Search NASA⌕ Search

Engineering topics

Krumholz, Mark R

Publications and source records attributed to Krumholz, Mark R.

At least 19 records

DIRECT NUMERICAL SIMULATION OF RADIATION PRESSURE-DRIVEN TURBULENCE AND WINDS IN STAR CLUSTERS AND GALACTIC DISKS

The pressure exerted by the radiation of young stars may be an important feedback mechanism that drives turbulence and winds in forming star clusters and the disks of starburst galaxies. However, there is great uncertainty in how efficiently radiation couples to matter in these high optical depth environments. In particular, it is unclear what levels of turbulence the radiation can produce, and whether the infrared radiation trapped by the dust opacity can give rise to heavily mass-loaded winds. In this paper, we report a series of two-dimensional flux-limited diffusion radiation-hydrodynamics calculations performed with the code ORION in which we drive strong radiation fluxes through columns of dusty matter confined by gravity in order to answer these questions. We consider both systems where the radiation flux is sub-Eddington throughout the gas column, and those where it is super-Eddington at the midplane but sub-Eddington in the atmosphere. In the latter, we find that the radiation-matter interaction gives rise to radiation-driven Rayleigh-Taylor instability, which drives supersonic turbulence at a level sufficient to fully explain the turbulence seen in Galactic protocluster gas clouds, and to make a non-trivial contribution to the turbulence observed in starburst galaxy disks. However, the instability also produces a channel structure in which the radiation-matter interaction is reduced compared to time-steady analytic models because the radiation field is not fully trapped. For astrophysical parameters relevant to forming star clusters and starburst galaxies, we find that this effect reduces the net momentum deposition rate in the dusty gas by a factor of {approx}2-6 compared to simple analytic estimates, and that in steady state the Eddington ratio reaches unity and there are no strong winds. We provide an approximation formula, appropriate for implementation in analytic models and non-radiative simulations, for the force exerted by the infrared radiation field in this regime.

79 ASTRONOMY AND ASTROPHYSICS↗

WAS THE SUN BORN IN A MASSIVE CLUSTER?

A number of authors have argued that the Sun must have been born in a cluster of no more than several thousand stars, on the basis that, in a larger cluster, close encounters between the Sun and other stars would have truncated the outer solar system or excited the outer planets into eccentric orbits. However, this dynamical limit is in tension with meteoritic evidence that the solar system was exposed to a nearby supernova during or shortly after its formation; a several-thousand-star cluster is much too small to produce a massive star whose lifetime is short enough to have provided the enrichment. In this paper, we revisit the dynamical limit in the light of improved observations of the properties of young clusters. We use a series of scattering simulations to measure the velocity-dependent cross-section for disruption of the outer solar system by stellar encounters, and use this cross-section to compute the probability of a disruptive encounter as a function of birth cluster properties. We find that, contrary to prior work, the probability of disruption is small regardless of the cluster mass, and that it actually decreases rather than increases with cluster mass. Our results differ from prior work for three main reasons: (1) unlike in most previous work, we compute a velocity-dependent cross-section and properly integrate over the cluster mass-dependent velocity distribution of incoming stars; (2) we recognize that {approx}90% of clusters have lifetimes of a few crossing times, rather than the 10-100 Myr adopted in many earlier models; and (3) following recent observations, we adopt a mass-independent surface density for embedded clusters, rather than a mass-independent radius as assumed many earlier papers. Our results remove the tension between the dynamical limit and the meteoritic evidence, and suggest that the Sun was born in a massive cluster. A corollary to this result is that close encounters in the Sun's birth cluster are highly unlikely to truncate the Kuiper Belt unless the Sun was born in one of the unusual clusters that survived for tens of Myr. However, we find that encounters could plausibly produce highly eccentric Kuiper Belt objects such as Sedna.

