arXiv subjects

M. P. Hobson

Publications and source records attributed to M. P. Hobson.

At least 37 records · Page 2Linked to original sources

arXiv2018

Kinetic Initial Conditions for Inflation

We consider the classical evolution of the inflaton field $ϕ(t)$ and the Hubble parameter $H(t)$ in homogeneous and isotropic single-field inflation models. Under an extremely broad assumption, we show that the universe generically emerges from an initial singularity in a non-inflating state where the kinetic energy of the inflaton dominates its potential energy, $\dotϕ^2 \gg V(ϕ)$. In this kinetically-dominated regime, the dynamical equations admit simple analytic solutions for $ϕ(t)$ and $H(t)$, which are independent of the form of $V(ϕ)$. In such models, these analytic solutions thus provide a simple way of setting the initial conditions from which to start the (usually numerical) integration of the coupled equations of motion for $ϕ(t)$ and $H(t)$. We illustrate this procedure by applying it to spatially-flat models with polynomial and exponential potentials, and determine the background evolution in each case; generically $H(t)$ and $|ϕ(t)|$ as well as their time derivatives decrease during kinetic dominance until $\dotϕ^2\sim V(ϕ)$, marking the onset of a brief period of fast-roll inflation prior to a slow roll phase. We also calculate the approximate spectrum of scalar perturbations produced in each model and show that it exhibits a generic damping of power on large scales. This may be relevant to the apparent low-$\ell$ falloff in the CMB power spectrum.

astro-ph.CO↗

arXiv2017

Constraining the dark energy equation of state using Bayes theorem and the Kullback-Leibler divergence

Data-driven model-independent reconstructions of the dark energy equation of state $w(z)$ are presented using Planck 2015 era CMB, BAO, SNIa and Lyman-$α$ data. These reconstructions identify the $w(z)$ behaviour supported by the data and show a bifurcation of the equation of state posterior in the range $1.5{<}z{<}3$. Although the concordance $Λ$CDM model is consistent with the data at all redshifts in one of the bifurcated spaces, in the other a supernegative equation of state (also known as `phantom dark energy') is identified within the $1.5 σ$ confidence intervals of the posterior distribution. To identify the power of different datasets in constraining the dark energy equation of state, we use a novel formulation of the Kullback--Leibler divergence. This formalism quantifies the information the data add when moving from priors to posteriors for each possible dataset combination. The SNIa and BAO datasets are shown to provide much more constraining power in comparison to the Lyman-$α$ datasets. Further, SNIa and BAO constrain most strongly around redshift range $0.1-0.5$, whilst the Lyman-$α$ data constrains weakly over a broader range. We do not attribute the supernegative favouring to any particular dataset, and note that the $Λ$CDM model was favoured at more than $2$ log-units in Bayes factors over all the models tested despite the weakly preferred $w(z)$ structure in the data.

astro-ph.CO↗

arXiv2016

Wide-band Profile Domain Pulsar Timing Analysis

We extend profile domain pulsar timing to incorporate wide-band effects such as frequency-dependent profile evolution and broadband shape variation in the pulse profile. We also incorporate models for temporal variations in both pulse width and in the separation in phase of the main pulse and interpulse. We perform the analysis with both nested sampling and Hamiltonian Monte Carlo methods. In the latter case we introduce a new parameterisation of the posterior that is extremely efficient in the low signal-to-noise regime and can be readily applied to a wide range of scientific problems. We apply this methodology to a series of simulations, and to between seven and nine yr of observations for PSRs J1713$+$0747, J1744$-$1134, and J1909$-$3744 with frequency coverage that spans 700-3600MHz. We use a smooth model for profile evolution across the full frequency range, and compare smooth and piecewise models for the temporal variations in DM. We find the profile domain framework consistently results in improved timing precision compared to the standard analysis paradigm by as much as 40% for timing parameters. Incorporating smoothness in the DM variations into the model further improves timing precision by as much as 30%. For PSR J1713+0747 we also detect pulse shape variation uncorrelated between epochs, which we attribute to variation intrinsic to the pulsar at a level consistent with previously published analyses. Not accounting for this shape variation biases the measured arrival times at the level of $\sim$30ns, the same order of magnitude as the expected shift due to gravitational-waves in the pulsar timing band.

