ISM Masses and Star Formation at z = 1 to 6 ALMA Observations of Dust Continuum in 180 Galaxies in COSMOS

ALMA Cycle 2 observations of the long wavelength dust emission in 180 star-forming (SF) galaxies are used to investigate the evolution of ISM masses at z = 1 to 6.4. The ISM masses exhibit strong increases from z = 0 to $rm <z>$ = 1.15 and further to $rm <z>$ = 2.2 and 4.8, particularly amongst galaxies above the SF galaxy main sequence (MS). The galaxies with highest SFRs at $rm <z>$ = 2.2 and 4.8 have gas masses 100 times that of the Milky Way and gas mass fractions reaching 50 to 80%, i.e. gas masses 1 - 4$times$ their stellar masses. For the full sample of galaxies, we find a single, very simple SF law: $rm SFR propto M_{rm ISM}^{0.9}$, i.e. a `linear dependence on the ISM mass -- on and above the MS. Thus, the galaxies above the MS are converting their larger ISM masses into stars on a timescale similar to those on the MS. At z $> 1$, the entire population of star-forming galaxies has $sim$5 - 10$times$ shorter gas depletion times ($sim0.2$ Gyr) than galaxies at low redshift. These {bf shorter depletion times are due to a different, dominant mode of SF in the early universe} -- dynamically driven by compressive, high dispersion gas motions and/or galaxy interactions. The dispersive gas motions are a natural consequence of the extraordinarily high gas accretion rates which must occur to maintain the prodigious SF.

A Classical Morphological Analysis of Galaxies in the Spitzer Survey of Stellar Structure in Galaxies (S4G)

The Spitzer Survey of Stellar Structure in Galaxies (S4G) is the largest available database of deep, homogeneous middle-infrared (mid-IR) images of galaxies of all types. The survey, which includes 2352 nearby galaxies, reveals galaxy morphology only minimally affected by interstellar extinction. This paper presents an atlas and classifications of S4G galaxies in the Comprehensive de Vaucouleurs revised Hubble-Sandage (CVRHS) system. The CVRHS system follows the precepts of classical de Vaucouleurs (1959) morphology, modified to include recognition of other features such as inner, outer, and nuclear lenses, nuclear rings, bars, and disks, spheroidal galaxies, X patterns and box/peanut structures, OLR subclass outer rings and pseudorings, bar ansae and barlenses, parallel sequence late-types, thick disks, and embedded disks in 3D early-type systems. We show that our CVRHS classifications are internally consistent, and that nearly half of the S4G sample consists of extreme late-type systems (mostly bulgeless, pure disk galaxies) in the range Scd-Im. The most common family classification for mid-IR types S0/a to Sc is SA while that for types Scd to Sm is SB. The bars in these two type domains are very different in mid-IR structure and morphology. This paper examines the bar, ring, and type classification fractions in the sample, and also includes several montages of images highlighting the various kinds of stellar structures seen in mid-IR galaxy morphology.

On the Markovian assumption in the excursion set approach: the approximation of Markov Velocities

The excursion set approach uses the statistics of the density field smoothed on a wide range of scales, to gain insight into a number of interesting processes in nonlinear structure formation, such as cluster assembly, merging and clustering. The approach treats the curve defined by the height of the overdensity fluctuation field when changing the smoothing scale as a random walk. The steps of the walks are often assumed to be uncorrelated, so that the walk heights are a Markov process, even though this assumption is known to be inaccurate for physically relevant filters. We develop a model in which the walk steps, rather than heights, are a Markov process, and correlations between steps arise because of nearest neighbour interactions. This model is a particular case of a general class, which we call Markov Velocity models. We show how these can approximate the walks generated by arbitrary power spectra and filters, and, unlike walks with Markov heights, provide a very good approximation to physically relevant models. We define a Markov Velocity Monte Carlo algorithm to generate walks whose first crossing distribution is very similar to that of TopHat-smoothed LCDM walks. Finally, we demonstrate that Markov Velocity walks generically exhibit a simple but realistic form of assembly bias, so we expect them to be useful in the construction of more realistic merger history trees.

Stochasticity in halo formation and the excursion set approach

The simplest stochastic halo formation models assume that the traceless part of the shear field acts to increase the initial overdensity (or decrease the underdensity) that a protohalo (or protovoid) must have if it is to form by the present time. Equivalently, it is the difference between the overdensity and (the square root of the) shear that must be larger than a threshold value. To estimate the effect this has on halo abundances using the excursion set approach, we must solve for the first crossing distribution of a barrier of constant height by the random walks associated with the difference, which is now (even for Gaussian initial conditions) a non-Gaussian variate. The correlation properties of such non-Gaussian walks are inherited from those of the density and the shear, and, since they are independent processes, the solution is in fact remarkably simple. We show that this provides an easy way to understand why earlier heuristic arguments about the nature of the solution worked so well. In addition to modelling halos and voids, this potentially simplifies models of the abundance and spatial distribution of filaments and sheets in the cosmic web.