79 ASTRONOMY AND ASTROPHYSICS↗

RADIATION-HYDRODYNAMIC SIMULATIONS OF THE FORMATION OF ORION-LIKE STAR CLUSTERS. II. THE INITIAL MASS FUNCTION FROM WINDS, TURBULENCE, AND RADIATION

We report a series of simulations of the formation of a star cluster similar to the Orion Nebula Cluster (ONC), including both radiative transfer and protostellar outflows, and starting from both smooth and self-consistently turbulent initial conditions. Each simulation forms >150 stars and brown dwarfs, yielding a stellar mass distribution that ranges from <0.1 M{sub Sun} to >10 M{sub Sun }. We show that a simulation that begins with self-consistently turbulent density and velocity fields embedded in a larger turbulent volume, and that includes protostellar outflows, produces an initial mass function (IMF) that is consistent both with that of the ONC and the Galactic field, at least within the statistical power provided by the number of stars formed in our simulations. This is the first simulation published to date that reproduces the observed IMF in a cluster large enough to contain massive stars, and where the peak of the mass function is determined by a fully self-consistent calculation of gas thermodynamics rather than a hand-imposed equation of state. This simulation also produces a star formation rate that, while still somewhat too high, is much closer to observed values than if we omit either the larger turbulent volume or the outflows. Moreover, we show that the combination of outflows, self-consistently turbulent initial conditions, and turbulence continually fed by motions on scales larger than that of the protocluster yields an IMF that is in agreement with observations and invariant with time, resolving the 'overheating' problem in which simulations without these features have an IMF peak that shifts to progressively higher masses over time as more and more of the gas is heated, inconsistent with the observed invariance of the IMF. The simulation that matches the observed IMF also qualitatively reproduces the observed trend of stellar multiplicity strongly increasing with mass. We show that this simulation produces massive stars from distinct massive cores whose properties are consistent with those of observed massive cores. However, the stars formed in these cores also undergo dynamical interactions as they accrete that naturally produce Trapezium-like hierarchical multiple systems of massive stars.

79 ASTRONOMY AND ASTROPHYSICS↗

METALLICITY-DEPENDENT QUENCHING OF STAR FORMATION AT HIGH REDSHIFT IN SMALL GALAXIES

The star formation rates (SFRs) of low-metallicity galaxies depend sensitively on the gas metallicity, because metals are crucial to mediating the transition from intermediate-temperature atomic gas to cold molecular gas, a necessary precursor to star formation. We study the impact of this effect on the star formation history of galaxies. We incorporate metallicity-dependent star formation and metal enrichment in a simple model that follows the evolution of a halo main progenitor. Our model shows that including the effect of metallicity leads to suppression of star formation at redshift z > 2 in dark halos with masses {approx}< 10{sup 11} M{sub Sun }, with the suppression becoming near total for halos below {approx}10{sup 9.5}-10{sup 10} M{sub Sun }. We find that at high redshift, until z {approx} 2, the SFR cannot catch up with the gas inflow rate (IR), because the SFR is limited by the free-fall time, and because it is suppressed further by a lack of metals in small halos. As a result, in each galaxy the SFR is growing in time faster than the IR, and the integrated cosmic SFR density is rising with time. The suppressed in situ SFR at high-z makes the growth of stellar mass dominated by ex situ SFR, meaning stars formed in lower mass progenitor galaxies and then accreted, which implies that the specific SFR (sSFR) remains constant with time. The intensely accreted gas at high-z is accumulating as an atomic gas reservoir. This provides additional fuel for star formation in 10{sup 10}-10{sup 12} M{sub Sun} halos at z {approx} 1-3, which allows the SFR to exceed the instantaneous IR, and may enable an even higher outflow rate. At z < 1, following the natural decline in IR with time due to the universal expansion, the SFR and sSFR are expected to drop. We specify the expected dependence of sSFR and metallicity on stellar mass and redshift. At a given z, and below a critical mass, these relations are predicted to be flat and rising, respectively. Our model predictions qualitatively match some of the puzzling features in the observed star formation history.

79 ASTRONOMY AND ASTROPHYSICS↗

DWARF GALAXY FORMATION WITH H{sub 2}-REGULATED STAR FORMATION

We describe cosmological galaxy formation simulations with the adaptive mesh refinement code Enzo that incorporate a star formation prescription regulated by the local abundance of molecular hydrogen. We show that this H{sub 2}-regulated prescription leads to a suppression of star formation in low-mass halos (M{sub h} {approx}< 10{sup 10} M{sub Sun }) at z > 4, alleviating some of the dwarf galaxy problems faced by theoretical galaxy formation models. H{sub 2} regulation modifies the efficiency of star formation of cold gas directly, rather than indirectly reducing the cold gas content with 'supernova feedback'. We determine the local H{sub 2} abundance in our most refined grid cells (76 proper parsec in size at z = 4) by applying the model of Krumholz, McKee, and Tumlinson, which is based on idealized one-dimensional radiative transfer calculations of H{sub 2} formation-dissociation balance in {approx}100 pc atomic-molecular complexes. Our H{sub 2}-regulated simulations are able to reproduce the empirical (albeit lower z) Kennicutt-Schmidt relation, including the low {Sigma}{sub gas} cutoff due to the transition from atomic to molecular phase and the metallicity dependence thereof, without the use of an explicit density threshold in our star formation prescription. We compare the evolution of the luminosity function, stellar mass density, and star formation rate density from our simulations to recent observational determinations of the same at z = 4-8 and find reasonable agreement between the two.

