Star-forming regions that are visible at 3.6 microns and Halpha but not in the u,g,r,i,z bands of the Sloan Digital Sky survey (SDSS), are measured in five nearby spiral galaxies to find extinctions averaging ~3.8 mag and stellar masses averaging ~5x
10^4 Msun. These regions are apparently young star complexes embedded in dark filamentary shock fronts connected with spiral arms. The associated cloud masses are ~10^7 Msun. The conditions required to make such complexes are explored, including gravitational instabilities in spiral shocked gas and compression of incident clouds. We find that instabilities are too slow for a complete collapse of the observed spiral filaments, but they could lead to star formation in the denser parts. Compression of incident clouds can produce a faster collapse but has difficulty explaining the semi-regular spacing of some regions along the arms. If gravitational instabilities are involved, then the condensations have the local Jeans mass. Also in this case, the near-simultaneous appearance of equally spaced complexes suggests that the dust lanes, and perhaps the arms too, are relatively young.
Local UV-bright galaxies in the Kiso survey include clumpy systems with kpc-size star complexes that resemble clumpy young galaxies in surveys at high redshift. We compare clump masses and underlying disks in several dozen galaxies from each of these
surveys to the star complexes and disks of normal spirals. Photometry and spectroscopy for the Kiso and spiral sample come from the Sloan Digital Sky Survey. We find that the largest Kiso clumpy galaxies resemble UDF clumpies in terms of the star formation rates, clump masses, and clump surface densities. Clump masses and surface densities in normal spirals are smaller. If the clump masses are proportional to the turbulent Jeans mass in the interstellar medium, then for the most luminous galaxies in the sequence of normal:Kiso:UDF, the turbulent speeds and surface densities increase in the proportions 1.0:4.7:5.0 and 1.0:4.0:5.1, respectively, for fixed restframe B-band absolute magnitude. For the least luminous galaxies in the overlapping magnitude range, the turbulent speed and surface density trends are 1.0:2.7:7.4 and 1.0:1.4:3.0, respectively. We also find that while all three types have radially decreasing disk intensities when measured with ellipse-fit azimuthal averages, the average profiles are more irregular for UDF clumpies (which are viewed in their restframe UV) than for Kiso galaxies (viewed at g-band), and major axis intensity scans are even more irregular for the UDF than Kiso galaxies. Local clumpy galaxies in the Kiso survey appear to be intermediate between UDF clumpies and normal spirals.
Tadpole galaxies have a head-tail shape with a large clump of star formation at the head and a diffuse tail or streak of stars off to one side. We measured the head and tail masses, ages, surface brightnesses, and sizes for 66 tadpoles in the Hubble
Ultra Deep Field (UDF), and we looked at the distribution of neighbor densities and tadpole orientations with respect to neighbors. The heads have masses of 10^7-10^8 Msun and photometric ages of ~0.1 Gyr for z~2. The tails have slightly larger masses than the heads, and comparable or slightly older ages. The most obvious interpretation of tadpoles as young merger remnants is difficult to verify. They have no enhanced proximity to other resolved galaxies as a class, and the heads, typically less than 0.2 kpc in diameter, usually have no obvious double-core structure. Another possibility is ram pressure interaction between a gas-rich galaxy and a diffuse cosmological flow. Ram pressure can trigger star formation on one side of a galaxy disk, giving the tadpole shape when viewed edge-on. Ram pressure can also strip away gas from a galaxy and put it into a tail, which then forms new stars and gravitationally drags along old stars with it. Such an effect might have been observed already in the Virgo cluster. Another possibility is that tadpoles are edge-on disks with large, off-center clumps. Analogous lop-sided star formation in UDF clump clusters are shown.
Cluster formation and gas dynamics in the central regions of barred galaxies are not well understood. This paper reviews the environment of three 10^7 Msun clusters near the inner Lindblad resonance of the barred spiral NGC 1365. The morphology, mass
, and flow of HI and CO gas in the spiral and barred regions are examined for evidence of the location and mechanism of cluster formation. The accretion rate is compared with the star formation rate to infer the lifetime of the starburst. The gas appears to move from inside corotation in the spiral region to looping filaments in the interbar region at a rate of ~6 Msun/yr before impacting the bar dustlane somewhere along its length. The gas in this dustlane moves inward, growing in flux as a result of the accretion to ~40 Msun/yr near the ILR. This inner rate exceeds the current nuclear star formation rate by a factor of 4, suggesting continued buildup of nuclear mass for another ~0.5 Gyr. The bar may be only 1-2 Gyr old. Extrapolating the bar flow back in time, we infer that the clusters formed in the bar dustlane outside the central dust ring at a position where an interbar filament currently impacts the lane. The ram pressure from this impact is comparable to the pressure in the bar dustlane, and both are comparable to the pressure in the massive clusters. Impact triggering is suggested. The isothermal assumption in numerical simulations seems inappropriate for the rare fraction parts of spiral and bar gas flows. The clusters have enough lower-mass counterparts to suggest they are part of a normal power law mass distribution. Gas trapping in the most massive clusters could explain their [NeII] emission, which is not evident from the lower-mass clusters nearby.
