arXiv subjects

S. L. Bridle

Publications and source records attributed to S. L. Bridle.

77 records · Page 5Linked to original sources

arXiv1999

Bayesian `Hyper-Parameters' Approach to Joint Estimation: The Hubble Constant from CMB Measurements

Recently several studies have jointly analysed data from different cosmological probes with the motivation of estimating cosmological parameters. Here we generalise this procedure to take into account the relative weights of various probes. This is done by including in the joint χ^2 function a set of `Hyper-Parameters', which are dealt with using Bayesian considerations. The resulting algorithm (in the case of uniform priors on the log of the Hyper-Parameters) is very simple: instead of minimising \sum χ_j^2 (where χ_j^2 is per data set j) we propose to minimise \sum N_j \ln (χ_j^2) (where N_j is the number of data points per data set j). We illustrate the method by estimating the Hubble constant H_0 from different sets of recent CMB experiments (including Saskatoon, Python V, MSAM1, TOCO and Boomerang).

astro-ph↗

arXiv1999

Cosmological parameters from cluster abundances, CMB and IRAS

We combine information on cosmological parameters from cluster abundances, CMB primordial anisotropies and the IRAS 1.2 Jy galaxy redshift survey. We take as free parameters the present values of the total matter density of the universe, Omega_m, the Hubble parameter, h, sigma_8, and the IRAS biasing factor, b_IRAS. We assume that the universe is spatially flat, with a cosmological constant, and that structure formed from adiabatic initial fluctuations with a Harrison-Zel'dovich power spectrum. The nucleosynthesis value Omega_b=0.019/h^2 is adopted. We use the full three- and four- dimensional likelihood functions for each data set and marginalise these to two- and one- dimensional distributions in a Bayesian way, integrating over the other parameters. It is shown that the three data sets are in excellent agreement, with a best fit point of Omega_m= 1-Omega_Lambda=0.36, h=0.54, sigma_8=0.74, and b_IRAS=1.08. This point is within one sigma of the minimum for each data set alone. Pairs of these data sets have their degeneracies in sufficiently different directions that using only two data sets at a time is sufficient to place good constraints on the cosmological parameters. We show that the results from each of the three possible pairings of the data are also in good agreement. Finally, we combine all three data sets to obtain marginalised 68 per cent confidence intervals of 0.30<Omega_m<0.43, 0.48<h<0.59, 0.69<sigma_8<0.79, and 1.01<b_IRAS<1.16. For the best fit parameters the CMB quadrupole is Q_rms-ps=18.0 uK, Gamma=0.15, Omega_b=0.066 and the age of the universe is 16.7 Gyr.

astro-ph↗

arXiv1999

Combining Supernovae and LSS Information with the CMB

Observations of the Cosmic Microwave Background (CMB), large scale structure (LSS) and standard candles such as Type 1a Supernovae (SN) each place different constraints on the values of cosmological parameters. We assume an inflationary Cold Dark Matter model with a cosmological constant, in which the initial density perturbations in the universe are adiabatic. We discuss the parameter degeneracies inherent in interpreting CMB or SN data, and derive their orthogonal nature. We then present our preliminary results of combining CMB and SN likelihood functions. The results of combining the CMB and IRAS 1.2 Jy survey information are given, with marginalised confidence regions in the H_0, Omega_m, b_IRAS and Q_rms-ps directions assuming n=1, Omega_Lambda+Omega_m=1 and Omega_b h^2=0.024. Finally we combine all three likelihood functions and find that the three data sets are consistent and suitably orthogonal, leading to tight constraints on H_0, Omega_m, b_IRAS and Q_rms-ps, given our assumptions.

astro-ph↗

arXiv1998

Joint estimation of cosmological parameters from CMB and IRAS data

Observations of large scale structure (LSS) and the Cosmic Microwave Background (CMB) each place separate constraints on the values of cosmological parameters. We calculate a joint likelihood based on various CMB experiments and the IRAS 1.2Jy galaxy redshift survey and use this to find an overall optimum with respect to the free parameters. Our formulation self-consistently takes account of the underlying mass distribution, which affects both the CMB potential fluctuations and the IRAS redshift distortion. This not only allows more accurate parameter estimation, but also removes the parameter degeneracy which handicaps calculations based on either approach alone. The family of Cold Dark Matter (CDM) models analysed corresponds to a spatially-flat universe with an initially scale-invariant spectrum and a cosmological constant. Free parameters in the joint model are the mass density due to all matter (Omega_m), Hubble's parameter (h = H_0 / 100km/s Mpc^-1), the quadrupole normalisation of the CMB power spectrum (Q) in uK, and the IRAS light-to-mass bias (b_i). Throughout the analysis, the baryonic density (Omega_b) is required is to satisfy the nucleosynthesis constraint Omega_b h^2 = 0.024. Results from the two data sets show good agreement, and the joint optimum lies at Omega_m=0.39, h=0.53, Q=16.96uK, and b_i=1.21. The 68 per cent confidence intervals are: 0.29<Omega_m<0.53, 0.39<h<0.58, 15.34<Q<17.60, and 0.98<b_i<1.56. For the above parameters the normalisation and shape of the mass power-spectrum are sigma_8=0.67 and Gamma=0.15, and the age of the Universe is 16.5Gyr.

astro-ph↗

arXiv1998

A maximum-entropy method for reconstructing the projected mass distribution of gravitational lenses

The maximum-entropy method is applied to the problem of reconstructing the projected mass density of a galaxy cluster using its gravitational lensing effects on background galaxies. We demonstrate the method by reconstructing the mass distribution in a model cluster using simulated shear and magnification data to which Gaussian noise is added. The mass distribution is reconstructed directly and the inversion is regularised using the entropic prior for this positive additive distribution. For realistic noise levels, we find that the method faithfully reproduces the main features of the cluster mass distribution not only within the observed field but also slightly beyond it. We estimate the uncertainties on the reconstruction by calculating an analytic approximation to the covariance matrix of the reconstruction values of each pixel. This result is compared with error estimates derived from Monte-Carlo simulations for different noise realisations and found to be in good agreement.

astro-ph↗