From Local Star Formation to Global Star Formation

In this brief report we summarise the most important points raised in the course of a two-hour evening discussion session on the above topic, organised by the author. Major questions that were debated included the universality of the IMF, the history of the star formation rate in the solar neighorhood, the star formation efficiency in molecular clouds, and the role of triggered star formation. The issue of a threshold gas surface density for star formation to occur was also critically discussed. This revised version was published online in September 2006 with corrections to the Cover Date.     

Turbulent Fragmentation and Star Formation

We review the main results from recent numerical simulations of turbulent fragmentation and star formation. Specifically, we discuss the observed scaling relationships, the “quiescent” (subsonic) nature of many star-forming cores, their energy balance, their synthesized polarized dust emission, the ages of stars associated with the molecular gas from which they have formed, the mass spectra of clumps, and the density and column density probability distribution function of the gas. We then give a critical discussion on recent attempts to explain and/or predict the star formation efficiency and the stellar initial mass function from the statistical nature of turbulent fields. Finally, it appears that turbulent fragmentation alone cannot account for the final stages of fragmentation: although the turbulent velocity field is able to produce filaments, the spatial distribution of cores in such filaments is better explained in terms of gravitational fragmentation.     

Star Formation and Cosmological Simulations

The inclusion of a detailed modeling of the short-scale baryonic physics in a large-scale cosmological simulation is crucial for a better comparison between observations and predictions from cosmological models. From a set of 3D hydrodynamical simulations which include a chemical model to account for the complex physics of the ISM at a sub-grid scale, we have been able to obtain a statistically significant sample of galaxy-type halos with observational properties, like colors and luminosities for different cosmological scenarios. From this data base, we have studied a number of different things, like Tully-Fisher relations, luminosity functions and environmental effects. Despite the progress made during the last few years in the modeling of the physics of ISM and star formation, more work is clearly needed. This revised version was published online in July 2006 with corrections to the Cover Date.     

What Drives Star Formation?

Current theoretical models for what drives star formation (especially low-mass star formation) are: (1) magnetic support of self-gravitating clouds with ambipolar diffusion removing support in cores and triggering collapse and (2) compressible turbulence forming self-gravitating clumps that collapse as soon as the turbulent cascade produces insufficient turbulent support. Observations of magnetic fields can distinguish between these two models because of different predictions in three areas: (1) magnetic field morphology, (2) the scaling of field strength with density and non-thermal velocities, and (3) the mass to magnetic flux ratio, M/Φ. We first discuss the techniques and limitations of methods for observing magnetic fields in star formation regions, then describe results for the L1544 prestellar core as an exemplar of the observational results. Application of the three tests leads to the following conclusions. The observational data show that both magnetic fields and turbulence are important in molecular cloud physics. Field lines are generally regular rather than chaotic, implying strong field strengths. But fields are not aligned with the minor axes of oblate spheroidal clouds, suggesting that turbulence is important. Field strengths appear to scale with non-thermal velocity widths, suggesting a significant turbulent support of clouds. Giant Molecular Clouds (GMCs) require mass accumulation over sufficiently large volumes that they would likely have an approximately critical M/Φ. Yet H I clouds are observed to be highly subcritical. If self-gravitating (molecular) clouds form with the subcritical M/Φ of H I clouds, the molecular clouds will be subcritical. However, the observations of molecular cloud cores suggest that they are approximately critical, with no direct evidence for subcritical molecular clouds or cloud envelopes. Hence, the observations remain inconclusive in deciding between the two extreme-case models of what drives star formation. What is needed to further advance our understanding of the role of magnetic fields in the star formation process are additional high sensitivity surveys of magnetic field strengths and other cloud properties in order to further refine the assessment of the importance of magnetic fields in molecular cores and envelopes.     

旋涡星系中恒星形成对旋涡结构的影响

本文利用N体数值模拟方法,讨论了旋涡星系中恒星形成对旋涡结构的维持作用。在假定恒星是在一定密度区域中形成以及恒星诞生率为1.0M_⊙/年的情况下,模拟结果表明,旋涡图象的维持时间是未考虑恒星形成模型的2.3倍。     

Circumnuclear Star Formation in M100

Near-infrared (JHK) images of the nucleus of the barred spiral galaxy M100 (NGC 4321) have been obtained using a high-bandwidth tip-tilt secondary and fast guider system on the 3.8 m United Kingdom Infrared Telescope (UKIRT). The resulting images, with a resolution of 0.34″ at K, reveal for the first time a host of compact ‘knots’, which appear to be the result of a recent burst (or bursts) of star formation. Confirmation of this comes from K-band spectroscopy of these knots with CGS4 on UKIRT, which shows Brγ emission and CO absorption features. A comparison with starburst evolutionary models suggests ages for these knots of between 17 and 27 Myr, and a stellar population dominated by late-type supergiants. This revised version was published online in July 2006 with corrections to the Cover Date.     