Clump clusters and chain galaxies in the Hubble Ultra Deep Field are examined for bulges in the NICMOS images. Approximately 50% of the clump clusters and 30% of the chains have relatively red and massive clumps that could be young bulges. Magnitudes
and colors are determined for these bulge-like objects and for the bulges in spiral galaxies, and for all of the prominent star-formation clumps in these three galaxy types. The colors are fitted to population evolution models to determine the bulge and clump masses, ages, star-formation rate decay times, and extinctions. The results indicate that bulge-like objects in clump cluster and chain galaxies have similar ages and 2 to 5 times larger masses compared to the star-formation clumps, while the bulges in spirals have ~6 times larger ages and 20 to 30 times larger masses than the clumps. All systems appear to have an underlying red disk population. The masses of star-forming clumps are typically in a range from 10^7 to 10^8 Msun; their ages have a wide range around ~10^2 Myr. Ages and extinctions both decrease with redshift. Star formation is probably the result of gravitational instabilities in the disk gas, in which case the large clump mass in the UDF is the result of a high gas velocity dispersion, 30 km/s or more, combined with a high gas mass column density, ~100 Msun/pc^2. Because clump clusters and chains dominate disk galaxies beyond z~1, the observations suggest that these types represent an early phase in the formation of modern spiral galaxies, when the bulge and inner disk formed.
Gas-rich disks in the early universe are highly turbulent and have giant star-forming clumps. Models suggest the clumps form by gravitational instabilities, and if they resist disruption by star formation, then they interact, lose angular momentum, a
nd migrate to the center to form a bulge. Here we study the properties of the bulges formed by this mechanism. They are all thick, slowly rotating, and have a high Sersic index, like classical bulges. Their rapid formation should also give them relatively high alpha-element abundances. We consider fourteen low-resolution models and four high-resolution models, three of which have supernova feedback. All models have an active halo, stellar disk, and gaseous disk, three of the models have a pre-existing bulge and three others have a cuspy dark matter halo. All show the same basic result except the one with the highest feedback, in which the clumps are quickly destroyed and the disk thickens too much. The coalescence of massive disk clumps in the center of a galaxy is like a major merger in terms of orbital mixing. It differs by leaving a bulge with no specific dark matter component, unlike the merger of individual galaxies. Normal supernova feedback has little effect because the high turbulent speed in the gas produces tightly bound clumps. A variety of indirect observations support the model, including clumpy disks with young bulges at high redshift and bulges with relatively little dark matter.
Massive stellar clumps in high redshift galaxies interact and migrate to the center to form a bulge and exponential disk in <1 Gyr. Here we consider the fate of intermediate mass black holes (BHs) that might form by massive-star coalescence in the de
nse young clusters of these disk clumps. We find that the BHs move inward with the clumps and reach the inner few hundred parsecs in only a few orbit times. There they could merge into a supermassive BH by dynamical friction. The ratio of BH mass to stellar mass in the disk clumps is approximately preserved in the final ratio of BH to bulge mass. Because this ratio for individual clusters has been estimated to be ~10^{-3}, the observed BH-to-bulge mass ratio results. We also obtain a relation between BH mass and bulge velocity dispersion that is compatible with observations of present-day galaxies.
K_s-band images of 20 barred galaxies show an increase in the peak amplitude of the normalized m=2 Fourier component with the R_25-normalized radius at this peak. This implies that longer bars have higher $m=2$ amplitudes. The long bars also correlat
e with an increased density in the central parts of the disks, as measured by the luminosity inside 0.25R_25 divided by the cube of this radius in kpc. Because denser galaxies evolve faster, these correlations suggest that bars grow in length and amplitude over a Hubble time with the fastest evolution occurring in the densest galaxies. All but three of the sample have early-type flat bars; there is no clear correlation between the correlated quantities and the Hubble type.
