arXiv subjects

Jan Hamann

Publications and source records attributed to Jan Hamann.

At least 19 recordsLinked to original sources

arXiv2026

(The) Wiggles going non-linear

The simplest models of slow-roll inflation predict a featureless, nearly scale-invariant power spectrum of primordial curvature perturbations, consistent with current observations. However, in many non-minimal realisations of inflation, one generically expects the primordial power spectrum (PPS) to be ''wiggly'' with features that strongly deviate from scale-invariance. Current and next generation large scale structure (LSS) surveys will probe the PPS with unprecedented accuracy and therefore also increase sensitivity to power spectrum wiggles. However, accessing the information contained in these data will require an understanding of the behaviour of wiggly power spectra beyond the linear regime of structure formation. In this work, we use high resolution $N$-body simulations to study the non-linear evolution of scenarios in which the PPS has superimposed oscillations, calibrate a one-parameter semi-analytic damping model to describe their signatures in the late-time matter power spectrum and test the relative improvement on constraints by implementing this modelling strategy in a likelihood analysis via Gaussian Process Regression (GPR) emulation. Paying special attention to identifying our approach's domain of validity and quantifying as well as propagating uncertainties, we demonstrate that as long as the frequency of the PPS modulation is large enough, we are able to predict the matter power spectrum with sub-percent accuracy -- thereby enabling us to search for inflationary wiggles in LSS data.

astro-ph.CO↗

arXiv2026

TUNeS: Neural Emulation of Large-Scale Structure Across Redshifts

In this work, we introduce TUNeS (Temporal UNet emulator for Structure formation), a neural network framework for accelerating N-body simulations by predicting the nonlinear evolution of the matter density field from an initial particle distribution. TUNeS employs a two-stage modeling strategy, combining particle-based inference with a density-field refinement on a regular grid, enabling accurate reconstruction of both large- and small-scale structures. The model is designed to operate across redshift, taking particle snapshots at arbitrary input redshifts and predicting density fields at arbitrary target redshifts. In this work, we evaluate its performance using simulations initialized at $z=100$, with predictions generated at multiple lower redshifts. Trained on only eight N-body simulations, TUNeS reproduces reference results with good agreement in both Gaussian and non-Gaussian statistics, including two-point correlations, one-point distributions, peak counts, and three-dimensional Minkowski functionals. In particular, at $k \simeq 1\,h\,\mathrm{Mpc}^{-1}$, the power spectrum error remains at the few-percent level. End-to-end inference from $256^3$ particles to a $256^3$ density grid can be completed in $\sim25\,\mathrm{second}$ on a single GPU. Thanks to its architectural design, the model naturally scales to larger particle numbers and larger volumes through particle batching and window-based refinement.

astro-ph.CO↗

arXiv2026

Bayesian optimisation for Bayesian evidence (BOBE) -- a fast and efficient likelihood emulator for model selection

The formalism of Bayesian model selection provides a very elegant way of ranking different physical models in terms of how compatible they are with a given set of observed data. However, its practical application is often hampered by the challenge of having to compute the Bayesian evidence - a multi-dimensional integral over the product of likelihood and prior probability. This usually necessitates a large number of function calls to the likelihood, which may become prohibitive in case of "slow", costly to evaluate likelihoods. A possible solution to this problem lies in approximating the slow full likelihood by a fast emulated likelihood. In this paper, we introduce BOBE (Bayesian Optimisation for Bayesian Evidence), a method to construct a Gaussian Process Regression (GPR)-based emulator. BOBE utilises a Bayesian Optimisation algorithm designed specifically to (i) provide a realistic estimate of the emulator's uncertainty and its impact on the evidence calculation, and (ii) minimise the number of likelihood evaluations required in order to meet a given evidence accuracy goal. We apply it to a number of toy examples as well as actual cosmological likelihoods, and demonstrate that training the emulator to a sufficient accuracy takes a factor of $O(10^3)$ fewer direct likelihood evaluations than would be needed if one were to directly compute the evidence integral via nested sampling. BOBE's overhead is independent of the likelihood computation time $t_L$, making it particularly useful for "expensive" likelihoods with $t_L \gtrsim 1$~s. BOBE is written in Python, supports MPI parallelisation, takes advantage of automatic differentiation and just-in-time-compilation provided by JAX, can straightforwardly be implemented with cosmological data analysis frameworks such as Cobaya, and is available for download from https://github.com/Ameek94/BOBE.

