arXiv subjects

Benjamin D. Wandelt

Publications and source records attributed to Benjamin D. Wandelt.

At least 145 records · Page 8Linked to original sources

arXiv2013

Probabilistic image reconstruction for radio interferometers

We present a novel, general-purpose method for deconvolving and denoising images from gridded radio interferometric visibilities using Bayesian inference based on a Gaussian process model. The method automatically takes into account incomplete coverage of the uv-plane, signal mode coupling due to the primary beam, and noise mode coupling due to uv sampling. Our method uses Gibbs sampling to efficiently explore the full posterior distribution of the underlying signal image given the data. We use a set of widely diverse mock images with a realistic interferometer setup and level of noise to assess the method. Compared to results from a proxy for point source- based CLEAN method we find that in terms of RMS error and signal-to-noise ratio our approach performs better than traditional deconvolution techniques, regardless of the structure of the source image in our test suite. Our implementation scales as O(np log np), provides full statistical and uncertainty information of the reconstructed image, requires no supervision, and provides a robust, consistent framework for incorporating noise and parameter marginalizations and foreground removal.

astro-ph.IM↗

arXiv2013

A response to arXiv:1310.2791: A self-consistent public catalogue of voids and superclusters in the SDSS Data Release 7 galaxy surveys

Recently, Nadathur & Hotchkiss (2013) submitted a paper discussing a new cosmic void catalog. This paper includes claims about the void catalog described in Sutter et al. (2012). In this note, we respond to those claims, clarify some discrepancies between the text of Sutter et al. (2012) and the most recent version of the catalog, and provide some comments on the differences between our catalog and that of Nadathur & Hotchkiss (2013). All updates and documentation for our catalog are available at http://www.cosmicvoids.net.

astro-ph.CO↗

arXiv2013

Searching for Oscillations in the Primordial Power Spectrum: Perturbative Approach (Paper I)

In this first of two papers, we present a new method for searching for oscillatory features in the primordial power spectrum. A wide variety of models predict these features in one of two different flavors: logarithmically spaced oscillations and linearly spaced oscillations. The proposed method treats the oscillations as perturbations on top of the scale-invariant power spectrum, allowing us to vary all cosmological parameters. This perturbative approach reduces the computational requirements for the search as the transfer functions and their derivatives can be precomputed. We show that the most significant degeneracy in the analysis is between the distance to last scattering and the overall amplitude at low frequencies. For models with logarithmic oscillations, this degeneracy leads to an uncertainty in the phase. For linear spaced oscillations, it affects the frequency of the oscillations. In this first of two papers, we test our code on simulated Planck-like data, and show we are able to recover fiducial input oscillations with an amplitude of a few times order 10^{-2}. We apply the code to WMAP9-year data and confirm the existence of two intriguing resonant frequencies for log spaced oscillations. For linear spaced oscillations we find a single resonance peak. We use numerical simulations to assess the significance of these features and conclude that the data do not provide compelling evidence for the existence of oscillatory features in the primordial spectrum.

astro-ph.CO↗

arXiv2013

Methods for Bayesian power spectrum inference with galaxy surveys

We derive and implement a full Bayesian large scale structure inference method aiming at precision recovery of the cosmological power spectrum from galaxy redshift surveys. Our approach improves over previous Bayesian methods by performing a joint inference of the three dimensional density field, the cosmological power spectrum, luminosity dependent galaxy biases and corresponding normalizations. We account for all joint and correlated uncertainties between all inferred quantities. Classes of galaxies with different biases are treated as separate sub samples. The method therefore also allows the combined analysis of more than one galaxy survey. In particular, it solves the problem of inferring the power spectrum from galaxy surveys with non-trivial survey geometries by exploring the joint posterior distribution with efficient implementations of multiple block Markov chain and Hybrid Monte Carlo methods. Our Markov sampler achieves high statistical efficiency in low signal to noise regimes by using a deterministic reversible jump algorithm. We test our method on an artificial mock galaxy survey, emulating characteristic features of the Sloan Digital Sky Survey data release 7, such as its survey geometry and luminosity dependent biases. These tests demonstrate the numerical feasibility of our large scale Bayesian inference frame work when the parameter space has millions of dimensions. The method reveals and correctly treats the anti-correlation between bias amplitudes and power spectrum, which are not taken into account in current approaches to power spectrum estimation, a 20 percent effect across large ranges in k-space. In addition, the method results in constrained realizations of density fields obtained without assuming the power spectrum or bias parameters in advance.