astro-ph.IM↗

arXiv2016

The Runge-Kutta-Wentzel-Kramers-Brillouin Method

We demonstrate the effectiveness of a novel scheme for numerically solving linear differential equations whose solutions exhibit extreme oscillation. We take a standard Runge-Kutta approach, but replace the Taylor expansion formula with a Wentzel-Kramers-Brillouin method. The method is demonstrated by application to the Airy equation, along with a more complicated burst-oscillation case. Finally, we compare our scheme to existing approaches.

physics.comp-ph↗

arXiv2016

Novel quantum initial conditions for inflation

We present a novel approach for setting initial conditions on the mode functions of the Mukhanov Sazaki equation. These conditions are motivated by minimisation of the renormalised stress-energy tensor, and are valid for setting a vacuum state even in a context where the spacetime is changing rapidly. Moreover, these alternative conditions are potentially observationally distinguishable. We apply this to the kinetically dominated universe, and compare with the more traditional approach.

gr-qc↗

arXiv2016

Bayesian analysis of cosmic-ray propagation: evidence against homogeneous diffusion

We present the results of the most complete ever scan of the parameter space for cosmic ray (CR) injection and propagation. We perform a Bayesian search of the main GALPROP parameters, using the MultiNest nested sampling algorithm, augmented by the BAMBI neural network machine learning package. This is the first such study to separate out low-mass isotopes ($p$, $\bar p$ and He) from the usual light elements (Be, B, C, N, O). We find that the propagation parameters that best fit $p$, $\bar p$, He data are significantly different from those that fit light elements, including the B/C and $^{10}$Be/$^9$Be secondary-to-primary ratios normally used to calibrate propagation parameters. This suggests each set of species is probing a very different interstellar medium, and that the standard approach of calibrating propagation parameters using B/C can lead to incorrect results. We present posterior distributions and best fit parameters for propagation of both sets of nuclei, as well as for the injection abundances of elements from H to Si. The input GALDEF files with these new parameters will be included in an upcoming public GALPROP update.

astro-ph.HE↗

arXiv2015

Comparison of Sunyaev-Zel'dovich measurements from Planck and from the Arcminute Microkelvin Imager for 99 galaxy clusters

We present observations and analysis of a sample of 123 galaxy clusters from the 2013 Planck catalogue of Sunyaev-Zel'dovich sources with the Arcminute Microkelvin Imager (AMI), a ground-based radio interferometer. AMI provides an independent measurement with higher angular resolution, 3 arcmin compared to the Planck beams of 5-10 arcmin. The AMI observations thus provide validation of the cluster detections, improved positional estimates, and a consistency check on the fitted 'size' ($θ_{s}$) and 'flux' ($Y_{\rm tot}$) parameters in the Generalised Navarro, Frenk and White (GNFW) model. We detect 99 of the clusters. We use the AMI positional estimates to check the positional estimates and error-bars produced by the Planck algorithms PowellSnakes and MMF3. We find that $Y_{\rm tot}$ values as measured by AMI are biased downwards with respect to the Planck constraints, especially for high Planck-SNR clusters. We perform simulations to show that this can be explained by deviation from the 'universal' pressure profile shape used to model the clusters. We show that AMI data can constrain the $α$ and $β$ parameters describing the shape of the profile in the GNFW model for individual clusters provided careful attention is paid to the degeneracies between parameters, but one requires information on a wider range of angular scales than are present in AMI data alone to correctly constrain all parameters simultaneously.

