Search

Forming high-mass stars have a significant effect on their natal environment. Their feedback pathways, including winds, outflows, and ionising radiation, shape the evolution of their surroundings which impacts the formation of the next generation of stars. They create or reveal dense pillars of gas and dust towards the edges of the cavities they clear. They are modelled in feedback simulations, and the sizes and shapes of the pillars produced are consistent with those observed. However, these models predict measurably different kinematics which provides testable discriminants. Here we present the first ALMA Compact Array (ACA) survey of 13 pillars in Carina, observed in $^{12}$CO, $^{13}$CO and C$^{18}$O J=2-1, and the 230 GHz continuum. The pillars in this survey were chosen to cover a wide range in properties relating to the amount and direction of incident radiation, proximity to nearby irradiating clusters and cloud rims, and whether they are detached from the cloud. With these data, we are able to discriminate between models. We generally find pillar velocity dispersions of $<$ 1 km s$^{-1}$ and that the outer few layers of molecular emission in these pillars show no significant offsets from each other, suggesting little bulk internal motions within the pillars. There are instances where the pillars are offset in velocity from their parental cloud rim, and some with no offset, hinting at a stochastic development of these motions., Comment: 24 Pages, 19 figures. Accepted to MNRAS

We present new MUSE/VLT observations of a small globule in the Carina H II region that hosts the HH 900 jet+outflow system. Data were obtained with the GALACSI ground-layer adaptive optics system in wide-field mode, providing spatially-resolved maps of diagnostic emission lines. These allow us to measure the variation of the physical properties in the globule and jet+outflow system. We find high temperatures ($T_e \approx 10^4$ K), modest extinction ($A_V \approx 2.5$ mag), and modest electron densities ($n_e \approx 200$ cm$^{-3}$) in the ionized gas. Higher excitation lines trace the ionized outflow; both the excitation and ionization in the outflow increase with distance from the opaque globule. In contrast, lower excitation lines that are collisionally de-excited at densities $\gtrsim 10^4$ cm$^{-3}$ trace the highly collimated protostellar jet. Assuming the globule is an isothermal sphere confined by the pressure of the ionization front, we compute a Bonnor-Ebert mass of $\sim 3.7$ M$_{\odot}$. This is two orders of magnitude higher than previous mass estimates, calling into question whether small globules like the Tadpole contribute to the bottom of the IMF. Derived globule properties are consistent with a cloud that has been and/or will be compressed by the ionization front on its surface. At the estimated globule photoevaporation rate of $\sim 5 \times 10^{-7}$ M$_{\odot}$ yr$^{-1}$, the globule will be completely ablated in $\sim 7$ Myr. Stars that form in globules like the Tadpole will emerge into the H II later and may help resolve some of the temporal tension between disk survival and enrichment., Comment: MNRAS, accepted

In this paper we provide a comprehensive description of the internal dynamics of G0.253+0.016 (a.k.a. 'the Brick'); one of the most massive and dense molecular clouds in the Galaxy to lack signatures of widespread star formation. As a potential host to a future generation of high-mass stars, understanding largely quiescent molecular clouds like G0.253+0.016 is of critical importance. In this paper, we reanalyse Atacama Large Millimeter Array cycle 0 HNCO $J=4(0,4)-3(0,3)$ data at 3 mm, using two new pieces of software which we make available to the community. First, scousepy, a Python implementation of the spectral line fitting algorithm scouse. Secondly, acorns (Agglomerative Clustering for ORganising Nested Structures), a hierarchical n-dimensional clustering algorithm designed for use with discrete spectroscopic data. Together, these tools provide an unbiased measurement of the line of sight velocity dispersion in this cloud, $\sigma_{v_{los}, {\rm 1D}}=4.4\pm2.1$ kms$^{-1}$, which is somewhat larger than predicted by velocity dispersion-size relations for the Central Molecular Zone (CMZ). The dispersion of centroid velocities in the plane of the sky are comparable, yielding $\sigma_{v_{los}, {\rm 1D}}/\sigma_{v_{pos}, {\rm 1D}}\sim1.2\pm0.3$. This isotropy may indicate that the line-of-sight extent of the cloud is approximately equivalent to that in the plane of the sky. Combining our kinematic decomposition with radiative transfer modelling we conclude that G0.253+0.016 is not a single, coherent, and centrally-condensed molecular cloud; 'the Brick' is not a \emph{brick}. Instead, G0.253+0.016 is a dynamically complex and hierarchically-structured molecular cloud whose morphology is consistent with the influence of the orbital dynamics and shear in the CMZ., Comment: 29 pages, 19 figures, 1 table (including appendix). Accepted for publication in MNRAS (February 4, 2019). Scousepy is available here: https://github.com/jdhenshaw/scousepy. Acorns is available here: https://github.com/jdhenshaw/acorns