astro-ph.CO↗

arXiv2024

Non-Gaussianity in CMB lensing from full-sky simulations

The lensing convergence field describing the weak lensing effect of the Cosmic Microwave Background (CMB) radiation is expected to be subject to mild deviations from Gaussianity. We perform a suite of full-sky lensing simulations using ray tracing through multiple lens planes - generated by combining $N$-body simulations on smaller scales and low-to-intermediate redshifts with realisations of Gaussian random fields on large scales and at high redshifts. We quantify the non-Gaussianity of the resulting convergence fields in terms of a set of skewness and kurtosis parameters and show that the non-Gaussian information in these maps can be used to constrain cosmological parameters such as the cold dark matter density $\Omega_\mathrm{c} h^2$ or the amplitude of primordial curvature perturbations $A_\mathrm{s}$. We forecast that for future CMB lensing observations, combining the non-Gaussian parameters with the Gaussian information can increase constraining power on $(\Omega_\mathrm{c} h^2, A_\mathrm{s})$ by a few per cent compared to constraints from Gaussian observables alone. We make the simulation code for the full-sky lensing simulation available for download from GitHub.

astro-ph.CO↗

arXiv2024

AB$\mathbb{C}$MB: Deep Delensing Assisted Likelihood-Free Inference from CMB Polarization Maps

The existence of a cosmic background of primordial gravitational waves (PGWB) is a robust prediction of inflationary cosmology, but it has so far evaded discovery. The most promising avenue of its detection is via measurements of Cosmic Microwave Background (CMB) $B$-polarization. However, this is not straightforward due to (a) the fact that CMB maps are distorted by gravitational lensing and (b) the high-dimensional nature of CMB data, which renders likelihood-based analysis methods computationally extremely expensive. In this paper, we introduce an efficient likelihood-free, end-to-end inference method to directly infer the posterior distribution of the tensor-to-scalar ratio $r$ from lensed maps of the Stokes $Q$ and $U$ polarization parameters. Our method employs a generative model to delense the maps and utilizes the Approximate Bayesian Computation (ABC) algorithm to sample $r$. We demonstrate that our method yields unbiased estimates of $r$ with well-calibrated uncertainty quantification.

astro-ph.CO↗

arXiv2023

A Minkowski Functional Analysis of the Cosmic Microwave Background Weak Lensing Convergence

Minkowski functionals are summary statistics that capture the geometric and morphological properties of fields. They are sensitive to all higher order correlations of the fields and can be used to complement more conventional statistics, such as the power spectrum of the field. We develop a Minkowski functional-based approach for a full likelihood analysis of mildly non-Gaussian sky maps with partial sky coverage. Applying this to the inference of cosmological parameters from the Planck mission's map of the Cosmic Microwave Background's lensing convergence, we find an excellent agreement with results from the power spectrum-based lensing likelihood. While the non-Gaussianity of current CMB lensing maps is dominated by reconstruction noise, a Minkowski functional-based analysis may be able to extract cosmological information from the non-Gaussianity of future lensing maps and thus go beyond what is accessible with a power spectrum-based analysis. We make the numerical code for the calculation of a map's Minkowski functionals, skewness and kurtosis parameters available for download from GitHub.

astro-ph.CO↗

arXiv2023

Removing the mask -- reconstructing a scalar field on the sphere from a masked field

The paper analyses a spectral approach to reconstructing a scalar field on the sphere, given only information about a masked version of the field together with precise information about the (smooth) mask. The theory is developed for a general mask, and later specialised to the case of an axially symmetric mask. Numerical experiments are given for the case of an axial mask motivated by the cosmic microwave background, assuming that the underlying field is a realisation of a Gaussian random field with an artificial angular power spectrum of moderate degree ($\ell \le 100$). The recovery is highly satisfactory in the absence of noise and even in the presence of moderate noise.

math.NA↗

arXiv2023

Revisiting Cosmological Constraints on Supersymmetric SuperWIMPs

SuperWIMPs are extremely weakly interacting massive particles that inherit their relic abundance from late decays of frozen-out parent particles. Within supersymmetric models, gravitinos and axinos represent two of the most well-motivated superWIMPs. In this paper we revisit constraints on these scenarios from a variety of cosmological observations that probe their production mechanisms as well as the superWIMP kinematic properties in the early Universe. We consider in particular observables of Big Bang Nucleosynthesis and the Cosmic Microwave Background (spectral distortion and anisotropies), which limit the fractional energy injection from the late decays, as well as warm and mixed dark matter constraints derived from the Lyman-$\alpha$ forest and other small-scale structure observables. We discuss complementary constraints from collider experiments, and argue that cosmological considerations rule out a significant part of the gravitino and the axino superWIMP parameter space.

