M. Ryan Joung

Turbulent Structure of a Stratified Supernova-driven Interstellar Medium

To study how supernova feedback structures the turbulent interstellar medium, we construct 3D models of vertically stratified gas stirred by discrete supernova explosions, including vertical gravitational fields and parameterized heating and cooling. The models reproduce many observed characteristics of the Galaxy, such as global circulation of gas (i.e., galactic fountain) and the existence of cold dense clouds in the galactic disk. Global quantities of the model such as warm and hot gas filling factors in the midplane, mass fraction of thermally unstable gas, and the averaged vertical density profile are compared directly with existing observations and shown to be broadly consistent. We find that energy injection occurs over a broad range of scales. There is no single effective driving scale, unlike the usual assumption for idealized models of incompressible turbulence. However, >90% of the total kinetic energy is contained in wavelengths shortward of 200 pc. The shape of the kinetic energy spectrum differs substantially from that of the velocity power spectrum, which implies that the velocity structure varies with the gas density. Velocity structure functions demonstrate that the phenomenological theory proposed by Boldyrev is applicable to the medium. We show that it can be misleading to predict physical properties such as the stellar initial mass function based on numerical simulations that do not include self-gravity of the gas. Even if all the gas in turbulently Jeans-unstable regions in our simulation is assumed to collapse and form stars in local free-fall times, the resulting total collapse rate is significantly lower than the value consistent with the input supernova rate. Supernova-driven turbulence inhibits star formation globally rather than triggering it.

Vertical Structure of a Supernova-driven Turbulent, Magnetized Interstellar Medium

Stellar feedback drives the circulation of matter from the disk to the halo of galaxies. We perform three-dimensional magnetohydrodynamic simulations of a vertical column of the interstellar medium with initial conditions typical of the solar circle in which supernovae drive turbulence and determine the vertical stratification of the medium. The simulations were run using a stable, positivity-preserving scheme for ideal MHD implemented in the FLASH code. We find that the majority (90%) of the mass is contained in thermally stable temperature regimes of cold molecular and atomic gas at T 3 kpc. The magnetic field in our models has no significant impact on the scale heights of gas in each temperature regime; the magnetic tension force is approximately equal to and opposite the magnetic pressure, so the addition of the field does not significantly affect the vertical support of the gas. The addition of a magnetic field does reduce the fraction of gas in the cold (<200 K) regime with a corresponding increase in the fraction of warm (~104 K) gas. However, our models lack rotational shear and thus have no large-scale dynamo, which reduces the role of the field in the models compared to reality. The supernovae drive oscillations in the vertical distribution of halo gas, with the period of the oscillations ranging from 30 Myr in the T < 200 K gas to ~100 Myr in the 106 K gas, in line with predictions by Walters & Cox.

GAS CONDENSATION IN THE GALACTIC HALO

Using adaptive mesh refinement (AMR) hydrodynamic simulations of vertically stratified hot halo gas, we examine the conditions under which clouds can form and condense out of the hot halo medium to potentially fuel star formation in the gaseous disk. We find that halo clouds do not develop from linear isobaric perturbations. This is a regime where the cooling time is longer than the Brunt-V?is?l? time, confirming previous linear analysis. We extend the analysis into the nonlinear regime by considering mildly or strongly nonlinear perturbations with overdensities up to 100, also varying the initial height, the cloud size, and the metallicity of the gas. Here, the result depends on the ratio of cooling time to the time required to accelerate the cloud to the sound speed (similar to the dynamical time). If the ratio exceeds a critical value near unity, the cloud is accelerated without further cooling and gets disrupted by Kelvin-Helmholtz and/or Rayleigh-Taylor instabilities. If it is less than the critical value, the cloud cools and condenses before disruption. Accreting gas with overdensities of 10-20 is expected to be marginally unstable; the cooling fraction will depend on the metallicity, the size of the incoming cloud, and the distance to the galaxy. Locally enhanced overdensities within cold streams have a higher likelihood of cooling out. Our results have implications on the evolution of clouds seeded by cold accretion that are barely resolved in current cosmological hydrodynamic simulations and absorption line systems detected in galaxy halos.