The evolution of molecular clouds in galactic centres is thought to differ from that in galactic discs due to a significant influence of the external gravitational potential. We present a set of numerical simulations of molecular clouds orbiting on the 100-pc stream of the Central Molecular Zone (the central $\sim500$ pc of the Galaxy) and characterise their morphological and kinematic evolution in response to the background potential and eccentric orbital motion. We find that the clouds are shaped by strong shear and torques, by tidal and geometric deformation, and by their passage through the orbital pericentre. Within our simulations, these mechanisms control cloud sizes, aspect ratios, position angles, filamentary structure, column densities, velocity dispersions, line-of-sight velocity gradients, spin angular momenta, and kinematic complexity. By comparing these predictions to observations of clouds on the Galactic Centre 'dust ridge', we find that our simulations naturally reproduce a broad range of key observed morphological and kinematic features, which can be explained in terms of well-understood physical mechanisms. We argue that the accretion of gas clouds onto the central regions of galaxies, where the rotation curve turns over and the tidal field is fully compressive, is accompanied by transformative dynamical changes to the clouds, leading to collapse and star formation. This can generate an evolutionary progression of cloud collapse with a common starting point, which either marks the time of accretion onto the tidally-compressive region or of the most recent pericentre passage. Together, these processes may naturally produce the synchronised starbursts observed in numerous (extra)galactic nuclei., Comment: 21 pages, 8 figures, 3 tables; accepted by MNRAS (February 5, 2019); Figures 2, 3, and 4 show the main results of the paper. An animated version of Figure 2 is available as an ancillary file in the arXiv source and will be included as online Supporting Information with the MNRAS publication

We present MUSE integral field data of two HII region complexes in the Large Magellanic Cloud (LMC), N44 and N180. Both regions consist of a main superbubble and a number of smaller, more compact HII regions that formed on the edge of the superbubble. For a total of 11 HII regions, we systematically analyse the radiative and mechanical feedback from the massive O-type stars on the surrounding gas. We exploit the integral field property of the data and the coverage of the HeII$\lambda$5412 line to identify and classify the feedback-driving massive stars, and from the estimated spectral types and luminosity classes we determine the stellar radiative output in terms of the ionising photon flux $Q_{0}$. We characterise the HII regions in terms of their sizes, morphologies, ionisation structure, luminosity and kinematics, and derive oxygen abundances via emission line ratios. We analyse the role of different stellar feedback mechanisms for each region by measuring the direct radiation pressure, the pressure of the ionised gas, and the pressure of the shock-heated winds. We find that stellar winds and ionised gas are the main drivers of HII region expansion in our sample, while the direct radiation pressure is up to three orders of magnitude lower than the other terms. We relate the total pressure to the star formation rate per unit area, $\Sigma_{SFR}$, for each region and find that stellar feedback has a negative effect on star formation, and sets an upper limit to $\Sigma_{SFR}$ as a function of increasing pressure., Comment: Accepted for publication in MNRAS, 27 pages, 21 figures

We present an analysis of the effect of feedback from O- and B-type stars with data from the integral field spectrograph MUSE mounted on the Very Large Telescope of pillar-like structures in the Carina Nebular Complex, one of the most massive star-forming regions in the Galaxy. For the observed pillars, we compute gas electron densities and temperatures maps, produce integrated line and velocity maps of the ionised gas, study the ionisation fronts at the pillar tips, analyse the properties of the single regions, and detect two ionised jets originating from two distinct pillar tips. For each pillar tip we determine the incident ionising photon flux $Q_\mathrm{0,pil}$ originating from the nearby massive O- and B-type stars and compute the mass-loss rate $\dot{M}$ of the pillar tips due to photo-evaporation caused by the incident ionising radiation. We combine the results of the Carina data set with archival MUSE data of a pillar in NGC 3603 and with previously published MUSE data of the Pillars of Creation in M16, and with a total of 10 analysed pillars, find tight correlations between the ionising photon flux and the electron density, the electron density and the distance from the ionising sources, and the ionising photon flux and the mass-loss rate. The combined MUSE data sets of pillars in regions with different physical conditions and stellar content therefore yield an empirical quantification of the feedback effects of ionising radiation. In agreement with models, we find that $\dot{M}\propto Q_\mathrm{0,pil}^{1/2}$., Comment: Accepted for publication in MNRAS, 36 pages, 30 figures

