arXiv subjects

Erwan Allys

Publications and source records attributed to Erwan Allys.

At least 19 recordsLinked to original sources

arXiv2026

Revealing Galactic dust beneath the cosmic infrared background anisotropies with Wavelet Phase Harmonics

Separating Galactic dust emission from cosmic infrared background (CIB) anisotropies is challenging because both the emission components have similar spectral properties. The primary objective of this work is to develop a component-separation algorithm to extract the Galactic dust emission from CIB anisotropies dominated sky regions using Wavelet Phase Harmonics statistics (WPH). We apply our algorithm to the \textit{Planck} $353\,\mathrm{GHz}$ frequency band over three sky regions spanning a range of neutral hydrogen ($\text{HI}$) column densities at high Galactic latitudes. In all three regions, we successfully recover the dust signal without significant leakage between the dust and CIB anisotropy components. We use the low signal-to-noise ($\mathrm{S/N}$) region to validate the component-separation algorithm and compare its performance with the standard template-fit approach that uses the linear dust-$\text{HI}$ correlation relation. For intermediate and high $\mathrm{S/N}$ regions, we separate the dust signal map from the contamination while preserving the statistical correlation with $\text{HI}$ column density map. Dust maps are used to analyze the relationship between dust and $\text{HI}$ emissions. We characterize the spatial variations in dust emission relative to the hydrogen column density as a function of the angular scale.

astro-ph.GA↗

arXiv2026

The Multi-phase HI of the Milky Way and Nearby Galaxies

Atomic hydrogen (HI) is the dominant baryonic component of the interstellar medium (ISM) in Milky Way-like galaxies and the reservoir from which molecular clouds and stars ultimately form. The condensation of diffuse HI into cold structures is governed by a complex interplay between radiative cooling, turbulence, magnetic fields, stellar feedback, and galactic dynamics, acting over scales ranging from astronomical units to kiloparsecs. Understanding how these processes regulate the thermal structure of the HI, the formation of cold clouds, and the transfer of matter and energy across scales is essential for connecting the small-scale physics of the ISM to the evolution of galaxies. Recent advances from SKA precursors have transformed our view of the atomic ISM, revealing a highly structured and filamentary cold medium, increasing the density of HI absorption measurements by orders of magnitude, and enabling new approaches to infer the thermodynamic and magnetic properties of the gas from spectral-line datasets. SKA-mid will provide the first comprehensive characterization of HI as a multi-phase, turbulent, and magnetized medium across the Milky Way and nearby galaxies. Its combination of sensitivity, angular resolution, spectral resolution, and survey speed will enable matched emission-absorption studies, dense optical-depth grids, and detailed mapping of the atomic-to-molecular transition over a broad range of environments. Combined with polarization, Zeeman, recombination-line, and multi-wavelength observations, SKA-mid will establish a unified observational framework to study the evolution of diffuse matter in galaxies, in connection with star formation, from the Solar neighborhood to galactic scales.

astro-ph.GA↗

arXiv2026

Competing with AI Scientists: Agent-Driven Approach to Astrophysics Research

We present an agent-driven approach to the construction of parameter inference pipelines for scientific data analysis. Our method leverages a multi-agent system, Cmbagent (the analysis system of the AI scientist Denario), in which specialized agents collaborate to generate research ideas, write and execute code, evaluate results, and iteratively refine the overall pipeline. As a case study, we apply this approach to the FAIR Universe Weak Lensing Uncertainty Challenge, a competition under time constraints focused on robust cosmological parameter inference with realistic observational uncertainties. While the fully autonomous exploration initially did not reach expert-level performance, the integration of human intervention enabled our agent-driven workflow to achieve a first-place result in the challenge. This demonstrates that semi-autonomous agentic systems can compete with, and in some cases surpass, expert solutions. We describe our workflow in detail, including both the autonomous and semi-autonomous exploration by Cmbagent. Our final inference pipeline utilizes parameter-efficient convolutional neural networks, likelihood calibration over a known parameter grid, and multiple regularization techniques. Our results suggest that agent-driven research workflows can provide a scalable framework to rapidly explore and construct pipelines for inference problems.

cs.AI↗

arXiv2026

Untangling dust emission and cosmic infrared background anisotropies with the scattering transform statistics

