Major Starbursts Research Articles

Erratum: The frequency of very young galaxies in the local Universe: II. The view from SDSS spectra

Only a handful of galaxies in the local Universe appear to be very young. We estimate the fraction of very young galaxies (VYGs), defined as those with more than half their stellar masses formed within the last Gyr. We fit non-parametric star formation histories (SFHs) to ~280 000 galaxy spectra from a flux- and volume-limited subsample of the Main Galaxy Sample (MGS) of the SDSS, which is also complete in mass-to-light ratio, thus properly accounting for passive galaxies of a given mass. The VYG fractions decrease with increasing galaxy stellar mass, from ~50% at $m = 10^8\,\rm M_\odot$ to ~0.1% at $m = 10^{11.5}\,\rm M_\odot$, with differences of up to 1 dex between the different spectral models used to estimate the SFH and on how we treat aperture effects. But old stellar populations may hide in our VYGs despite our conservative VYG sample built with galaxies that are globally bluer than within the region viewed by the SDSS fibre. The VYG fractions versus mass decrease more gradually compared to the Tweed et al. predictions using analytical and semi-analytical models of galaxy formation, but agree better with the SIMBA hydrodynamical simulation. These discrepancies highlight the usefulness of VYGs in constraining the strong uncertainties in both galaxy formation models and spectral modelling of galaxy SFHs. Given the lognormal cosmic SFH, these mean VYG fractions suggest that galaxies with $m > 10^8\,\rm M_\odot$ undergo at most 4 major starbursts on average.

The frequency of very young galaxies in the local Universe – II. The view from SDSS spectra

ABSTRACT Only a handful of galaxies in the local Universe appear to be very young. We estimate the fraction of very young galaxies (VYGs), defined as those with more than half their stellar masses formed within the last Gyr. We fit non-parametric star formation histories (SFHs) to ∼280 000 galaxy spectra from a flux- and volume-limited subsample of the Main Galaxy Sample (MGS) of the SDSS, which is also complete in mass-to-light ratio, thus properly accounting for passive galaxies of a given mass. The VYG fractions decrease with increasing galaxy stellar mass, from ∼50 per cent at $m = 10^8\, {\rm M}_{\odot }$ to ∼0.1 per cent at $m = 10^{11.5}\, {\rm M}_{\odot }$, with differences of up to 1 dex between the different spectral models used to estimate the SFH and on how we treat aperture effects. But old stellar populations may hide in our VYGs despite our conservative VYG sample built with galaxies that are globally bluer than within the region viewed by the SDSS fibre. The VYG fractions versus mass decrease more gradually compared to the Tweed et al. predictions using analytical and semi-analytical models of galaxy formation, but agree better with the SIMBA hydrodynamical simulation. These discrepancies highlight the usefulness of VYGs in constraining the strong uncertainties in both galaxy formation models and spectral modelling of galaxy SFHs. Given the lognormal cosmic SFH, these mean VYG fractions suggest that galaxies above $10^8\, \rm M_\odot$ undergo at most four major starbursts on average.

The NGC 1614 interacting galaxy

Minor mergers frequently occur between giant and gas-rich low mass galaxies and can provide significant amounts of interstellar matter to refuel star formation and power AGN in the giant systems. Major starbursts and/or AGN result when fresh gas is transported and compressed in the central regions of the giant galaxy. This is the situation in NGC1614, whose molecular medium we explore at half arcsecond angular resolution through our observations of 12CO(2-1) emission using the SMA. We compare our maps with optical and Pa alpha, HST and high angular resolution radio continuum images to study the relationships between dense molecular gas and the starburst region. The most intense CO emission occurs in a partial ring with ~230pc radius around the center, with an extension to the north-west into the dust lane that contains diffuse molecular gas. We resolve 10 GMAs in the ring which has an integrated molecular mass of ~8x10^8M_sun. Our observations filter out a large part of the CO(1-0) emission mapped at shorter spacings, indicating that most of the molecular gas is diffuse and that GMAs only exist near and within the circumnuclear ring. The molecular ring is uneven with most of the mass on the western side, which also contains GMAs extending into a pronounced tidal dust lane. The spatial and kinematic patterns suggest that the northwest extension of the ring is a cosmic umbilical cord that is feeding molecular gas associated with the dust lane and tidal debris into the nuclear ring. The astrophysical process for producing a ring structure is not fully understood, but the presence of numerous GMAs suggests an orbit crowding or resonance phenomenon. There is some evidence that star formation is progressing radially outwards within the ring, indicating that a self-triggering mechanism may also affect these processes.

Top-heavy integrated galactic stellar initial mass functions in starbursts

Star formation rates (SFR) larger than 1000 Msun/ yr are observed in extreme star bursts. This leads to the formation of star clusters with masses > 10^6 Msun in which crowding of the pre-stellar cores may lead to a change of the stellar initial mass function (IMF). Indeed, the large mass-to-light ratios of ultra-compact dwarf galaxies and recent results on globular clusters suggest the IMF to become top-heavy with increasing star-forming density. We explore the implications of top-heavy IMFs in these very massive and compact systems for the integrated galactic initial mass function (IGIMF), which is the galaxy-wide IMF, in dependence of the star-formation rate of galaxies. The resulting IGIMFs can have slopes, alpha_3, for stars more massive than about 1 Msun between 1.5 and the Salpeter slope of 2.3 for an embedded cluster mass function (ECMF) slope (beta) of 2.0, but only if the ECMF has no low-mass clusters in galaxies with major starbursts. Alternatively, beta would have to decrease with increasing SFR >10 Msun/ yr such that galaxies with major starbursts have a top-heavy ECMF. The resulting IGIMFs are within the range of observationally deduced IMF variations with redshift.