Using spectral-line observations of HNCO, N2H+, and HNC, we investigate the kinematics of dense gas in the central ~250 pc of the Galaxy. We present SCOUSE (Semi-automated multi-COmponent Universal Spectral-line fitting Engine), a line fitting algorithm designed to analyse large volumes of spectral-line data efficiently and systematically. Unlike techniques which do not account for complex line profiles, SCOUSE accurately describes the {l, b, v_LSR} distribution of CMZ gas, which is asymmetric about Sgr A* in both position and velocity. Velocity dispersions range from 2.6 km/s<\sigma<53.1 km/s. A median dispersion of 9.8 km/s, translates to a Mach number, M_3D>28. The gas is distributed throughout several "streams", with projected lengths ~100-250 pc. We link the streams to individual clouds and sub-regions, including Sgr C, the 20 and 50 km/s clouds, the dust ridge, and Sgr B2. Shell-like emission features can be explained by the projection of independent molecular clouds in Sgr C and the newly identified conical profile of Sgr B2 in {l ,b, v_LSR} space. These features have previously invoked supernova-driven shells and cloud-cloud collisions as explanations. We instead caution against structure identification in velocity-integrated emission maps. Three geometries describing the 3-D structure of the CMZ are investigated: i) two spiral arms; ii) a closed elliptical orbit; iii) an open stream. While two spiral arms and an open stream qualitatively reproduce the gas distribution, the most recent parameterisation of the closed elliptical orbit does not. Finally, we discuss how proper motion measurements of masers can distinguish between these geometries, and suggest that this effort should be focused on the 20 km/s and 50 km/s clouds and Sgr C., Comment: Accepted for publication in MNRAS. 30 pages (including appendices), 22 figures, 3 tables. This updated version includes minor corrections (including typos and updated references). SCOUSE is available for download at https://github.com/jdhenshaw/SCOUSE. A selection of PPV movies can be downloaded at https://github.com/jdhenshaw/PPV_movies

A nebular analysis of the central Orion Nebula and its main structures is presented. We exploit MUSE integral field observations in the wavelength range 4595-9366 \r{A} to produce the first O, S and N ionic and total abundance maps of a region spanning 6' x 5' with a spatial resolution of 0.2". We use the S$_{23}$ ( = ([SII]$\lambda$6717,31+[SIII]$\lambda$9068)/H$\beta$) parameter, together with [OII]/[OIII] as an indicator of the degree of ionisation, to distinguish between the various small-scale structures. The only Orion Bullet covered by MUSE is HH 201, which shows a double component in the [FeII]$\lambda$8617 line throughout indicating an expansion, and we discuss a scenario in which this object is undergoing a disruptive event. We separate the proplyds located south of the Bright Bar into four categories depending on their S$_{23}$ values, propose the utility of the S$_{23}$ parameter as an indicator of the shock-contribution to the excitation of line-emitting atoms, and show that the MUSE data is able to identify the proplyds associated with disks and microjets. We compute the second order structure function for the H$\alpha$, [OIII]$\lambda$5007, [SII]$\lambda$6731 and [OI]$\lambda$6300 emission lines to analyse the turbulent velocity field of the region covered with MUSE. We find that the spectral and spatial resolution of MUSE is not able to faithfully reproduce the structure functions of previous works., Comment: 34 pages, 31 figures, published in MNRAS, added missing citation

Filaments represent a key structure during the early stages of the star formation process. Simulations show filamentary structure commonly formed before and during the formation of cores. Aims. The Serpens Core represents an ideal laboratory to test the state-of-the-art of simulations of turbulent Giant Molecular Clouds. We use Herschel observations of the Serpens Core to compute temperature and column density maps of the region. Among the simulations of Dale et al. (2012), we select the early stages of their Run I, before stellar feedback is initiated, with similar total mass and physical size as the Serpens Core. We derive temperature and column density maps also from the simulations. The observed distribution of column densities of the filaments has been analysed first including and then masking the cores. The same analysis has been performed on the simulations as well. A radial network of filaments has been detected in the Serpens Core. The analysed simulation shows a striking morphological resemblance to the observed structures. The column density distribution of simulated filaments without cores shows only a log-normal distribution, while the observed filaments show a power-law tail. The power-law tail becomes evident in the simulation if one focuses just on the column density distribution of the cores. In contrast, the observed cores show a flat distribution. Even though the simulated and observed filaments are subjectively similar-looking, we find that they behave in very different ways. The simulated filaments are turbulence-dominated regions, the observed filaments are instead self-gravitating structures that will probably fragment into cores., Comment: Accepted for publication on A&A. Main text: 12 pages, 7 figures

