared with trends indicated by available observations. In both approaches, we find remarkable correspondence with the observed statistical description of well-studied components of the ISM, wherein the spatial spectra have been found to show significant deviations from the Kolmogorov spectrum. One of the key indications of this study, regardless of whether or not the SN-induced turbulence is the dominant process in the ISM, is that the apparent non-Kolmogorov spectral characteristics (of H I and/or electron column density across thick or thin screens) needed to explain related observations may not be in conflict at all with the underlying turbulence (i.e., the velocity structure) being of a Kolmogorov nature. We briefly discuss the limitations of our simulations and the various implications of our results