arXiv subjects
Publications and source records attributed to Carlos Frenk.
The main ingredients of recent semi-analytic models of galaxy formation are summarised. We present predictions for the galaxy clustering properties of a well specified LCDM model whose parameters are constrained by observed local galaxy properties. We present preliminary predictions for evolution of clustering that can be probed with deep pencil beam surveys.
We present the results for the predicted density and peculiar velocity fields and the dipole from the PSCz survey of 15,000 IRAS galaxies over 84% of the sky. We find a significant component to the dipole arising between 6000 and 15,000 km/s, but no significant component from greater distances. The misalignment with the CMB is 20 degrees. The most remarkable feature of the PSCz model velocity field is a coherent large-scale flow along the baseline connecting Perseus-Pisces, the Local Supercluster, Great Attractor and the Shapley Concentration. We have measured the parameter beta using the amplitude of the dipole, bulk flow and point by point comparisons between the individual velocities of galaxies in the MarkIII and SFI datasets, and the large-scale clustering distortion in redshift space.All our results are consistent with beta = 0.6 +- 0.1.
We present the catalogue and redshift data for the PSCz survey of 15,000 IRAS galaxies over 84% of the sky. The uniformity, completeness and data quality are assessed, and guidelines and caveats for its use are given.
We perform a detailed comparison of the IRAS PSCz and 1.2-Jy spherical harmonic coefficients of the density and velocity fields in redshift space. The monopole terms predicted from the two surveys show some differences. The mismatch between the velocity monopoles arises from faint sources and disappears when extracting a PSCz subsample of galaxies with fluxes larger than 1.2 Jy. The analysis of PSCz dipole moments confirms the same inconsistencies found by Davis, Nusser and Willick (1996) between the IRAS 1.2-Jy gravity field and MARK III peculiar velocities. We conclude that shot-noise, which is greatly reduced in our PSCz gravity field, cannot be responsible for the observed mismatch.
Semi-analytical models of galaxy formation based on hierarchical clustering now make a wide range of predictions for observable properties of galaxies at low and high redshift. This article concentrates on 2 aspects: (1) Self-consistent modelling of dust absorption predicts a mean UV extinction A_{UV} ~ 1 mag, depending only weakly on redshift, and similar to observational estimates. (2) The models predict that the Lyman-break galaxies found at z ~ 3 should be strongly clustered with a comoving correlation length r_0 = 4-7 Mpc/h (depending on the cosmology), in good agreement with subsequent observational determinations.
I describe a semi-analytical model for the formation and evolution of galaxies in hierarchical clustering models, and its predictions for the properties of the galaxy population at high z. The predictions are found to agree well with the observed properties of the Lyman break galaxies found at z sim 3 by Steidel et al. The models predict that the star formation rate per comoving volume should have peaked at z sim 1-2, which also agrees well with recent observational data.
We present a revised and expanded catalog of satellite galaxies of a set of isolated spiral galaxies similar in luminosity to the Milky Way. This sample of 115 satellites, 69 of which were discovered in our multifiber redshift survey, is used to further probe the results obtained from the original sample (Zaritsky et al. 1993). The satellites are, by definition, at projected separations <~ 500 kpc, have absolute recessional velocity differences with respect to the parent spiral of less than 500 km/s and are at least 2.2 mag fainter than their associated primary galaxy. A key characteristic of this survey is the strict isolation of these systems, which simplifies any dynamical analysis. We find no evidence for a decrease in the velocity dispersion of the satellite system as a function of radius out to galactocentric radii of 400 kpc, suggesting that the halo extends well beyond 200 kpc. Furthermore, the new sample affirms our previous conclusions (Zaritsky et al. 1993) that (1) the velocity difference between a satellite and its primary is not strongly correlated with the rotation speed of the primary, (2) the system of satellites has a slight net rotation (34 \pm 14 km\s) in the same sense as the primary's disk, and (3) that the halo mass of an ~ L* spiral galaxy is in excess of 2 x 10^{12} solar masses.
We investigate the clustering of galaxies in the QDOT redshift survey of IRAS galaxies. We find that the redshift space two-point correlation function is well approximated by a power-law of slope -1.11+\-0.09, with clustering length 3.87+\-0.32 Mpc/h out to pair separations of 25 Mpc\h. On scales larger than 40 Mpc/h, the correlation function is consistent with zero. The r.m.s. fluctuation in the count of QDOT galaxies above the Poisson level in spheres of radius 8 Mpc/h is σ_8^{IRAS}=0.58+\-0.14, showing that fluctuations in the distribution of IRAS galaxies on these scales are smaller than those of optical galaxies by a factor of about 0.65, with an uncertainty of 25%. We find no detectable difference between the correlation functions measured in redshift space and in real space, leading to a $2σ$ limit of b_{IRAS}/Ω^{0.6} >1.05, where b_{IRAS} is the bias factor for IRAS galaxies and Ωis the cosmological density parameter. The QDOT autocorrelation function calculated in concentric shells increases significantly with shell radius. This difference is more likely due to sampling fluctuations than to an increase of the clustering strength with galaxy luminosity, but the two effects are difficult to disentangle; our data allow at most an increase of \sim 20% in clustering strength for each decade in luminosity. For separations greater than \sim 3 Mpc/h, the cross-correlation function of Abell clusters (with Richness R\ge 1) and QDOT galaxies is well approximated by a power-law of slope -1.81+\-0.10, with clustering length 10.10+\-0.45 Mpc/h, and no significant signal beyond \gsim 50 Mpc\h. This cross-correlation depends only weakly on cluster richness.