Galaxy Size Problem at z = 3: Simulated Galaxies are too Small

Using state-of-the-art adaptive mesh refinement cosmological hydrodynamic simulations with a spatial resolution of proper 0.21h –1 73 kpc in refined subregions embedded within a comoving cosmological volume (27.4h –1 73 Mpc)3, we investigate the sizes of galaxies at z = 3 in the standard cold dark matter model. Our simulated galaxies are found to be significantly smaller than the observed ones: while more than one half of the galaxies observed by Hubble Space Telescope and Very Large Telescope ranging from rest-frame UV to optical bands with stellar masses larger than 2 × 1010 M ☉ have half-light radii larger than ~2h –1 73 kpc, none of the simulated massive galaxies in the same mass range have half-light radii larger than ~2h –1 73 kpc, after taking into account dust extinction. Corroborative evidence is provided by the rotation curves of the simulated galaxies with total masses of 1011-1012 M ☉, which display values (300-1000 km s–1) at small radii (~0.5h –1 73 kpc) due to high stellar concentration in the central regions that are larger than those of any well observed galaxies. Possible physical mechanisms to resolve this serious problem include: (1) an early reionization at zri 6 to suppress gas condensation and hence star formation, (2) a strong, internal energetic feedback from stars or central black holes to reduce the overall star formation efficiency, or (3) a substantial small-scale cutoff in the matter power spectrum.

Chondrule Formation and Protoplanetary Disk Heating by Current Sheets in Nonideal Magnetohydrodynamic Turbulence

We study magnetic field steepening due to ambipolar diffusion in protoplanetary disk environments and draw the following conclusions. Current sheets are generated in magnetically active regions of the disk where the ionization fraction is high enough for the magnetorotational instability to operate. In late stages of solar nebula evolution, the surface density is expected to decrease and dust grains are expected to gravitationally settle to the midplane. If the local dust-to-gas mass ratio near the midplane is increased above cosmic abundances by factors 103, current sheets reach high enough temperatures to melt millimeter-sized dust grains and hence may provide the mechanism to form meteoritic chondrules. In addition, these current sheets possibly explain the near-infrared excesses observed in spectral energy distributions (SEDs) of young stellar objects. Direct imaging of protoplanetary disks via a nulling interferometer or, in the future, a multiband, adaptive optics coronagraph can test this hypothesis.

PHOTOIONIZATION OF HIGH-ALTITUDE GAS IN A SUPERNOVA-DRIVEN TURBULENT INTERSTELLAR MEDIUM

We investigate models for the photoionization of the widespread diffuse ionized gas (DIG) in galaxies. In particular, we address the long standing question of the penetration of Lyman continuum photons from sources close to the galactic midplane to large heights in the galactic halo. We find that recent hydrodynamical simulations of a supernova-driven interstellar medium (ISM) have low-density paths and voids that allow for ionizing photons from midplane OB stars to reach and ionize gas many kiloparsecs above the midplane. We find that ionizing fluxes throughout our simulation grids are larger than predicted by one-dimensional slab models, thus allowing for photoionization by O stars of low altitude neutral clouds in the Galaxy that are also detected in Hα. In previous studies of such clouds, the photoionization scenario had been rejected and the Hα had been attributed to enhanced cosmic ray ionization or scattered light from midplane H II regions. We do find that the emission measure distributions in our simulations are wider than those derived from Hα observations in the Milky Way. In addition, the horizontally averaged height dependence of the gas density in the hydrodynamical models is lower than inferred in the Galaxy. These discrepancies are likely due to the absence of magnetic fields in the hydrodynamic simulations and we discuss how magnetohydrodynamic effects may reconcile models and observations. Nevertheless, we anticipate that the inclusion of magnetic fields in the dynamical simulations will not alter our primary finding that midplane OB stars are capable of producing high-altitude DIG in a realistic three-dimensional ISM.

The Origin of the Hot Gas in the Galactic Halo: Confronting Models with XMM-Newton Observations

We compare the predictions of three physical models for the origin of the hot halo gas with the observed halo X-ray emission, derived from 26 high-latitude XMM-Newton observations of the soft X-ray background between l = 120° and l = 240°. These observations were chosen from a much larger set of observations as they are expected to be the least contaminated by solar wind charge exchange emission. We characterize the halo emission in the XMM-Newton band with a single-temperature plasma model. We find that the observed halo temperature is fairly constant across the sky (~(1.8-2.4) × 106 K), whereas the halo emission measure varies by an order of magnitude (~0.0005-0.006 cm–6 pc). When we compare our observations with the model predictions, we find that most of the hot gas observed with XMM-Newton does not reside in isolated extraplanar supernova (SN) remnants—this model predicts emission an order of magnitude too faint. A model of an SN-driven interstellar medium, including the flow of hot gas from the disk into the halo in a galactic fountain, gives good agreement with the observed 0.4-2.0 keV surface brightness. This model overpredicts the halo X-ray temperature by a factor of ~2, but there are a several possible explanations for this discrepancy. We therefore conclude that a major (possibly dominant) contributor to the halo X-ray emission observed with XMM-Newton is a fountain of hot gas driven into the halo by disk SNe. However, we cannot rule out the possibility that the extended hot halo of accreted material predicted by disk galaxy formation models also contributes to the emission.