astro-ph.CO↗

arXiv2015

PolyChord: next-generation nested sampling

PolyChord is a novel nested sampling algorithm tailored for high-dimensional parameter spaces. This paper coincides with the release of PolyChord v1.3, and provides an extensive account of the algorithm. PolyChord utilises slice sampling at each iteration to sample within the hard likelihood constraint of nested sampling. It can identify and evolve separate modes of a posterior semi-independently, and is parallelised using openMPI. It is capable of exploiting a hierarchy of parameter speeds such as those present in CosmoMC and CAMB, and is now in use in the CosmoChord and ModeChord codes. PolyChord is available for download at: http://ccpforge.cse.rl.ac.uk/gf/project/polychord/

astro-ph.IM↗

arXiv2015

PolyChord: nested sampling for cosmology

PolyChord is a novel nested sampling algorithm tailored for high dimensional parameter spaces. In addition, it can fully exploit a hierarchy of parameter speeds such as is found in CosmoMC and CAMB. It utilises slice sampling at each iteration to sample within the hard likelihood constraint of nested sampling. It can identify and evolve separate modes of a posterior semi-independently and is parallelised using openMPI. PolyChord is available for download at: http://ccpforge.cse.rl.ac.uk/gf/project/polychord/

astro-ph.CO↗

arXiv2015

Generative pulsar timing analysis

A new Bayesian method for the analysis of folded pulsar timing data is presented that allows for the simultaneous evaluation of evolution in the pulse profile in either frequency or time, along with the timing model and additional stochastic processes such as red spin noise, or dispersion measure variations. We model the pulse profiles using `shapelets' - a complete ortho-normal set of basis functions that allow us to recreate any physical profile shape. Any evolution in the profiles can then be described as either an arbitrary number of independent profiles, or using some functional form. We perform simulations to compare this approach with established methods for pulsar timing analysis, and to demonstrate model selection between different evolutionary scenarios using the Bayesian evidence. %s The simplicity of our method allows for many possible extensions, such as including models for correlated noise in the pulse profile, or broadening of the pulse profiles due to scattering. As such, while it is a marked departure from standard pulsar timing analysis methods, it has clear applications for both new and current datasets, such as those from the European Pulsar Timing Array (EPTA) and International Pulsar Timing Array (IPTA).

astro-ph.IM↗

arXiv2014

Bayes-X: a Bayesian inference tool for the analysis of X-ray observations of galaxy clusters

We present the first public release of our Bayesian inference tool, Bayes-X, for the analysis of X-ray observations of galaxy clusters. We illustrate the use of Bayes-X by analysing a set of four simulated clusters at z=0.2-0.9 as they would be observed by a Chandra-like X-ray observatory. In both the simulations and the analysis pipeline we assume that the dark matter density follows a spherically-symmetric Navarro, Frenk and White (NFW) profile and that the gas pressure is described by a generalised NFW (GNFW) profile. We then perform four sets of analyses. By numerically exploring the joint probability distribution of the cluster parameters given simulated Chandra-like data, we show that the model and analysis technique can robustly return the simulated cluster input quantities, constrain the cluster physical parameters and reveal the degeneracies among the model parameters and cluster physical parameters. We then analyse Chandra data on the nearby cluster, A262, and derive the cluster physical profiles. To illustrate the performance of the Bayesian model selection, we also carried out analyses assuming an Einasto profile for the matter density and calculated the Bayes factor. The results of the model selection analyses for the simulated data favour the NFW model as expected. However, we find that the Einasto profile is preferred in the analysis of A262. The Bayes-X software, which is implemented in Fortran 90, is available at http://www.mrao.cam.ac.uk/facilities/software/bayesx/.