astro-ph.CO↗

arXiv2013

Maximum likelihood analysis of systematic errors in interferometric observations of the cosmic microwave background

We investigate the impact of instrumental systematic errors in interferometric measurements of the cosmic microwave background (CMB) temperature and polarization power spectra. We simulate interferometric CMB observations to generate mock visibilities and estimate power spectra using the statistically optimal maximum likelihood technique. We define a quadratic error measure to determine allowable levels of systematic error that do not induce power spectrum errors beyond a given tolerance. As an example, in this study we focus on differential pointing errors. The effects of other systematics can be simulated by this pipeline in a straightforward manner. We find that, in order to accurately recover the underlying B-modes for r=0.01 at 28<l<384, Gaussian-distributed pointing errors must be controlled to 0.7^\circ rms for an interferometer with an antenna configuration similar to QUBIC, in agreement with analytical estimates. Only the statistical uncertainty for 28<l<88 would be changed at ~10% level. With the same instrumental configuration, we find the pointing errors would slightly bias the 2-σupper limit of the tensor-to-scalar ratio r by ~10%. We also show that the impact of pointing errors on the TB and EB measurements is negligibly small.

astro-ph.CO↗

arXiv2013

Systematic Effects in Interferometric Observations of the CMB Polarization

The detection of the primordial $B$-mode spectrum of the polarized cosmic microwave background (CMB) signal may provide a probe of inflation. However, observation of such a faint signal requires excellent control of systematic errors. Interferometry proves to be a promising approach for overcoming such a challenge. In this paper we present a complete simulation pipeline of interferometric observations of CMB polarization, including systematic errors. We employ two different methods for obtaining the power spectra from mock data produced by simulated observations: the maximum likelihood method and the method of Gibbs sampling. We show that the results from both methods are consistent with each other, as well as, within a factor of 6, with analytical estimates. Several categories of systematic errors are considered: instrumental errors, consisting of antenna gain and antenna coupling errors, and beam errors, consisting of antenna pointing errors, beam cross-polarization and beam shape (and size) errors. In order to recover the tensor-to-scalar ratio, $r$, within a 10% tolerance level, which ensures the experiment is sensitive enough to detect the $B$-signal at $r=0.01$ in the multipole range $28 < \ell < 384$, we find that, for a QUBIC-like experiment, Gaussian-distributed systematic errors must be controlled with precisions of $|g_{rms}| = 0.1$ for antenna gain, $|ε_{rms}| = 5 \times 10^{-4}$ for antenna coupling, $δ_{rms} \approx 0.7^\circ$ for pointing, $ζ_{rms} \approx 0.7^\circ$ for beam shape, and $μ_{rms} = 5 \times 10^{-4}$ for beam cross-polarization.

astro-ph.CO↗

arXiv2013

Bayesian Inference of Polarized CMB Power Spectra from Interferometric Data

Detection of B-mode polarization of the cosmic microwave background (CMB) radiation is one of the frontiers of observational cosmology. Because they are an order of magnitude fainter than E-modes, it is quite a challenge to detect B-modes. Having more manageable systematics, interferometers prove to have a substantial advantage over imagers in detecting such faint signals. Here, we present a method for Bayesian inference of power spectra and signal reconstruction from interferometric data of the CMB polarization signal by using the technique of Gibbs sampling. We demonstrate the validity of the method in the flat-sky approximation for a simulation of an interferometric observation on a finite patch with incomplete uv-plane coverage, a finite beam size and a realistic noise model. With a computational complexity of O(n^{3/2}), n being the data size, Gibbs sampling provides an efficient method for analyzing upcoming cosmology observations.