We investigate the longevity of broad bridge features in position-velocity diagrams that appear as a result of cloud-cloud collisions. Broad bridges will have a finite lifetime due to the action of feedback, conversion of gas into stars and the timescale of the collision. We make a series of analytic arguments with which to estimate these lifetimes. Our simple analytic arguments suggest that for collisions between clouds larger than R~10 pc the lifetime of the broad bridge is more likely to be determined by the lifetime of the collision rather than the radiative or wind feedback disruption timescale. However for smaller clouds feedback becomes much more effective. This is because the radiative feedback timescale scales with the ionising flux Nly as R^{7/4}Nly^{-1/4} so a reduction in cloud size requires a relatively large decrease in ionising photons to maintain a given timescale. We find that our analytic arguments are consistent with new synthetic observations of numerical simulations of cloud-cloud collisions (including star formation and radiative feedback). We also argue that if the number of observable broad bridges remains ~ constant, then the disruption timescale must be roughly equivalent to the collision rate. If this is the case our analytic arguments also provide collision rate estimates, which we find are readily consistent with previous theoretical models at the scales they consider (clouds larger than about 10 pc) but are much higher for smaller clouds., Comment: 11 pages, 10 figures, accepted for publication in the MNRAS main journal

StarBench is a project focused on benchmarking and validating different star-formation and stellar feedback codes. In this first StarBench paper we perform a comparison study of the D-type expansion of an HII region. The aim of this work is to understand the differences observed between the twelve participating numerical codes against the various analytical expressions examining the D-type phase of HII region expansion. To do this, we propose two well-defined tests which are tackled by 1D and 3D grid- and SPH- based codes. The first test examines the `early phase' D-type scenario during which the mechanical pressure driving the expansion is significantly larger than the thermal pressure of the neutral medium. The second test examines the `late phase' D-type scenario during which the system relaxes to pressure equilibrium with the external medium. Although they are mutually in excellent agreement, all twelve participating codes follow a modified expansion law that deviates significantly from the classical Spitzer solution in both scenarios. We present a semi-empirical formula combining the two different solutions appropriate to both early and late phases that agrees with high-resolution simulations to $\lesssim2\%$. This formula provides a much better benchmark solution for code validation than the Spitzer solution. The present comparison has validated the participating codes and through this project we provide a dataset for calibrating the treatment of ionizing radiation hydrodynamics codes., Comment: 20 pages, 12 Figures, 4 Tables. Accepted for publication in MNRAS. Comments are welcome. Participation in future StarBench tests is also welcome

We examine the combined effects of winds and photoionizing radiation from O--type stars on embedded stellar clusters formed in model turbulent molecular clouds covering a range of masses and radii. We find that feedback is able to increase the quantities of dense gas present, but decreases the rate and efficiency of the conversion of gas to stars relative to control simulations in which feedback is absent. Star formation in these calculations often proceeds at a rate substantially slower than the freefall rate in the dense gas. This decoupling is due to the weakening of, and expulsion of gas from, the deepest parts of the clouds' potential wells where most of the star formation occurs in the control simulations. This results in large fractions of the stellar populations in the feedback simulation becoming dissociated from dense gas. However, where star formation \emph{does} occur in both control and feedback simulations, it does so in dense gas, so the correlation between star formation activity and dense gas is preserved. The overall dynamical effects of feedback on the \emph{clusters} are minimal, with only small fraction of stars becoming unbound, despite large quantities of gas being expelled from some clouds. This owes to the settling of the stars into virialised and stellar--dominated configurations before the onset of feedback. By contrast, the effects of feedback on the observable properties of the clusters -- their U--, B-- and V--band magnitudes -- are strong and sudden. The timescales on which the clusters become visible and unobscured are short compared with the timescales which the clouds are actually destroyed., Comment: 19 pages, 17 figures (many degraded to get below submission size limit), accepted by MNRAS

Integral field unit (IFU) data of the iconic Pillars of Creation in M16 are presented. The ionisation structure of the pillars was studied in great detail over almost the entire visible wavelength range, and maps of the relevant physical parameters, e.g. extinction, electron density, electron temperature, line-of-sight velocity of the ionised and neutral gas are shown. In agreement with previous authors, we find that the pillar tips are being ionised and photo-evaporated by the massive members of the nearby cluster NGC 6611. They display a stratified ionisation structure where the emission lines peak in a descending order according to their ionisation energies. The IFU data allowed us to analyse the kinematics of the photo-evaporative flow in terms of the stratified ionisation structure, and we find that, in agreement with simulations, the photo-evaporative flow is traced by a blueshift in the position-velocity profile. The gas kinematics and ionisation structure have allowed us to produce a sketch of the 3D geometry of the Pillars, positioning the pillars with respect to the ionising cluster stars. We use a novel method to detect a previously unknown bipolar outflow at the tip of the middle pillar and suggest that it has an embedded protostar as its driving source. Furthermore we identify a candidate outflow in the leftmost pillar. With the derived physical parameters and ionic abundances, we estimate a mass loss rate due to the photo-evaporative flow of 70 M$_{\odot}$ Myr$^{-1}$ which yields an expected lifetime of approximately 3 Myr., Comment: 25 pages, 21 figures, appendix 10 pages and 10 figures, MNRAS in press