79 ASTRONOMY AND ASTROPHYSICS↗

WHAT SETS THE INITIAL ROTATION RATES OF MASSIVE STARS?

The physical mechanisms that set the initial rotation rates in massive stars are a crucial unknown in current star formation theory. Observations of young, massive stars provide evidence that they form in a similar fashion to their low-mass counterparts. The magnetic coupling between a star and its accretion disk may be sufficient to spin down low-mass pre-main-sequence (PMS) stars to well below breakup at the end stage of their formation when the accretion rate is low. However, we show that these magnetic torques are insufficient to spin down massive PMS stars due to their short formation times and high accretion rates. We develop a model for the angular momentum evolution of stars over a wide range in mass, considering both magnetic and gravitational torques. We find that magnetic torques are unable to spin down either low-mass or high-mass stars during the main accretion phase, and that massive stars cannot be spun down significantly by magnetic torques during the end stage of their formation either. Spin-down occurs only if massive stars' disk lifetimes are substantially longer or their magnetic fields are much stronger than current observations suggest.

79 ASTRONOMY AND ASTROPHYSICS↗

EVOLUTION OF BLISTER-TYPE H II REGIONS IN A MAGNETIZED MEDIUM

We use the three-dimensional Athena ionizing radiation-magnetohydrodynamics code to simulate blister-type H II regions driven by stars on the edge of magnetized gas clouds. We compare these to simulations of spherical H II regions where the star is embedded deep within a cloud, and to non-magnetized simulations of both types, in order to compare their ability to drive turbulence and influence star formation. We find that magnetized blister H II regions can be very efficient at injecting energy into clouds. This is partly a magnetic effect: the magnetic energy added to a cloud by an H II region is comparable to or larger than the kinetic energy, and magnetic fields can also help collimate the ejected gas, increasing its energy yield. As a result of these effects, a blister H II region expanding into a cloud with a magnetic field perpendicular to its edge injects twice as much energy by 5 Myr as a non-magnetized blister H II region driven by a star of the same luminosity. Blister H II regions are also more efficient at injecting kinetic energy than spherical H II regions, due to the recoil provided by escaping gas, but not as much as predicted by some analytic approximations.

79 ASTRONOMY AND ASTROPHYSICS↗

A UNIVERSAL, LOCAL STAR FORMATION LAW IN GALACTIC CLOUDS, NEARBY GALAXIES, HIGH-REDSHIFT DISKS, AND STARBURSTS

Star formation laws are rules that relate the rate of star formation in a particular region, either an entire galaxy or some portion of it, to the properties of the gas, or other galactic properties, in that region. While observations of Local Group galaxies show a very simple, local star formation law in which the star formation rate per unit area in each patch of a galaxy scales linearly with the molecular gas surface density in that patch, recent observations of both Milky Way molecular clouds and high-redshift galaxies apparently show a more complicated relationship in which regions of equal molecular gas surface density can form stars at quite different rates. These data have been interpreted as implying either that different star formation laws may apply in different circumstances, that the star formation law is sensitive to large-scale galaxy properties rather than local properties, or that there are high-density thresholds for star formation. Here we collate observations of the relationship between gas and star formation rate from resolved observations of Milky Way molecular clouds, from kpc-scale observations of Local Group galaxies, and from unresolved observations of both disk and starburst galaxies in the local universe and at high redshift. We show that all of these data are in fact consistent with a simple, local, volumetric star formation law. The apparent variations stem from the fact that the observed objects have a wide variety of three-dimensional size scales and degrees of internal clumping, so even at fixed gas column density the regions being observed can have wildly varying volume densities. We provide a simple theoretical framework to remove this projection effect, and we use it to show that all the data, from small solar neighborhood clouds with masses {approx}10{sup 3} M{sub Sun} to submillimeter galaxies with masses {approx}10{sup 11} M{sub Sun }, fall on a single star formation law in which the star formation rate is simply {approx}1% of the molecular gas mass per local free-fall time. In contrast, proposed star formation laws in which the star formation timescale is set by the galactic rotation period are inconsistent with the data from the Milky Way and the Local Group, while those in which the star formation rate is linearly proportional to the gas mass above some density threshold fail both in the Local Group and for starburst galaxies.