astro-ph.CO↗

arXiv2013

Fast Wiener filtering of CMB maps

We present the application of a new method to compute the Wiener filter solution of large and complex data sets. Contrary to the iterative solvers usually employed in signal processing, our algorithm does not require the use of preconditioners to be computationally efficient. The new scheme is conceptually very simple and therefore easy to implement, numerically absolutely stable, and guaranteed to converge. We introduce a messenger field to mediate between the different preferred bases in which signal and noise properties can be specified most conveniently, and rephrase the signal reconstruction problem in terms of this auxiliary variable. We demonstrate the capabilities of the algorithm by applying it to cosmic microwave background (CMB) radiation data obtained by the WMAP satellite.

astro-ph.CO↗

arXiv2012

A first application of the Alcock-Paczynski test to stacked cosmic voids

We report on the first application of the Alcock-Paczynski test to stacked voids in spectroscopic galaxy redshift surveys.We use voids from the Sutter et al. (2012) void catalog, which was derived from the Sloan Digital Sky Survey Data Release 7 main sample and luminous red galaxy catalogs. The construction of that void catalog removes potential shape measurement bias by using a modified version of the ZOBOV algorithm and by removing voids near survey boundaries and masks. We apply the shape-fitting procedure presented in Lavaux & Wandelt (2012) to ten void stacks out to redshift z=0.36. Combining these measurements, we determine the mean cosmologically induced "stretch" of voids in three redshift bins, with 1-sigma errors of 5-15%. The mean stretch is consistent with unity, providing no indication of a distortion induced by peculiar velocities. While the statistical errors are too large to detect the Alcock-Paczynski effect over our limited redshift range, this proof-of-concept analysis defines procedures that can be applied to larger spectroscopic galaxy surveys at higher redshifts to constrain dark energy using the expected statistical isotropy of structures that are minimally affected by uncertainties in galaxy velocity bias.

astro-ph.CO↗

arXiv2012

Efficient Wiener filtering without preconditioning

We present a new approach to calculate the Wiener filter solution of general data sets. It is trivial to implement, flexible, numerically absolutely stable, and guaranteed to converge. Most importantly, it does not require an ingenious choice of preconditioner to work well. The method is capable of taking into account inhomogeneous noise distributions and arbitrary mask geometries. It iteratively builds up the signal reconstruction by means of a messenger field, introduced to mediate between the different preferred bases in which signal and noise properties can be specified most conveniently. Using cosmic microwave background (CMB) radiation data as a showcase, we demonstrate the capabilities of our scheme by computing Wiener filtered WMAP7 temperature and polarization maps at full resolution for the first time. We show how the algorithm can be modified to synthesize fluctuation maps, which, combined with the Wiener filter solution, result in unbiased constrained signal realizations, consistent with the observations. The algorithm performs well even on simulated CMB maps with Planck resolution and dynamic range.

astro-ph.CO↗

arXiv2012

Accelerating convolutions on the sphere with hybrid GPU/CPU kernel splitting

We present a general method for accelerating by more than an order of magnitude the convolution of pixelated function on the sphere with a radially-symmetric kernel. Our method splits the kernel into a compact real-space, and a compact spherical harmonic space component that can then be convolved in parallel using an inexpensive commodity GPU and a CPU, respectively. We provide models for the computational cost of both real-space and Fourier space convolutions and an estimate for the approximation error. Using these models we can determine the optimum split that minimizes the wall clock time for the convolution while satisfying the desired error bounds. We apply this technique to the problem of simulating a cosmic microwave background sky map at the resolution typical of the high resolution maps of the cosmic microwave background anisotropies produced by the Planck space craft. For the main Planck CMB science channels we achieve a speedup of over a factor of ten, assuming an acceptable fractional rms error of order 10^-5 in the (power spectrum of the) output map.