hep-ph↗

arXiv2023

Enhancing Bispectrum Estimators for Galaxy Redshift Surveys with Velocities

We forecast the ability of bispectrum estimators to constrain primordial non-Gaussianity using future photometric galaxy redshift surveys. A full-sky survey with photometric redshift resolution of $\sigma_z/(1+z)=0.05$ in the redshift range $0.2<z<2$ can provide constraints $\sigma(f^\mathrm{local}_\mathrm{NL})=3.4$, $\sigma(f^\mathrm{equil}_\mathrm{NL})=15$, and $\sigma(f^\mathrm{orth}_\mathrm{NL})=17$ for the local, equilateral, and orthogonal shapes respectively, delivering constraints on primordial non-Gaussianities competitive to those from the cosmic microwave background. We generalize these results by deriving a scaling relation for the constraints on the amplitude of primordial non-Gaussianity as a function of redshift error, depth, sky coverage, and nonlinear scale cutoff. Finally, we investigate the impact that photometric calibration errors on the largest scales will have on the constraining power of future experiments. We show that peculiar velocities reconstructed via kinetic Sunyaev Zeldovich tomography can be used to mitigate the impact of calibration errors on primordial non-Gaussianity constraints.

astro-ph.CO↗

arXiv2022

Constraining primordial tensor features with the anisotropies of the Cosmic Microwave Background

It is commonly assumed that the stochastic background of gravitational waves on cosmological scales follows an almost scale-independent power spectrum, as generically predicted by the inflationary paradigm. However, it is not inconceivable that the spectrum could have strongly scale-dependent features, generated, e.g., via transient dynamics of spectator axion-gauge fields during inflation. Using the temperature and polarisation maps from the \textit{Planck} and BICEP/Keck datasets, we search for such features, taking the example of a log-normal bump in the primordial tensor spectrum at CMB scales. We do not find any evidence for the existence of bump-like tensor features at present, but demonstrate that future CMB experiments such as LiteBIRD and CMB-S4 will greatly improve our prospects of determining the amplitude, location and width of such a bump. We also highlight the role of delensing in constraining these features at angular scales $\ell\gtrsim 100$.

astro-ph.CO↗

arXiv2022

Approximate Equivariance SO(3) Needlet Convolution

This paper develops a rotation-invariant needlet convolution for rotation group SO(3) to distill multiscale information of spherical signals. The spherical needlet transform is generalized from $\mathbb{S}^2$ onto the SO(3) group, which decomposes a spherical signal to approximate and detailed spectral coefficients by a set of tight framelet operators. The spherical signal during the decomposition and reconstruction achieves rotation invariance. Based on needlet transforms, we form a Needlet approximate Equivariance Spherical CNN (NES) with multiple SO(3) needlet convolutional layers. The network establishes a powerful tool to extract geometric-invariant features of spherical signals. The model allows sufficient network scalability with multi-resolution representation. A robust signal embedding is learned with wavelet shrinkage activation function, which filters out redundant high-pass representation while maintaining approximate rotation invariance. The NES achieves state-of-the-art performance for quantum chemistry regression and Cosmic Microwave Background (CMB) delensing reconstruction, which shows great potential for solving scientific challenges with high-resolution and multi-scale spherical signal representation.

eess.IV↗

arXiv2021

Optimising Inflationary Features the Bayesian Way

Modern cosmological data demand modern data analysis techniques. We introduce BayOp, a new likelihood sampling and maximisation method which is based on the Bayesian Optimisation algorithm and learns a function instead of randomly sampling from it. We apply BayOp to analyse Planck data for traces of inflationary features models with global periodic modulations of the primordial power spectrum. While we do not find any new evidence for features, we demonstrate that BayOp provides an extremely efficient way of sampling likelihoods over low-to-moderate-dimensional parameter spaces, even for very complex likelihood landscapes.

astro-ph.CO↗

arXiv2020

CosmoVAE: Variational Autoencoder for CMB Image Inpainting

Cosmic microwave background radiation (CMB) is critical to the understanding of the early universe and precise estimation of cosmological constants. Due to the contamination of thermal dust noise in the galaxy, the CMB map that is an image on the two-dimensional sphere has missing observations, mainly concentrated on the equatorial region. The noise of the CMB map has a significant impact on the estimation precision for cosmological parameters. Inpainting the CMB map can effectively reduce the uncertainty of parametric estimation. In this paper, we propose a deep learning-based variational autoencoder --- CosmoVAE, to restoring the missing observations of the CMB map. The input and output of CosmoVAE are square images. To generate training, validation, and test data sets, we segment the full-sky CMB map into many small images by Cartesian projection. CosmoVAE assigns physical quantities to the parameters of the VAE network by using the angular power spectrum of the Gaussian random field as latent variables. CosmoVAE adopts a new loss function to improve the learning performance of the model, which consists of $\ell_1$ reconstruction loss, Kullback-Leibler divergence between the posterior distribution of encoder network and the prior distribution of latent variables, perceptual loss, and total-variation regularizer. The proposed model achieves state of the art performance for Planck \texttt{Commander} 2018 CMB map inpainting.

eess.IV↗

arXiv2019

CMB constraints on ultra-light primordial black holes with extended mass distributions