Collisions between giant molecular clouds are a potential mechanism for triggering the formation of massive stars, or even super star clusters. The trouble is identifying this process observationally and distinguishing it from other mechanisms. We produce synthetic position-velocity diagrams from models of: cloud-cloud collisions, non-interacting clouds along the line of sight, clouds with internal radiative feedback and a more complex cloud evolving in a galactic disc, to try and identify unique signatures of collision. We find that a broad bridge feature connecting two intensity peaks, spatially correlated but separated in velocity, is a signature of a high velocity cloud-cloud collision. We show that the broad bridge feature is resilient to the effects of radiative feedback, at least to around 2.5Myr after the formation of the first massive (ionising) star. However for a head on 10km/s collision we find that this will only be observable from 20-30 per cent of viewing angles. Such broad-bridge features have been identified towards M20, a very young region of massive star formation that was concluded to be a site of cloud-cloud collision by Torii et al (2011), and also towards star formation in the outer Milky Way by Izumi et al (2014)., Comment: 12 pages. Accepted for publication in the MNRAS main journal

We examine the evidence offered for triggered star formation against the backdrop provided by recent numerical simulations of feedback from massive stars at or below giant molecular cloud sizescales. We compile a catalogue of sixty--seven observational papers, mostly published over the last decade, and examine the signposts most commonly used to infer the presence of triggered star formation. We then determine how well these signposts perform in a recent suite of hydrodynamic simulations of star formation including feedback from O--type stars performed by Dale et al (2012a, b, 2013a, b, 2014). We find that none of the observational markers improve the chances of correctly identifying a given star as triggered by more than factors of two at most. This limits the fidelity of these techniques in interpreting star formation histories. We therefore urge caution in interpreting observations of star formation near feedback--driven structures in terms of triggering., Comment: 16 pages, 6 figures, accepted by MNRAS

Aims. We investigate the effects of ionising photons on accretion and stellar mass growth in a young star forming region, using a Monte Carlo radiation transfer code coupled to a smoothed particle hydrodynamics (SPH) simulation. Methods. We introduce the framework with which we correct stellar cluster masses for the effects of photoionising (PI) feedback and compare to the results of a full ionisation hydrodynamics code. Results. We present results of our simulations of star formation in the spiral arm of a disk galaxy, including the effects of photoionising radiation from high mass stars. We find that PI feedback reduces the total mass accreted onto stellar clusters by approximately 23 per cent over the course of the simulation and reduces the number of high mass clusters, as well as the maximum mass attained by a stellar cluster. Mean star formation rates (SFRs) drop from 0.042 solar masses per year in our control run to 0.032 solar masses per year after the inclusion of PI feedback with a final instantaneous SFR reduction of 62 per cent. The overall cluster mass distribution appears to be affected little by PI feedback. Conclusions. We compare our results to the observed extra-galactic Schmidt-Kennicutt relation and the observed properties of local star forming regions in the Milky Way and find that internal photoionising (PI) feedback is unlikely to reduce star formation rates by more than a factor of approximately 2 and thus may play only a minor role in regulating star formation., Comment: 10 pages, 11 figures, A&A, in press

Giant Molecular Clouds (GMCs) are observed to be turbulent, but theory shows that without a driving mechanism turbulence should quickly decay. The question arises by which mechanisms turbulence is driven or sustained. It has been shown that photoionising feedback from massive stars has an impact on the surrounding GMC and can for example create vast HII bubbles. We therefore address the question of whether turbulence is a consequence of this effect of feedback on the cloud. To investigate this, we analyse the velocity field of simulations of high mass star forming regions by studying velocity structure functions and power spectra. We find that clouds whose morphology is strongly affected by photoionising feedback also show evidence of driving of turbulence by preserving or recovering a Kolmogorov-type velocity field. On the contrary, control run simulations without photoionising feedback have a velocity distribution that bears the signature of gravitational collapse and of the dissipation of energy, where the initial Kolmogorov-type structure function is erased., Comment: accepted by MNRAS, 13 pages, 12 figures

Young stars show evidence of accretion discs which evolve quickly and disperse with an e-folding time of $\sim$ 3Myr. This is in striking contrast with recent observations that suggest evidence for numerous $>30$ Myr old stars with an accretion disc in large star-forming complexes. We consider whether these observations of apparently old accretors could be explained by invoking Bondi-Hoyle accretion to rebuild a new disc around these stars during passage through a clumpy molecular cloud. We combine a simple Monte Carlo model to explore the capture of mass by such systems with a viscous evolution model to infer the levels of accretion that would be observed. We find that a significant fraction of stars may capture enough material via the Bondi-Hoyle mechanism to rebuild a disc of mass $\gtrsim$ 1 minimum-mass solar nebula, and $\lesssim 10\%$ accrete at observable levels at any given time. A significant fraction of the observed old accretors may be explained with our proposed mechanism. Such accretion may provide a chance for a second epoch of planet formation, and have unpredictable consequences for planetary evolution., Comment: 5 pages, 2 figures. Accepted by A&A