astro-ph.IM↗

arXiv2012

Robust, data-driven inference in non-linear cosmostatistics

We discuss two projects in non-linear cosmostatistics applicable to very large surveys of galaxies. The first is a Bayesian reconstruction of galaxy redshifts and their number density distribution from approximate, photometric redshift data. The second focuses on cosmic voids and uses them to construct cosmic spheres that allow reconstructing the expansion history of the Universe using the Alcock-Paczynski test. In both cases we find that non-linearities enable the methods or enhance the results: non-linear gravitational evolution creates voids and our photo-z reconstruction works best in the highest density (and hence most non-linear) portions of our simulations.

astro-ph.CO↗

arXiv2012

A public void catalog from the SDSS DR7 Galaxy Redshift Surveys based on the watershed transform

We produce the most comprehensive public void catalog to date using the Sloan Digital Sky Survey Data Release 7 main sample out to redshift z=0.2 and the luminous red galaxy sample out to z=0.44. Using a modified version of the parameter-free void finder ZOBOV, we fully take into account the presence of the survey boundary and masks. Our strategy for finding voids is thus appropriate for any survey configuration. We produce two distinct catalogs: a complete catalog including voids near any masks, which would be appropriate for void galaxy surveys, and a bias-free catalog of voids away from any masks, which is necessary for analyses that require a fair sampling of void shapes and alignments. Our discovered voids have effective radii from 5 to 135 h^-1 Mpc. We discuss basic catalog statistics such as number counts and redshift distributions and describe some additional data products derived from our catalog, such as radial density profiles and projected density maps. We find that radial profiles of stacked voids show a qualitatively similar behavior across nearly two decades of void radii and throughout the full redshift range.

astro-ph.CO↗

arXiv2012

Bayesian angular power spectrum analysis of interferometric data

We present a Bayesian angular power spectrum and signal map inference engine which can be adapted to interferometric observations of anisotropies inthe cosmic microwave background, 21 cm emission line mapping of galactic brightness fluctuations, or 21 cm absorption line mapping of neutral hydrogen in the dark ages. The method uses Gibbs sampling to generate a sampled representation of the angular power spectrum posterior and the posterior of signal maps given a set of measured visibilities in the uv-plane. We use a mock interferometric CMB observation to demonstrate the validity of this method in the flat-sky approximation when adapted to take into account arbitrary coverage of the uv-plane, mode-mode correlations due to observations on a finite patch, and heteroschedastic visibility errors. The computational requirements scale as O(n_p log n_p) where n_p measures the ratio of the size of the detector array to the inter-detector spacing, meaning that Gibbs sampling is a promising technique for meeting the data analysis requirements of future cosmology missions.

astro-ph.CO↗

arXiv2012

Precision cosmography with stacked voids

We present a purely geometrical method for probing the expansion history of the Universe from the observation of the shape of stacked voids in spectroscopic redshift surveys. Our method is an Alcock-Paczynski (AP) test based on the average sphericity of voids posited on the local isotropy of the Universe. It works by comparing the temporal extent of cosmic voids along the line of sight with their angular, spatial extent. We describe the algorithm that we use to detect and stack voids in redshift shells on the light cone and test it on mock light cones produced from N-body simulations. We establish a robust statistical model for estimating the average stretching of voids in redshift space and quantify the contamination by peculiar velocities. Finally, assuming that the void statistics that we derive from N-body simulations is preserved when considering galaxy surveys, we assess the capability of this approach to constrain dark energy parameters. We report this assessment in terms of the figure of merit (FoM) of the dark energy task force and in particular of the proposed EUCLID mission which is particularly suited for this technique since it is a spectroscopic survey. The FoM due to stacked voids from the EUCLID wide survey may double that of all other dark energy probes derived from EUCLID data alone (combined with Planck priors). In particular, voids seem to outperform Baryon Acoustic Oscillations by an order of magnitude. This result is consistent with simple estimates based on mode-counting. The AP test based on stacked voids may be a significant addition to the portfolio of major dark energy probes and its potentialities must be studied in detail.