We examine the effects ultra-light primordial black holes (PBHs) have on the anisotropies of the cosmic microwave background (CMB). PBHs in the mass range of $10^{15}$ to $10^{17}$ g emit Hawking radiation in the early Universe, modifying the standard recombination history. This leads to a damping of small-scale temperature and polarisation anisotropies and enhances large-scale polarisation fluctuations. As some models of inflation predict PBHs with a range of masses, we investigate the impacts of extended mass distributions on PBH abundance constraints. We model PBH energy injection using a ground-up approach incorporating species-dependent deposition efficiencies. By allowing the $\Lambda$CDM parameters to vary simultaneously with the PBH fraction and mass, we show that exclusion bounds on the PBH fraction of DM $f_\text{PBH}$ are relaxed by up to an order of magnitude, compared to the case of fixed $\Lambda$CDM parameters. We also give 95% exclusion regions for $f_\text{PBH}$ for a variety of mass distributions. In particular, for a uniform mass distribution between $10^{15}$ and $10^{17}$ g, we find $f_\text{PBH} < 1.6 \times 10^{-5}$ when allowing $\Lambda$CDM parameters to vary.

astro-ph.CO↗

arXiv2017

A non-parametric consistency test of the $\Lambda$CDM model with Planck CMB data

Non-parametric reconstruction methods, such as Gaussian process (GP) regression, provide a model-independent way of estimating an underlying function and its uncertainty from noisy data. We demonstrate how GP-reconstruction can be used as a consistency test between a given data set and a specific model by looking for structures in the residuals of the data with respect to the model's best-fit. Applying this formalism to the Planck temperature and polarisation power spectrum measurements, we test their global consistency with the predictions of the base $\Lambda$CDM model. Our results do not show any serious inconsistencies, lending further support to the interpretation of the base $\Lambda$CDM model as cosmology's gold standard.

astro-ph.CO↗

arXiv2016

Precise measurements of inflationary features with 21 cm observations

Future observations of 21~cm emission using HI intensity mapping will enable us to probe the large scale structure of the Universe over very large survey volumes within a reasonable observation time. We demonstrate that the three-dimensional information contained in such surveys will be an extremely powerful tool in searching for features that were imprinted in the primordial power spectrum and bispectrum during inflation. Here we focus on the "resonant" and "step" inflation models, and forecast the potential of upcoming 21~cm experiments to detect these inflationary features in the observable power- and bispectrum. We find that the full scale Tianlai experiment and the Square Kilometre Array (SKA) have the potential to improve on the sensitivity of current Cosmic Microwave Background (CMB) experiments by several orders of magnitude.

astro-ph.CO↗

arXiv2015

Features and New Physical Scales in Primordial Observables: Theory and Observation

All cosmological observations to date are consistent with adiabatic, Gaussian and nearly scale invariant initial conditions. These findings provide strong evidence for a particular symmetry breaking pattern in the very early universe (with a close to vanishing order parameter, $ε$), widely accepted as conforming to the predictions of the simplest realizations of the inflationary paradigm. However, given that our observations are only privy to perturbations, in inferring something about the background that gave rise to them, it should be clear that many different underlying constructions project onto the same set of cosmological observables. Features in the primordial correlation functions, if present, would offer a unique and discriminating window onto the parent theory in which the mechanism that generated the initial conditions is embedded. In certain contexts, simple linear response theory allows us to infer new characteristic scales from the presence of features that can break the aforementioned degeneracies among different background models, and in some cases can even offer a limited spectroscopy of the heavier degrees of freedom that couple to the inflaton. In this review, we offer a pedagogical survey of the diverse, theoretically well grounded mechanisms which can imprint features into primordial correlation functions in addition to reviewing the techniques one can employ to probe observations. These observations include cosmic microwave background anisotropies and spectral distortions as well as the matter two and three point functions as inferred from large-scale structure and potentially, 21 cm surveys.

astro-ph.CO↗

arXiv2015

Future cosmological sensitivity for hot dark matter axions

We study the potential of a future, large-volume photometric survey to constrain the axion mass $m_a$ in the hot dark matter limit. Future surveys such as Euclid will have significantly more constraining power than current observations for hot dark matter. Nonetheless, the lowest accessible axion masses are limited by the fact that axions lighter than $\sim 0.15$ eV decouple before the QCD epoch, assumed here to occur at a temperature $T_{\rm QCD} \sim 170$ MeV; this leaves an axion population of such low density that its late-time cosmological impact is negligible. For larger axion masses, $m_a \gtrsim 0.15$ eV, where axions remain in equilibrium until after the QCD phase transition, we find that a Euclid-like survey combined with Planck CMB data can detect $m_a$ at very high significance. Our conclusions are robust against assumptions about prior knowledge of the neutrino mass. Given that the proposed IAXO solar axion search is sensitive to $m_a\lesssim 0.2$ eV, the axion mass range probed by cosmology is nicely complementary.

astro-ph.CO↗