The Evolution of ISM Mass Probed by Dust Emission -- ALMA Observations at z = 0.3 to 2

The use of submm dust continuum emission to probe the mass of interstellar dust and gas in galaxies is empirically calibrated using samples of local star forming galaxies, Planck observations of the Milky Way and high redshift submm galaxies (SMGs). All of these objects suggest a similar calibration, strongly supporting the view that the Rayleigh-Jeans (RJ) tail of the dust emission can be used as an accurate and very fast probe of the ISM in galaxies. We present ALMA Cycle 0 observations of the Band 7 (350 GHz) dust emission in 107 galaxies from z = 0.2 to 2.5. Three samples of galaxies with a total of 101 galaxies were stellar mass-selected from COSMOS to have $M_* simeq10^{11}$msun: 37 at z$sim0.4$, 33 at z$sim0.9$ and 31 at z$=2$. A fourth sample with 6 IR luminous galaxies at z = 2 was observed for comparison with the purely mass-selected samples. From the fluxes detected in the stacked images for each sample, we find that the ISM content has decreased a factor $sim 6$ from $1 - 2 times 10^{10}$msun at both z = 2 and 0.9 down to $sim 2 times 10^9$msun at z = 0.4. The IR luminous sample at z = 2 shows a further $sim 4$ times increase in M$_{ISM}$ compared to the equivalent non-IR bright sample at the same redshift. The gas mass fractions are $sim 2pm0.5, 12pm3, 14pm2 ~rm{and} ~53pm3$ $%$ for the four subsamples (z = 0.4, 0.9, 2 and IR bright galaxies).

The importance of stepping up in the excursion set approach

Recently, we provided a simple but accurate formula which closely approximates the first crossing distribution associated with random walks having correlated steps. The approximation is accurate for the wide range of barrier shapes of current interest and is based on the requirement that, in addition to having the right height, the walk must cross the barrier going upwards. Therefore, it only requires knowledge of the bivariate distribution of the walk height and slope, and is particularly useful for excursion set models of the massive end of the halo mass function. However, it diverges at lower masses. We show how to cure this divergence by using a formulation which requires knowledge of just one other variable. While our analysis is general, we use examples based on Gaussian initial conditions to illustrate our results. Our formulation, which is simple and fast, yields excellent agreement with the considerably more computationally expensive Monte-Carlo solution of the first crossing distribution, for a wide variety of moving barriers, even at very low masses.

The excursion set approach in non-Gaussian random fields

Insight into a number of interesting questions in cosmology can be obtained from the first crossing distributions of physically motivated barriers by random walks with correlated steps. We write the first crossing distribution as a formal series, ordered by the number of times a walk upcrosses the barrier. Since the fraction of walks with many upcrossings is negligible if the walk has not taken many steps, the leading order term in this series is the most relevant for understanding the massive objects of most interest in cosmology. This first term only requires knowledge of the bivariate distribution of the walk height and slope, and provides an excellent approximation to the first crossing distribution for all barriers and smoothing filters of current interest. We show that this simplicity survives when extending the approach to the case of non-Gaussian random fields. For non-Gaussian fields which are obtained by deterministic transformations of a Gaussian, the first crossing distribution is simply related to that for Gaussian walks crossing a suitably rescaled barrier. Our analysis shows that this is a useful way to think of the generic case as well. Although our study is motivated by the possibility that the primordial fluctuation field was non-Gaussian, our results are general. In particular, they do not assume the non-Gaussianity is small, so they may be viewed as the solution to an excursion set analysis of the late-time, nonlinear fluctuation field rather than the initial one. They are also useful for models in which the barrier height is determined by quantities other than the initial density, since most other physically motivated variables (such as the shear) are usually stochastic and non-Gaussian. We use the Lognormal transformation to illustrate some of our arguments.

The impact of bars on disk breaks as probed by S4G imaging

We have analyzed the radial distribution of old stars in a sample of 218 nearby face-on disks, using deep 3.6um images from the Spitzer Survey of Stellar Structure in Galaxies (S4G). In particular, we have studied the structural properties of those disks with a broken or down-bending profile. We find that, on average, disks with a genuine single exponential profile have a scale-length and a central surface brightness which are intermediate to those of the inner and outer components of a down-bending disk with the same total stellar mass. In the case of barred galaxies, the ratio between the break and the bar radii (Rbr/Rbar) depends strongly on the total stellar mass of the galaxy. For galaxies more massive than 10^10 Msun, the distribution is bimodal, peaking at Rbr/Rbar~2 and ~3.5. The first peak, which is the most populated one, is linked to the Outer Lindblad Resonance of the bar, whereas the second one is consistent with a dynamical coupling between the bar and the spiral pattern. For galaxies below 10^10 Msun, breaks are found up to ~10 Rbar, but we show that they could still be caused by resonances given the rising nature of rotation curves in these low-mass disks. While not ruling out star formation thresholds, our results imply that radial stellar migration induced by non-axysymmetric features can be responsible not only for those breaks at 2 Rbar, but also for many of those found at larger radii.

Peak-peak correlations in the cosmic background radiation from cosmic strings

We examine the two-point correlation function of local maxima in temperature fluctuations at the last scattering surface when this stochastic field is modified by the additional fluctuations produced by straight cosmic strings via the Kaiser-Stebbins effect. We demonstrate that one can detect the imprint of cosmic strings with tension $Gmu gtrsim 1.2 times 10^{-8}$ on noiseless $1^prime$ resolution cosmic microwave background (CMB) maps at 95% confidence interval. Including the effects of foregrounds and anticipated systematic errors increases the lower bound to $Gmu gtrsim 9.0times 10^{-8}$ at $2sigma$ confidence level. Smearing by beams of order 4 degrades the bound further to $Gmu gtrsim 1.6 times 10^{-7}$. Our results indicate that two-point statistics are more powerful than 1-point statistics (e.g. number counts) for identifying the non-Gaussianity in the CMB due to straight cosmic strings.