Star Formation and the Hall Effect

The breakdown of flux-freezing in molecular clouds and protostellar discs is usually approximated by ambipolar diffusion at low densities or by resistive diffusion at high densities. Here an intermediate regime is discussed in which the Hall term in the conductivity tensor is significant, and the vector evolution of the magnetic field, and therefore the evolution of the system under consideration is dramatically altered. Calculations of charged particle abundances in dense gas in molecular clouds and protostellar discs demonstrate that Hall diffusion is important over a surprisingly broad range of conditions.     

On a Formation Scenario of Star Clusters

Most formation scenarios of globular clusters assume a molecular cloud as the progenitor of the stellar system. However, it is still unclear, how this cloud is transformed into a star cluster, i.e. how the destructive processes related to gas removal or low star formation effiency can be avoided. Here a scheme of supernova (SN) induced cluster formation is studied. According to this scenario an expanding SN shell accumulates the mass of the cloud. This is accompanied by fragmentation resulting in star formation in the shell. Provided the stellar shell expands sufficiently slow, its self-gravity stops the expansion and the shell recollapses, by this forming a stellar system. I present N-body simulations of collapsing shells which move in a galactic potential on circular and elliptic orbits. It is shown that typical shells (10⁵ M_⊙, 30 pc) evolve to twin clusters over a large range of galactocentric distances. Outside this range single stellar systems are formed, whereas at small galactocentric distances the shells are tidally disrupted. In that case many small fragments formed during the collapse survive as single bound entities. About 1/3 of the twin cluster systems formed on circular orbits merge within 400 Myr. On elliptic orbits the merger rate reduces to less than 4%. Thus, there could be a significant number of twin clusters even in our Galaxy, which, however, might be undetected as twins due to a large phase shift on their common orbit. This revised version was published online in September 2006 with corrections to the Cover Date.     

The Triggered Star Formation in Rotating Disks

The gravitational instability of expanding shells triggering the formation of clouds and stars is analyzed. Disks with different scale-heights, ambient and shell velocity dispersions, mid-plane densities, rotation rates and shear rates are explored with three dimensional numerical simulations in the thin shell approximation. Three conditions for the shell collapse are specified: the first is that it happens before a significant blow-out, the second requires that the shell collapses before it is distorted by Coriolis forces and shear, and the third requires that the internal pressure in the accumulated gas is small and the fragmentation is achieved within the expansion time. The gas-rich and slowly rotating galaxies are the best sites of the triggered star formation, concluding that its importance has been much larger at the times of galaxy formation compared to the present epoch. This revised version was published online in September 2006 with corrections to the Cover Date.     

Star Formation Histories From Faint Galaxy Counts

Multispectral faint galaxy counts, including the deepest Hubble DeepField, are interpreted with the help of our evolution model PEGASE(Fioc and Rocca-Volmerange, 1997). The best fits correspond to galaxyformations at high redshifts, a pure luminosity evolution andclassical luminosity functions. The adopted cosmology is a flatuniverse with the matter density parameter Ω_M =0.3 and acosmological constant Ω_Λ =0.7. A solution with Ω_M=0.01 (open universe) is also acceptable. But a flat universe withΩ_M =1 is clearly excluded. The star formation histories for galaxytypes are derived from scenarios of evolution. The comparison with resultsalready published in the litterature, arises puzzling problems needinga further analysis of star formation tracers, specifically for bright galaxies. This revised version was published online in July 2006 with corrections to the Cover Date.     

Enhanced Star Formation in Seyfert 2 Galaxies

In this paper, we report our preliminary results on enhanced star formation activity in Seyfert 2 galaxies. By re-analysing the Tully-Fisher relation for Whittle's (1992) sample and for a Seyfert 2s' sample selecting from Veron-Cetty and Veron (1996), we find that (1) almost all Seyfert 2 galaxies with circumnuclear star formation have a ratio of far infrared (FIR) to blue luminosities (L_FIR/L_B) to be larger than 1/3; (2) for Seyfert 2 galaxies with L_FIR/L_B > 1/3, the Tuly-Fisher relation is similar to that of the normal spiral galaxies; while for those with L_FIR/L_B ≥ 1/3, they are significantly different from the normal ones, which confirms Whittle's suggestion of enhanced star formation activities in the circumnuclear regions of these Seyfert 2 galaxies. This revised version was published online in July 2006 with corrections to the Cover Date.