astro-ph.CO↗

arXiv2012

Likelihood, Fisher information, and systematics of cosmic microwave background experiments

Every experiment is affected by systematic effects that hamper the data analysis and have the potential to ultimately degrade its performance. In the case of probes of the cosmic microwave background (CMB) radiation, a minimal set of issues to consider includes asymmetric beam functions, correlated noise, and incomplete sky coverage. Presuming a simplified scanning strategy that allows for an exact analytical treatment of the problem, we study the impact of systematic effects on the likelihood function of the CMB power spectrum. We use the Fisher matrix, a measure of the information content of a data set, for a quantitative comparison of different experimental configurations. In addition, for various power spectrum coefficients, we explore the functional form of the likelihood directly, and obtain the following results: The likelihood function can deviate systematically from a Gaussian distribution up to the highest multipole values considered in our analysis. Treated exactly, realistic levels of asymmetric beam functions and correlated noise do not by themselves decrease the information yield of CMB experiments nor do they induce noticeable coupling between multipoles. Masking large fractions of the sky, on the other hand, results in a considerably more complex correlation structure of the likelihood function. Combining adjacent power spectrum coefficients into bins can partially mitigate these problems.

astro-ph.CO↗

arXiv2012

Bayesian physical reconstruction of initial conditions from large scale structure surveys

We present a fully probabilistic, physical model of the non-linearly evolved density field, as probed by realistic galaxy surveys. Our model is valid in the linear and mildly non-linear regimes and uses second order Lagrangian perturbation theory to connect the initial conditions with the final density field. Our parameter space consists of the 3D initial density field and our method allows a fully Bayesian exploration of the sets of initial conditions that are consistent with the galaxy distribution sampling the final density field. A natural byproduct of this technique is an optimal non-linear reconstruction of the present density and velocity fields, including a full propagation of the observational uncertainties. A test of these methods on simulated data mimicking the survey mask, selection function and galaxy number of the SDSS DR7 main sample shows that this physical model gives accurate reconstructions of the underlying present-day density and velocity fields on scales larger than ~6 Mpc/h. Our method naturally and accurately reconstructs non-linear features corresponding to three-point and higher order correlation functions such as walls and filaments. Simple tests of the reconstructed initial conditions show statistical consistency with the Gaussian simulation inputs. Our test demonstrates that statistical approaches based on physical models of the large scale structure distribution are now becoming feasible for realistic current and future surveys.

astro-ph.CO↗

arXiv2012

Fast calculation of the Fisher matrix for cosmic microwave background experiments

The Fisher information matrix of the cosmic microwave background (CMB) radiation power spectrum coefficients is a fundamental quantity that specifies the information content of a CMB experiment. In the most general case, its exact calculation scales with the third power of the number of data points N and is therefore computationally prohibitive for state-of-the-art surveys. Applicable to a very large class of CMB experiments without special symmetries, we show how to compute the Fisher matrix in only O(N^2 log N) operations as long as the inverse noise covariance matrix can be applied to a data vector in time O(l_max^3 log l_max). This assumption is true to a good approximation for all CMB data sets taken so far. The method takes into account common systematics such as arbitrary sky coverage and realistic noise correlations. As a consequence, optimal quadratic power spectrum estimation also becomes feasible in O(N^2 log N) operations for this large group of experiments. We discuss the relevance of our findings to other areas of cosmology where optimal power spectrum estimation plays a role.

astro-ph.CO↗