We model the combined effects of photoionization and momentum--driven winds from O--stars on molecular clouds spanning a parameter space of initial conditions. The dynamical effects of the winds are very modest. However, in the lower--mass clouds, they influence the morphologies of the HII regions by creating 10pc--scale central cavities.\\ The inhomogeneous structures of the model GMCs make them highly permeable to photons, ionized gas and supernova ejecta, and the leaking of ionized gas in particular strongly affects their evolution, reducing the effectiveness of feedback. Nevertheless, feedback is able to expel large fractions of the mass of the lower escape--velocity clouds. Its impact on star formation is more modest, decreasing final star formation efficiencies by 10--20$\%$, and the rate of change of the star formation efficiency per freefall time by about one third. However, the clouds still form stars substantially faster than observed GMCs., Comment: 21 pages, 19 figures, accepted for publication in MNRAS

The major physical processes responsible for shaping and sculpting pillars in the clouds surrounding massive stars (i.e. the `Pillars of Creation') are now being robustly incorporated into models quantifying the ionizing radiation from massive stars. The detailed gas dynamics within these pillars can now be compared with observations. Our goal is to quantify the gas dynamics in a pillar being sculpted by a nearby massive star. To do this, we use the CO, $^{13}$CO, and C$^{18}$O J=1-0 emission towards a pillar in the Vulpecula Rift. These data are a combination of CARMA and FCRAO observations providing high resolution ($\sim5''$) imaging of large scale pillar structures ($>100''$). We find that this cold ($\sim18$ K), low density material ($8\times10^3$ cm$^{-3}$) material is fragmenting on Jeans scales, has very low velocity dispersions ($\sim0.5$ km s$^{-1}$), and appears to be moving away from the ionizing source. We are able to draw direct comparisons with three models from the literature, and find that those with lower velocity dispersions best fit our data, although the dynamics of any one model do not completely agree with our observations. We do however, find that our observed pillar exhibits many of the characteristics expected from simulations., Comment: 8 pages, including 7 figures. Accepted for publication in MNRAS

We examine the effect of momentum-driven OB-star stellar winds on a parameter space of simulated turbulent Giant Molecular Clouds using SPH hydrodynamical simulations. By comparison with identical simulations in which ionizing radiation was included instead of winds, we show that momentum-driven winds are considerably less effective in disrupting their host clouds than are HII regions. The wind bubbles produced are smaller and generally smoother than the corresponding ionization-driven bubbles. Winds are roughly as effective in destroying the very dense gas in which the O-stars are embedded, and thus shutting down the main regions of star-forming activity in the model clouds. However, their influence falls off rapidly with distance from the sources, so they are not as good at sweeping up dense gas and triggering star formation further out in the clouds. As a result, their effect on the star formation rate and efficiency is generally more negative than that of ionization, if they exert any effect at all., Comment: Accepted by MNRAS, 18 pages, 18 figures

We present optical integral field unit (IFU) observations of two gas pillars surrounding the Galactic young massive star cluster NGC 3603. The high S/N and spectral resolution of these data have allowed us to accurately quantify the H-alpha, [NII] and [SII] emission line shapes, and we find a mixture of broad (FWHM~70-100 km/s) and narrow (<50 km/s) components. The broad components are found close to the edges of both pillars, suggesting that they originate in turbulent mixing layers (TMLs) driven by the effect of the star cluster wind. Both pillars exhibit surprisingly high ionized gas densities of >10000 cm-3. In one pillar we found that these high densities are only found in the narrow component, implying it must originate from deeper within the pillar than the broad component. From this, together with our kinematical data, we conclude that the narrow component traces a photoevaporation flow, and that the TML forms at the interface with the hot wind. On the pillar surfaces we find a consistent offset in radial velocity between the narrow (brighter) components of H-alpha and [NII] of ~5-8 km/s, for which we were unable to find a satisfactory explanation. We urge the theoretical community to simulate mechanical and radiative cloud interactions in more detail to address the many unanswered questions raised by this study., Comment: 17 pages, 19 figures, accepted to MNRAS