astro-ph.CO↗

arXiv2014

Generalised Fisher Matrices

The Fisher Information Matrix formalism is extended to cases where the data is divided into two parts (X,Y), where the expectation value of Y depends on X according to some theoretical model, and X and Y both have errors with arbitrary covariance. In the simplest case, (X,Y) represent data pairs of abscissa and ordinate, in which case the analysis deals with the case of data pairs with errors in both coordinates, but X can be any measured quantities on which Y depends. The analysis applies for arbitrary covariance, provided all errors are gaussian, and provided the errors in X are small, both in comparison with the scale over which the expected signal Y changes, and with the width of the prior distribution. This generalises the Fisher Matrix approach, which normally only considers errors in the `ordinate' Y. In this work, we include errors in X by marginalising over latent variables, effectively employing a Bayesian hierarchical model, and deriving the Fisher Matrix for this more general case. The methods here also extend to likelihood surfaces which are not gaussian in the parameter space, and so techniques such as DALI (Derivative Approximation for Likelihoods) can be generalised straightforwardly to include arbitrary gaussian data error covariances. For simple mock data and theoretical models, we compare to Markov Chain Monte Carlo experiments, illustrating the method with cosmological supernova data. We also include the new method in the Fisher4Cast software.

astro-ph.CO↗

arXiv2014

Testing the mutual consistency of different supernovae surveys

It is now common practice to constrain cosmological parameters using supernovae (SNe) catalogues constructed from several different surveys. Before performing such a joint analysis, however, one should check that parameter constraints derived from the individual SNe surveys that make up the catalogue are mutually consistent. We describe a statistically-robust mutual consistency test, which we calibrate using simulations, and apply it to each pairwise combination of the surveys making up, respectively, the UNION2 catalogue and the very recent JLA compilation by Betoule et al. We find no inconsistencies in the latter case, but conclusive evidence for inconsistency between some survey pairs in the UNION2 catalogue.

astro-ph.IM↗

arXiv2013

Constraints on the Tensor-to-Scalar ratio for non-power-law models

Recent cosmological observations hint at a deviation from the simple power-law form of the primordial spectrum of curvature perturbations. In this paper we show that in the presence of a tensor component, a turn-over in the initial spectrum is preferred by current observations, and hence non-power-law models ought to be considered. For instance, for a power-law parameterisation with both a tensor component and running parameter, current data show a preference for a negative running at more than $2.5σ$ C.L. As a consequence of this deviation from a power-law, constraints on the tensor-to-scalar ratio $r$ are slightly broader. We also present constraints on the inflationary parameters for a model-independent reconstruction and the Lasenby & Doran (LD) model. In particular, the constraints on the tensor-to-scalar ratio from the LD model are: $r_{\rm LD}=0.11\pm{0.024}$. In addition to current data, we show expected constraints from Planck-like and CMB-Pol sensitivity experiments by using Markov-Chain-Monte-Carlo sampling chains. For all the models, we have included the Bayesian Evidence to perform a model selection analysis. The Bayes factor, using current observations, shows a strong preference for the LD model over the standard power-law parameterisation, and provides an insight into the accuracy of differentiating models through future surveys.

astro-ph.CO↗

arXiv2013

Exploring local fNL estimators based on the binned bispectrum

We explore different estimators of the local non-linear coupling parameter, fNL, based on the binned bispectrum presented in Bucher et al. Using simulations of Wilkinson Microwave Anisotropy Probe (WMAP)-7yr data, we compare the performance of a regression neural network with a χ^2-minimization and study the dependence of the results on the presence of the linear term in the analysis and on the use of inpainting for masked regions. Both methods obtain similar results and are robust to the use of inpainting, but the neural network estimator converges considerably faster. We also examine the performance of a simplified χ^2 estimator that assumes a diagonal matrix and has the linear term subtracted, which considerably reduces the computational time; in this case inpainting is found to be crucial. The estimators are also applied to real WMAP-7yr data, yielding constraints at 95% confidence level of -3< fNL <83.

astro-ph.CO↗

arXiv2013

Bayesian constraints on dark matter halo properties using gravitationally-lensed supernovae

