Galaxy formation and evolution

Abstract:

We use 3035 Herschel-SPIRE 500$\mu$m sources from 20.3 sq deg of sky in the HerMES Lockman, ES1 and XMM-LSS areas to estimate the star-formation rate density at z = 1-6. 500 mu sources are associated first with 350 and 250 mu sources, and then with Spitzer 24 mu sources from the SWIRE photometric redshift catalogue. The infrared and submillimetre data are fitted with a set of radiative-transfer templates corresponding to cirrus (quiescent) and starburst galaxies. Lensing candidates are removed via a set of colour-colour and colour-redshift constraints. Star-formation rates are found to extend from < 1 to 20,000 Mo/yr. Such high values were also seen in the all-sky IRAS Faint Source Survey. Star-formation rate functions are derived in a series of redshift bins from 0-6, combined with earlier far-infrared estimates, where available, and fitted with a Saunders et al (1990) functional form. The star-formation-rate density as a function of redshift is derived and compared with other estimates. There is reasonable agreement with both infrared and ultraviolet estimates for z < 3, but we find higher star-formation-rate densities than ultraviolet estimates at z = 3-6. Given the considerable uncertainties in the submillimetre estimates, we can not rule out the possibility that the ultraviolet estimates are correct. But the possibility that the ultraviolet estimates have seriously underestimated the contribution of dust-shrouded star-formation can also not be excluded.

Abstract:

© 2016. The American Astronomical Society. All rights reserved.. The dynamical formation of stellar-mass black hole-black hole binaries has long been a promising source of gravitational waves for the Laser Interferometer Gravitational-Wave Observatory (LIGO). Mass segregation, gravitational focusing, and multibody dynamical interactions naturally increase the interaction rate between the most massive black holes in dense stellar systems, eventually leading them to merge. We find that dynamical interactions, particularly three-body binary formation, enhance the merger rate of black hole binaries with total mass M tot roughly as ∝Mtotβ, with β ≳ 4. We find that this relation holds mostly independently of the initial mass function, but the exact value depends on the degree of mass segregation. The detection rate of such massive black hole binaries is only further enhanced by LIGO's greater sensitivity to massive black hole binaries with M tot ≲ 80 . We find that for power-law BH mass functions dN/dM ∝ M -α with α ≤ 2, LIGO is most likely to detect black hole binaries with a mass twice that of the maximum initial black hole mass and a mass ratio near one. Repeated mergers of black holes inside the cluster result in about ∼5% of mergers being observed between two and three times the maximum initial black hole mass. Using these relations, one may be able to invert the observed distribution to the initial mass function with multiple detections of merging black hole binaries.

Abstract:

The next generation of cosmology experiments will be required to use photometric redshifts rather than spectroscopic redshifts. Obtaining accurate and well-characterized photometric redshift distributions is therefore critical for Euclid, the Large Synoptic Survey Telescope and the Square Kilometre Array. However, determining accurate variance predictions alongside single point estimates is crucial, as they can be used to optimize the sample of galaxies for the specific experiment (e.g. weak lensing, baryon acoustic oscillations, supernovae), trading off between completeness and reliability in the galaxy sample. The various sources of uncertainty in measurements of the photometry and redshifts put a lower bound on the accuracy that any model can hope to achieve. The intrinsic uncertainty associated with estimates is often non-uniform and input-dependent, commonly known in statistics as heteroscedastic noise. However, existing approaches are susceptible to outliers and do not take into account variance induced by non-uniform data density and in most cases require manual tuning of many parameters. In this paper, we present a Bayesian machine learning approach that jointly optimizes the model with respect to both the predictive mean and variance we refer to as Gaussian processes for photometric redshifts (GPz). The predictive variance of the model takes into account both the variance due to data density and photometric noise. Using the SDSS DR12 data, we show that our approach substantially outperforms other machine learning methods for photo-z estimation and their associated variance, such as tpz and annz2. We provide a matlab and python implementations that are available to download at https://github.com/OxfordML/GPz.