79 ASTRONOMY AND ASTROPHYSICS↗

RADIATIVELY EFFICIENT MAGNETIZED BONDI ACCRETION

We have carried out a numerical study of the effect of large-scale magnetic fields on the rate of accretion from a uniform, isothermal gas onto a resistive, stationary point mass. Only mass, not magnetic flux, accretes onto the point mass. The simulations for this study avoid complications arising from boundary conditions by keeping the boundaries far from the accreting object. Our simulations leverage adaptive refinement methodology to attain high spatial fidelity close to the accreting object. Our results are particularly relevant to the problem of star formation from a magnetized molecular cloud in which thermal energy is radiated away on timescales much shorter than the dynamical timescale. Contrary to the adiabatic case, our simulations show convergence toward a finite accretion rate in the limit in which the radius of the accreting object vanishes, regardless of magnetic field strength. For very weak magnetic fields, the accretion rate first approaches the Bondi value and then drops by a factor of {approx}2 as magnetic flux builds up near the point mass. For strong magnetic fields, the steady-state accretion rate is reduced by a factor of {approx}0.2 {beta}{sup 1/2} compared to the Bondi value, where {beta} is the ratio of the gas pressure to the magnetic pressure. We give a simple expression for the accretion rate as a function of the magnetic field strength. Approximate analytic results are given in the Appendices for both time-dependent accretion in the limit of weak magnetic fields and steady-state accretion for the case of strong magnetic fields.

79 ASTRONOMY AND ASTROPHYSICS↗

ON THE ORIGIN OF STELLAR MASSES

It has been a longstanding problem to determine, as far as possible, the characteristic masses of stars in terms of fundamental constants; the almost complete invariance of this mass as a function of the star-forming environment suggests that this should be possible. Here I provide such a calculation. The typical stellar mass is set by the characteristic fragment mass in a star-forming cloud, which depends on the cloud's density and temperature structure. Except in the very early universe, the latter is determined mainly by the radiation released as matter falls onto seed protostars. The energy yield from this process is ultimately set by the properties of deuterium burning in protostellar cores, which determines the stars' radii. I show that it is possible to combine these considerations to compute a characteristic stellar mass almost entirely in terms of fundamental constants, with an extremely weak residual dependence on the interstellar pressure and metallicity. This result not only explains the invariance of stellar masses, it resolves a second mystery: why fragmentation of a cold, low-density interstellar cloud, a process with no obvious dependence on the properties of nuclear reactions, happens to select a stellar mass scale such that stellar cores can ignite hydrogen. Finally, the weak residual dependence on the interstellar pressure and metallicity may explain recent observational hints of a smaller characteristic mass in the high-pressure, high-metallicity cores of giant elliptical galaxies.

79 ASTRONOMY AND ASTROPHYSICS↗

STOCHASTIC STAR FORMATION AND A (NEARLY) UNIFORM STELLAR INITIAL MASS FUNCTION

Recent observations indicate a lower H{alpha} to FUV ratio in dwarf galaxies than in brighter systems, a trend that could be explained by a truncated and/or steeper initial mass function (IMF) in small galaxies. However, at low star formation rates (SFRs), the H{alpha} to FUV ratio can vary due to stochastic sampling even for a universal IMF, a hypothesis that has, prior to this work, received limited investigation. Using SLUG, a fully stochastic code for synthetic photometry in star clusters and galaxies, we compare the H{alpha} and FUV luminosity in a sample of {approx}450 nearby galaxies with models drawn from a universal Kroupa IMF and a modified IMF, the integrated galactic initial mass function (IGIMF). Once random sampling and time evolution are included, a Kroupa IMF convolved with the cluster mass function (CMF) reproduces the observed H{alpha} distribution at all FUV luminosities, while a truncated IMF as implemented in current IGIMF models underpredicts the H{alpha} luminosity by more than an order of magnitude at the lowest SFRs. We conclude that the observed luminosity is the result of the joint probability distribution function of the SFR, CMF, and a universal IMF, consistent with parts of the IGIMF theory, but that a truncation in the IMF in clusters is inconsistent with the observations. Future work will examine stochastic star formation and its time dependence in detail to study whether random sampling can explain other observations that suggest a varying IMF.