We present the fourth in a series of papers detailing our SPH study of the effects of ionizing feedback from O--type stars on turbulent star forming clouds. Here, we study the effects of photoionization on a series of initially partially unbound clouds with masses ranging from $10^{4}$--$10^{6}$M$_{\odot}$ and initial sizes from 2.5-45pc. We find that ionizing feedback profoundly affects the structure of the gas in most of our model clouds, creating large and often well-cleared bubble structures and pillars. However, changes in the structures of the embedded clusters produced are much weaker and not well correlated to the evolution of the gas. We find that in all cases, star formation efficiencies and rates are reduced by feedback and numbers of objects increased, relative to control simulations. We find that local triggered star formation does occur and that there is a good correlation between triggered objects and pillars or bubble walls, but that triggered objects are often spatially-mixed with those formed spontaneously. Some triggered objects acquire large enough masses to become ionizing sources themselves, lending support to the concept of propagating star formation. We find scant evidence for spatial age gradients in most simulations, and where we do see them, they are not a good indicator of triggering, as they apply equally to spontaneously-formed objects as triggered ones. Overall, we conclude that inferring the global or local effects of feedback on stellar populations from observing a system at a single epoch is very problematic., Comment: 17 pages, 11 figures (mostly degraded to get under the submission size limit), accepted by MNRAS

We extend our previous SPH parameter study of the effects of photoionization from O-stars on star-forming clouds to include initially unbound clouds. We generate a set of model clouds in the mass range $10^{4}-10^{6}$M$_{\odot}$ with initial virial ratios $E_{\rm kin}/E_{\rm pot}$=2.3, allow them to form stars, and study the impact of the photoionizing radiation produced by the massive stars. We find that, on the 3Myr timescale before supernovae are expected to begin detonating, the fractions of mass expelled by ionizing feedback is a very strong function of the cloud escape velocities. High-mass clouds are largely unaffected dynamically, while lower-mass clouds have large fractions of their gas reserves expelled on this timescale. However, the fractions of stellar mass unbound are modest and significant portions of the unbound stars are so only because the clouds themselves are initially partially unbound. We find that ionization is much more able to create well-cleared bubbles in the unbound clouds, owing to their intrinsic expansion, but that the presence of such bubbles does not necessarily indicate that a given cloud has been strongly influenced by feedback. We also find, in common with the bound clouds from our earlier work, that many of the systems simulated here are highly porous to photons and supernova ejecta, and that most of them will likely survive their first supernova explosions., Comment: 14 pages, 13 figures (some degraded and greyscaled), accepted by MNRAS

We present a detailed study of star formation occurring in bound star--forming clouds under the influence of internal ionizing feedback from massive stars across a spectrum of cloud properties. We infer which objects are triggered by comparing our feedback simulations with control simulations in which no feedback was present. We find feedback always results in a lower star--formation efficiency and usually but not always results in a larger number of stars or clusters. Cluster mass functions are not strongly affected by feedback, but stellar mass functions are biased towards lower masses. Ionization also affects the geometrical distribution of stars in ways that are robust against projection effects, but may make the stellar associations more or less subclustered depending on the background cloud environment. We observe a prominent pillar in one simulation which is the remains of an accretion flow feeding the central ionizing cluster of its host cloud and suggest that this may be a general formation mechanism for pillars such as those observed in M16. We find that the association of stars with structures in the gas such as shells or pillars is a good but by no means foolproof indication that those stars have been triggered and we conclude overall that it is very difficult to deduce which objects have been induced to form and which formed spontaneously simply from observing the system at a single time., Comment: 15 pages, 12 figures, to appear in MNRAS

We present an SPH parameter study of the dynamical effect of photoionization from O--type stars on star--forming clouds of a range of masses and sizes during the time window before supernovae explode. Our model clouds all have the same degree of turbulent support initially, the ratio of turbulent kinetic energy to gravitational potential energy being set to $E_{\rm kin}/|E_{\rm pot}|$=0.7. We allow the clouds to form stars and study the dynamical effects of the ionizing radiation from the massive stars or clusters born within them. We find that dense filamentary structures and accretion flows limit the quantities of gas that can be ionized, particularly in the higher density clusters. More importantly, the higher escape velocities in our more massive (10$^{6}$M$_{\odot}$) clouds prevent the HII regions from sweeping up and expelling significant quantities of gas, so that the most massive clouds are largely dynamically unaffected by ionizing feedback. However, feedback has a profound effect on the lower--density 10$^{4}$ and 10$^{5}$M$_{\odot}$ clouds in our study, creating vast evacuated bubbles and expelling tens of percent of the neutral gas in the 3Myr timescale before the first supernovae are expected to detonate, resulting in clouds highly porous to both photons and supernova ejecta., Comment: 18 pages, 17 figures (mostly degraded to get under the filesize limit), accepted by MNRAS