A hierarchical Bayesian method is applied to the analysis of Type-Ia supernovae (SNIa) observations to constrain the properties of the dark matter haloes of galaxies along the SNIa lines-of-sight via their gravitational lensing effect. The full joint posterior distribution of the dark matter halo parameters is explored using the nested sampling algorithm {\sc MultiNest}, which also efficiently calculates the Bayesian evidence, thereby facilitating robust model comparison. We first demonstrate the capabilities of the method by applying it to realistic simulated SNIa data, based on the real 3-year data release from the Supernova Legacy Survey (SNLS3). Assuming typical values for the halo parameters in our simulations, we find that a catalogue analogous to the existing SNLS3 data set is incapable of detecting the lensing signal, but a catalogue containing approximately three times as many SNIa does produce robust and accurate parameter constraints and model selection results for two halo models: a truncated singular isothermal sphere (SIS) and a Navarro--Frenk--White (NFW) profile, thereby validating our analysis methodology. In the analysis of the real SNLS3 data, contrary to previous studies, we obtain only a very marginal detection of a lensing signal and weak constraints on the halo parameters for the truncated SIS model, although these constraints are tighter than those obtained from the equivalent simulated SNIa data set. This difference is driven by a preferred value of $η\approx 1$ in the assumed scaling-law $σ\propto L^η$ between velocity dispersion and luminosity, which is somewhat higher than the canonical values of $η= \tfrac{1}{4}$ and $η= \tfrac{1}{3}$ for early and late-type galaxies, respectively, and leads to a stronger lensing effect by the halo. No detection of a lensing signal is made for the NFW model.

astro-ph.CO↗

arXiv2012

A simple and robust method for automated photometric classification of supernovae using neural networks

A method is presented for automated photometric classification of supernovae (SNe) as Type-Ia or non-Ia. A two-step approach is adopted in which: (i) the SN lightcurve flux measurements in each observing filter are fitted separately; and (ii) the fitted function parameters and their associated uncertainties, along with the number of flux measurements, the maximum-likelihood value of the fit and Bayesian evidence for the model, are used as the input feature vector to a classification neural network (NN) that outputs the probability that the SN under consideration is of Type-Ia. The method is trained and tested using data released following the SuperNova Photometric Classification Challenge (SNPCC). We consider several random divisions of the data into training and testing sets: for instance, for our sample D_1 (D_4), a total of 10% (40%) of the data are involved in training the algorithm and the remainder used for blind testing of the resulting classifier; we make no selection cuts. Assigning a canonical threshold probability of p_th=0.5 on the NN output to classify a SN as Type-Ia, for the sample D_1 (D_4) we obtain a completeness of 0.78 (0.82), purity of 0.77 (0.82), and SNPCC figure-of-merit of 0.41 (0.50). Including the SN host-galaxy redshift and its uncertainty as additional inputs to the NN results in a modest 5-10% increase in these values. We find that the classification quality does not vary significantly with SN redshift. Moreover, our probabilistic classification method allows one to calculate the expected completeness, purity or other measures of classification quality as a function of the p_th, without knowing the true classes of the SNe in the testing sample. The method may thus be improved further by optimising p_th and can easily be extended to divide non-Ia SNe into their different classes.

astro-ph.CO↗

arXiv2012

Reconstruction of the Dark Energy equation of state

One of the main challenges of modern cosmology is to investigate the nature of dark energy in our Universe. The properties of such a component are normally summarised as a perfect fluid with a (potentially) time-dependent equation-of-state parameter $w(z)$. We investigate the evolution of this parameter with redshift by performing a Bayesian analysis of current cosmological observations. We model the temporal evolution as piecewise linear in redshift between `nodes', whose $w$-values and redshifts are allowed to vary. The optimal number of nodes is chosen by the Bayesian evidence. In this way, we can both determine the complexity supported by current data and locate any features present in $w(z)$. We compare this node-based reconstruction with some previously well-studied parameterisations: the Chevallier-Polarski-Linder (CPL), the Jassal-Bagla-Padmanabhan (JBP) and the Felice-Nesseris-Tsujikawa (FNT). By comparing the Bayesian evidence for all of these models we find an indication towards possible time-dependence in the dark energy equation-of-state. It is also worth noting that the CPL and JBP models are strongly disfavoured, whilst the FNT is just significantly disfavoured, when compared to a simple cosmological constant $w=-1$. We find that our node-based reconstruction model is slightly disfavoured with respect to the $Λ$CDM model.

astro-ph.CO↗