79 ASTRONOMY AND ASTROPHYSICS↗

RADIATION-HYDRODYNAMIC SIMULATIONS OF THE FORMATION OF ORION-LIKE STAR CLUSTERS. I. IMPLICATIONS FOR THE ORIGIN OF THE INITIAL MASS FUNCTION

One model for the origin of typical Galactic star clusters such as the Orion Nebula Cluster (ONC) is that they form via the rapid, efficient collapse of a bound gas clump within a larger, gravitationally unbound giant molecular cloud. However, simulations in support of this scenario have thus far not included the radiation feedback produced by the stars; radiative simulations have been limited to significantly smaller or lower-density regions. Here we use the ORION AMR code to conduct the first ever radiation-hydrodynamic simulations of the global collapse scenario for the formation of an ONC-like cluster. We show that radiative feedback has a dramatic effect on the evolution: once the first {approx}10%-20% of the gas mass is incorporated into stars, their radiative feedback raises the gas temperature high enough to suppress any further fragmentation. However, gas continues to accrete onto existing stars, and, as a result, the stellar mass distribution becomes increasingly top-heavy, eventually rendering it incompatible with the observed initial mass function (IMF). Systematic variation in the location of the IMF peak as star formation proceeds is incompatible with the observed invariance of the IMF between star clusters, unless some unknown mechanism synchronizes the IMFs in different clusters by ensuring that star formation is always truncated when the IMF peak reaches a particular value. We therefore conclude that the global collapse scenario, at least in its simplest form, is not compatible with the observed stellar IMF. We speculate that processes that slow down star formation, and thus reduce the accretion luminosity, may be able to resolve the problem.

79 ASTRONOMY AND ASTROPHYSICS↗

THE GLOBAL EVOLUTION OF GIANT MOLECULAR CLOUDS. II. THE ROLE OF ACCRETION

We present virial models for the global evolution of giant molecular clouds (GMCs). Focusing on the presence of an accretion flow and accounting for the amount of mass, momentum, and energy supplied by accretion and star formation feedback, we are able to follow the growth, evolution, and dispersal of individual GMCs. Our model clouds reproduce the scaling relations observed in both galactic and extragalactic clouds. We find that accretion and star formation contribute roughly equal amounts of turbulent kinetic energy over the lifetime of the cloud. Clouds attain virial equilibrium and grow in such a way as to maintain roughly constant surface densities, with typical surface densities of order 50-200 M{sub sun} pc{sup -2}, in good agreement with observations of GMCs in the Milky Way and nearby external galaxies. We find that as clouds grow, their velocity dispersion and radius must also increase, implying that the linewidth-size relation constitutes an age sequence. Lastly, we compare our models to observations of GMCs and associated young star clusters in the Large Magellanic Cloud and find good agreement between our model clouds and the observed relationship between H II regions, young star clusters, and GMCs.

79 ASTRONOMY AND ASTROPHYSICS↗

METALLICITY AND THE UNIVERSALITY OF THE INITIAL MASS FUNCTION

The stellar initial mass function (IMF), along with the star formation rate, is one of the fundamental properties that any theory of star formation must explain. An interesting feature of the IMF is that it appears to be remarkably universal across a wide range of environments. Particularly, there appears to be little variation in either the characteristic mass of the IMF or its high-mass tail between clusters with different metallicities. Previous attempts to understand this apparent independence of metallicity have not accounted for radiation feedback from high-mass protostars, which can dominate the energy balance of the gas in star-forming regions. We extend this work, showing that the fragmentation of molecular gas should depend only weakly on the amount of dust present, even when the primary heating source is radiation from massive protostars. First, we report a series of core collapse simulations using the ORION AMR code that systematically vary the dust opacity and show explicitly that this has little effect on the temperature or fragmentation of the gas. Then, we provide an analytic argument for why the IMF varies so little in observed star clusters, even as the metallicity varies by a factor of 100.