THE GLOBULAR CLUSTER SYSTEM IN M87: A WIDE-FIELD STUDY WITH CFHT/MEGACAM

CFHT Megacam data in (g',r',i') are used to obtain deep, wide-field photometry of the globular cluster system (GCS) around M87. A total of 6200 GCs brighter than M_I = -8.5 are included in the study, essentially containing almost the entire bright half of the total GC population in the galaxy. The classic bimodal metal-poor and metal-rich sequences of GCs show up clearly; while the spatial distribution of the GCs can be traced detectably outward to more than 100 kpc, the blue, metal-poor subpopulation is very much more spatially extended than the red subpopulation. Both the red and blue GC subsystems have radial metallicity gradients, where mean heavy-element abundance scales with projected galactocentric distance as Z ~ R^{-0.12} (blue) and R^{-0.17} (red). The blue sequence exhibits a strongly significant mass/metallicity relation (MMR) in which the mean metallicity gradually increases with cluster luminosity as Z ~ L^{0.25 \pm 0.05} for the luminosity range M_I < -10. However, this relation is also clearly nonlinear, becoming vertical at fainter levels. The MMR can be understood as the result of self-enrichment within the most massive metal-poor GCs during their formation. In giant elliptical galaxies, the red GCs are likely to represent a broad composite population formed during several major starbursts. If so, the red sequence might display a population-based MMR that could in principle be either positive or negative.

AEGIS: Galaxy Spectral Energy Distributions from the X-Ray to Radio

The All-Wavelength Extended Groth Strip International Survey (AEGIS) team presents broadband spectral energy distributions (SEDs), from X-ray to radio wavelengths, for 71 galaxies spanning the redshift range 0.55-1.16 (⟨z⟩ ~ 0.7). Galaxies with secure redshifts are selected from a small (22 arcmin2) subsection of the Keck/DEIMOS Galaxy Redshift Survey in the Extended Groth Strip field that has also been targeted for deep panchromatic imaging by ultraviolet (120-250 nm), optical (360-900 nm), optical/near-infrared (440-1600 nm), near-infrared (1200-2200 nm), mid/far-infrared (3.6-70 μm), and radio (6-20 cm). A typical galaxy in our sample is MB = -19.82. The ultraviolet to mid-infrared portion of their SEDs are found to be bracketed by two stellar-only model SEDs: (1) an early burst followed by passive evolution and (2) a constant star formation rate since early times. This suggests that few of these galaxies are undergoing major starbursts. Approximately half the galaxies show a mid- to far-infrared excess relative to the model SEDs, consistent with thermal emission from interstellar dust. Two objects have power-law SEDs, indicating that they are dominated by active galactic nuclei; both are detected in X-rays. SEDs, from the ultraviolet to the infrared, follow expected trends: redder SEDs are associated with red U - B, early-type morphology, and low [O II] emission, and vice versa for blue SEDs.

Modelling ISO galaxy counts with luminosity and merger rate evolution

We model galaxy evolution in the 6.75- and 15-μm passbands of the ISO satellite, by combining models of galaxy evolution at optical wavelengths (which are consistent with the optical galaxy counts) with observed spectral energy distributions in the infrared. Our model is consistent with the local 12-μm galaxy luminosity function (from IRAS), including the bright end if we place a small fraction (∼3.5 per cent at z=0) of the model's spiral galaxies in interacting pairs with mid-infrared luminosities enhanced by major starbursts. Source counts from deep ISO surveys in the 6.75- and 15-μm passbands significantly exceed non-evolving predictions. We find the number counts, redshift distributions and the wide range of mid-infrared/optical flux ratios of the detected galaxies to be reasonably consistent with our evolving model. Our models suggest that the steep number count of 6.75-μm sources can be explained primarily by the evolving E/S0 galaxies. About one-third of 6.75-μm sources do appear to be E/S0s or early-type spirals, but starburst galaxies at 0.6<z<1.1 are more prominent in this passband than we expected. In the 15-μm passband the main contributors appear to be evolving, star-forming spirals and starbursting galaxy mergers, as in our model. Visibly interacting galaxies form ∼31 per cent of the 15-μm detections to 200 μJy, and have high mid-infrared/optical flux ratios consistent with our models for major starbursts with high dust extinction (up to ∼1.8 mag in the rest-frame blue band). The numbers and high mean redshift (0.68±0.08) of merging/interacting galaxies detected in these ISO surveys are only consistent with the local mid-infrared luminosity function if the merger-triggered starbursts undergo significant luminosity evolution, of ∼(1+z)2 to z∼1, in addition to the ∼(1+z)2 number evolution reflecting the increase with redshift in the fraction of merging/interacting galaxies (as estimated from optical surveys).