We present SPH simulations of protoclusters including the effects of winds from massive stars. Using a particle-injection method, we investigate the effect of structure close to the wind sources and the short-timescale influence of winds on protoclusters. Structures such as disks and gaseous filaments have a strong collimating effect. By a different technique of injecting momentum from point sources, we compare the large-scale, long-term effects of isotropic and intrinsically-collimated winds and find them to be similar. Both types of wind dramatically slow the global star formation process, but the timescale on which they expel significant mass from the cluster is rather long (approaching 10 freefall times). Clusters may then experience rapid star formation early in their lifetimes, before switching to a mode where gas is gradually expelled, while star formation proceeds much more slowly. This complicates conclusions regarding slow star formation derived from measuring the star-formation efficiency per freefall time. Estimates of the efficacy of winds in dispersing clusters derived simply from the total wind momentum flux may not be very reliable. This is due to material being expelled from deep within stellar potential wells, often to velocities well in excess of the cluster escape velocity, and also to the loss of momentum flux through holes in the gas distribution. Winds have little effect on the accretion--driven stellar IMF except at the very high-mass end, where they serve to prevent some of the most massive objects accreting. We also find that the morphology of the gas, the rapid motions of the wind sources and the action of accretion flows prevent the formation of bubble-like structures. This may make it difficult to discern the influence of winds on very young clusters., Comment: 14 pages, 17 figures, accepted for publication in MNRAS, replaced with revised version as references missing in original

In this paper, we examine numerically the difference between triggered and revealed star formation. We present Smoothed Particle Hydrodynamics (SPH) simulations of the impact on a turbulent 10^4 solar-mass molecular cloud of irradiation by an external source of ionising photons. In particular, using a control model, we investigate the triggering of star formation within the cloud. We find that, although feedback has a dramatic effect on the morphology of our model cloud, its impact on star formation is relatively minor. We show that external irradiation has both positive and negative effects, accelerating the formation of some objects, delaying the formation of others, and inducing the formation of some that would not otherwise have formed. Overall, the calculation in which feedback is included forms nearly twice as many objects over a period of \sim0.5 freefall times (\sim2.4 Myr), resulting in a star--formation efficiency approximately one third higher (\sim4% as opposed to \sim3% at this epoch) as in the control run in which feedback is absent. Unfortunately, there appear to be no observable characteristics which could be used to differentiate objects whose formation was triggered from those which were forming anyway and which were simply revealed by the effects of radiation, although this could be an effect of poor statistics., Comment: 12 pages, 9 figures, accepted by MNRAS

We conduct Smoothed Particle Hydrodynamics simulations of the `collect and collapse' scenario (Elmegreen and Lada, 1977) for star formation triggered by an expanding HII region. We simulate the evolution of a spherical uniform molecular cloud with an ionising source at its centre. The gas in the cloud is self-gravitating, although the cloud is prevented from globally collapsing. We find that the shell driven by the HII region fragments to form numerous self--gravitating objects. We repeat our calculations at four numerical resolutions to ensure that they are converged. We compare our results to the analytical model of Whitworth et al, 1994 and show that our simulations and the predictions of Whitworth et al are in good agreement in the sense that the shell fragments at the time and radius predicted by Whitworth et al to within 20% and 25% respectively. Most of the fragments produced in our two highest resolution calculations are approximately half the mass of those predicted, but this conclusion is robust against both numerical resolution and the presence of random noise (local fluctuations in density of a factor of $\sim2$) in the initial gas distribution. We conclude that such noise has little impact on the fragmentation process., Comment: laTeX, 9 pages, 8 figures, accepted by MNRAS

We present numerical simulations of the runaway fractions expected amongst O and Wolf-Rayet star populations resulting from stars ejected from binaries by the supernova of the companion. Observationally the runaway fraction for both types of star is similar, prompting the explanation that close dynamical interactions are the main cause of these high-velocity stars. We show that, provided that the initial binary fraction is high, a scenario in which two-thirds of massive runaways are from supernovae is consistent with these observations. Our models also predict a low frequency of runaways with neutron star companions and a very low fraction of observable Wolf-Rayet--compact companion systems., Comment: MNRAS accepted, 10 pages, 5 figures

We present the first ever hydrodynamic calculations of star cluster formation that incorporate the effect of feedback from ionising radiation. In our simulations, the ionising source forms in the cluster core at the intersection of several dense filaments of inflowing gas. We show that these filaments collimate ionised outflows and suggest such an environmental origin for at least some observed outflows in regions of massive star formation. Our simulations show both positive feedback (i.e. promotion of star formation in neutral gas compressed by expanding HII regions) and negative feedback (i.e. suppression of the accretion flow in to the central regions). We show that the volume filling factor of ionised gas is very different in our simulations than would result from the case where the central source interacted with an azimuthally smoothed gas density distribution. As expected, gas density is the key parameter in determining whether clusters are unbound by photoionising radiation. Nevertheless, we find - on account of the acceleration of a small fraction of the gas to high velocities in the outflows - that the deposition in the gas of an energy that exceeds the binding energy of the cluster is not a sufficient criterion for unbinding the bulk of the cluster mass., Comment: 16 pages, 21 figures, LaTeX. Accepted for publication in MNRAS