79 ASTRONOMY AND ASTROPHYSICS↗

WHAT DRIVES THE EXPANSION OF GIANT H II REGIONS?: A STUDY OF STELLAR FEEDBACK IN 30 DORADUS

Observations show that star formation is an inefficient and slow process. This result can be attributed to the injection of energy and momentum by stars that prevents free-fall collapse of molecular clouds. The mechanism of this stellar feedback is debated theoretically; possible sources of pressure include the classical warm H II gas, the hot gas generated by shock heating from stellar winds and supernovae, direct radiation of stars, and the dust-processed radiation field trapped inside the H II shell. In this paper, we measure observationally the pressures associated with each component listed above across the giant H II region 30 Doradus in the Large Magellanic Cloud. We exploit high-resolution, multi-wavelength images (radio, infrared, optical, ultraviolet, and X-ray) to map these pressures as a function of position. We find that radiation pressure dominates within 75 pc of the central star cluster, R136, while the H II gas pressure dominates at larger radii. By contrast, the dust-processed radiation pressure and hot gas pressure are generally weak and not dynamically important, although the hot gas pressure may have played a more significant role at early times. Based on the low X-ray gas pressures, we demonstrate that the hot gas is only partially confined and must be leaking out the H II shell. Additionally, we consider the implications of a dominant radiation pressure on the early dynamics of 30 Doradus.

79 ASTRONOMY AND ASTROPHYSICS↗

WHICH PHASE OF THE INTERSTELLAR MEDIUM CORRELATES WITH THE STAR FORMATION RATE?

Nearby spiral galaxies show an extremely tight correlation between tracers of molecular hydrogen (H{sub 2}) in the interstellar medium and tracers of recent star formation, but it is unclear whether this correlation is fundamental or accidental. In the galaxies that have been surveyed to date, H{sub 2} resides predominantly in gravitationally bound clouds cooled by carbon monoxide (CO) molecules, but in galaxies of low metal content the correlations between bound clouds, CO, and H{sub 2} break down, and it is unclear if the star formation rate (SFR) will then correlate with H{sub 2} or with some other quantity. Here, we show that star formation will continue to follow H{sub 2} independent of metallicity. This is not because H{sub 2} is directly important for cooling, but instead because the transition from predominantly atomic hydrogen (H I) to H{sub 2} occurs under the same conditions as a dramatic drop in gas temperature and Bonnor-Ebert mass that destabilizes clouds and initiates collapse. We use this model to compute how SFR will correlate with total gas mass, with mass of gas where the hydrogen is H{sub 2}, and with mass of gas where the carbon is CO in galaxies of varying metallicity, and show that preliminary observations match the trend we predict.

79 ASTRONOMY AND ASTROPHYSICS↗

THE FORMATION OF LOW-MASS BINARY STAR SYSTEMS VIA TURBULENT FRAGMENTATION

We characterize the infall rate onto protostellar systems forming in self-gravitating radiation-hydrodynamics simulations. Using two dimensionless parameters to determine the disks' susceptibility to gravitational fragmentation, we infer limits on protostellar system multiplicity and the mechanism of binary formation. We show that these parameters give robust predictions even in the case of marginally resolved protostellar disks. We find that protostellar systems with radiation feedback predominately form binaries via turbulent fragmentation, not disk instability, and predict that turbulent fragmentation is the dominant channel for binary formation for low-mass stars. We clearly demonstrate that systems forming in simulations including radiative feedback have fundamentally different parameters than those in purely hydrodynamics simulations.

79 ASTRONOMY AND ASTROPHYSICS↗

TESTING MODELS FOR MOLECULAR GAS FORMATION IN GALAXIES: HYDROSTATIC PRESSURE OR GAS AND DUST SHIELDING?

Stars in galaxies form in giant molecular clouds that coalesce when the atomic hydrogen is converted into molecules. There are currently two dominant models based on the property of the galactic disk that determines its molecular fraction: either hydrostatic pressure driven by the gravity of gas and stars, or a combination of gas column density and metallicity. To assess the validity of these models, we compare theoretical predictions to the observed atomic gas content of low-metallicity dwarf galaxies with high stellar densities. The extreme conditions found in these systems are optimal for distinguishing the two models, otherwise degenerate in nearby spirals. Locally, on scales <100 pc, we find that the state of the interstellar medium is mostly sensitive to the gas column density and metallicity rather than hydrostatic pressure. On larger scales where the average stellar density is considerably lower, both pressure and shielding models reproduce the observations, even at low metallicity. We conclude that models based on gas and dust shielding more closely describe the process of molecular formation, especially at the high resolution that can be achieved in modern galaxy simulations or with future radio/millimeter arrays.

79 ASTRONOMY AND ASTROPHYSICS↗