Gas Accretion is Dominated by Warm Ionized Gas in Milky Way Mass Galaxies at z ~ 0

We perform high-resolution hydrodynamic simulations of a Milky Way mass galaxy in a fully cosmological setting using the adaptive mesh refinement code, Enzo, and study the kinematics of gas in the simulated galactic halo. We find that the gas inflow occurs mostly along filamentary structures in the halo. The warm-hot (105 K 106 K) ionized gases are found to dominate the overall mass accretion in the system (with -5 M ? yr?1) over a large range of distances, extending from the virial radius to the vicinity of the disk. Most of the inflowing gas (by mass) does not cool, and the small fraction that manages to cool does so primarily close to the galaxy (R 100?kpc, with more pronounced cooling at smaller R), perhaps comprising the neutral gas that may be detectable as, e.g., high-velocity clouds. The neutral clouds are embedded within larger, accreting filamentary flows, and represent only a small fraction of the total mass inflow rate. The inflowing gas has relatively low metallicity (Z/Z ? < 0.2). The outer layers of the filamentary inflows are heated due to compression as they approach the disk. In addition to the inflow, we find high-velocity, metal-enriched outflows of hot gas driven by supernova feedback. Our results are consistent with observations of halo gas at low z.

THE ORIGIN AND DISTRIBUTION OF COLD GAS IN THE HALO OF A MILKY-WAY-MASS GALAXY

We analyze an adaptive mesh refinement hydrodynamic cosmological simulation of a Milky Way-sized galaxy to study the cold gas in the halo. HI observations of the Milky Way and other nearby spirals have revealed the presence of such gas in the form of clouds and other extended structures, which indicates on-going accretion. We use a high-resolution simulation (136-272 pc throughout) to study the distribution of cold gas in the halo, compare it with observations, and examine its origin. The amount (10^8 Msun in HI), covering fraction, and spatial distribution of the cold halo gas around the simulated galaxy at z=0 are consistent with existing observations. At z=0 the HI mass accretion rate onto the disk is 0.2 Msun/yr. We track the histories of the 20 satellites that are detected in HI in the redshift interval 0.5>z>0 and find that most of them are losing gas, with a median mass loss rate per satellite of 3.1 x 10^{-3} Msun/yr. This stripped gas is a significant component of the HI gas seen in the simulation. In addition, we see filamentary material coming into the halo from the IGM at all redshifts. Most of this gas does not make it directly to the disk, but part of the gas in these structures is able to cool and form clouds. The metallicity of the gas allows us to distinguish between filamentary flows and satellite gas. We find that the former accounts for at least 25-75% of the cold gas in the halo seen at any redshift analyzed here. Placing constraints on cloud formation mechanisms allows us to better understand how galaxies accrete gas and fuel star formation at z=0.

SIMULATED VOID GALAXIES IN THE STANDARD COLD DARK MATTER MODEL

We analyze a (120 h –1 Mpc)3 adaptive mesh refinement hydrodynamic simulation that contains a higher resolution 31 × 31 × 35 h –3 Mpc subvolume centered on a ~30 Mpc diameter void. Our detailed ~1 kpc resolution allows us to identify 1300 galaxies within this void to a limiting halo mass of ~1010 M ☉. Nearly 1000 galaxies are found to be in underdense regions, with 300 galaxies residing in regions less than half the mean density of the simulation volume. We construct mock observations of the stellar and gas properties of these systems and reproduce the range of colors and luminosities observed in the Sloan Digital Sky Survey for nearby (z –16), though they are less reliably resolved, typically appear bluer, with higher rates of star formation and specific star formation and lower mean stellar ages than galaxies in average density environments. We find a significant population of low-luminosity (M r ~ –14) dwarf galaxies that is preferentially located in low-density regions and specifically in the void center. This population may help to reduce, but not remove, the discrepancy between the predicted and observed number of void galaxies.