The template-fit approach is often used to separate Galactic dust emission from cosmic infrared background (CIB) anisotropies in low $\text{HI}$ column density regions, under the assumption that gas and dust are tightly correlated. However, this method fails in regions where additional Galactic emission from molecular hydrogen, diffuse ionized gas, and dark gas is present. We develop and test a statistical component-separation method to extract the dust signal from contaminated $\textit{Planck}$ $353\,\rm GHz$ observations using Scattering Covariance (SC) statistics. We first obtain a set of CIB maps over $25$ square patches, each covering an area of $222\,{\rm deg}^{2}$, using the linear correlation between dust and Galactic $21\,\rm cm$ $\text{HI}$ emission, which is valid in low $\text{HI}$ column density regions, through the template-fit approach. We then construct, from these $25$ maps, a generative model of the CIB using SC statistics. Finally, we rely on this contamination model to perform component separation of dust and CIB in the $\textit{Planck}$ data for different sky regions. Applying our algorithm to the $\textit{Planck}$ $353\,\rm GHz$ observations, we recover a dust map for a test sky region that exhibits more structure than the corrected SFD map at $100\,\mu \rm m$. The differences observed at the map level can be explained by decomposing the recovered $\textit{Planck}$ dust map into two gas phases: dust associated with $N_{\text{HI}}$ and dust associated with $N_{\text{H}_{2}}$. This work provides a clear pathway for mapping Galactic interstellar reddening over intermediate and high Galactic latitudes.

astro-ph.GA↗

arXiv2026

Bayesian imaging inverse problem with scattering transform

Bayesian imaging inverse problems in astrophysics and cosmology remain challenging, particularly in low-data regimes, due to complex forward operators and the frequent lack of well-motivated priors for non-Gaussian signals. In this paper, we introduce a Bayesian approach that addresses these difficulties by relying on a low-dimensional representation of physical fields built from Scattering Transform statistics. This representation enables inference to be performed in a compact model space, where we recover a posterior distribution over signal models that are consistent with the observed data. We propose an iterative adaptive algorithm to efficiently approximate this posterior distribution. We apply our method to a large-scale structure column density field from the Quijote simulations, using a realistic instrumental forward operator. We demonstrate both accurate statistical inference and deterministic signal reconstruction from a single contaminated image, without relying on any external prior distribution for the field of interest. These results demonstrate that Scattering Transform statistics provide an effective representation for solving complex imaging inverse problems in challenging low-data regimes. Our approach opens the way to new applications for non-Gaussian astrophysical and cosmological signals for which little or no prior modeling is available.

astro-ph.IM↗

arXiv2026

Separation of polarized dust emission in Planck observations with Scattering Transforms

Polarized dust emission is a major astrophysical foreground contaminant of the cosmic microwave background polarization (CMB), which must be accurately measured to look for the faint primordial polarization B-modes of inflationary origin. The available maps to date, obtained from Planck space mission data, are noise-dominated in the high Galactic latitude regions that are most relevant for CMB observations. The goal of this work is to obtain better dust polarization maps from Planck observations, by exploiting both the dependence between polarization and total intensity, as well as the non-Gaussian filamentary structure of the dust emission. To this end, we use scattering transforms, which provide a stable and interpretable representation of complex non-Gaussian textures, allowing for a data-driven analysis approach requiring no explicit priors on dust. The analysis is performed locally on Cartesian patches of sky, where Stokes linear polarization parameters, redefined in a local reference frame, are modeled as the sum of a signal of interest and a nuisance term. Using multiple realizations of the random nuisance term, we recover the polarized dust maps by minimizing a composite objective function that enforces multiple statistical constraints in scattering space. The proposed algorithm reconstructs maps of polarized dust emission whose statistics are consistent with those expected from the Planck data once random nuisance realizations are added. This is confirmed in a validation test using a high signal-to-noise sky region as a test case. Comparisons with existing dust polarization maps and models show that our approach better recovers small-scale polarized dust emission, and that our reconstructed power and cross-spectra closely match those of the dust polarization maps. A second set of maps that deterministically reproduce the features of the dust polarized emission is also produced.

astro-ph.CO↗

arXiv2025

Multiscale Turbulence Synthesis: Validation in 2D Hydrodynamics

Numerical simulations can follow the evolution of fluid motions through the intricacies of developed turbulence. However, they are rather costly to run, especially in 3D. In the past two decades, generative models have emerged which produce synthetic random flows at a computational cost equivalent to no more than a few time-steps of a simulation. These simplified models qualitatively bear some characteristics of turbulent flows in specific contexts (incompressible 3D hydrodynamics or magnetohydrodynamics), but generally struggle with the synthesis of coherent structures. We aim at generating random fields (e.g. velocity, density, magnetic fields, etc.) with realistic physical properties for a large variety of governing partial differential equations and at a small cost relative to time-resolved simulations. We propose a set of approximations applied to given sets of partial differential equations, and test the validity of our method in the simplest framework: 2D decaying incompressible hydrodynamical turbulence. We compare results of 2D decaying simulations with snapshots of our synthetic turbulence. We assess quantitatively the difference first with standard statistical tools: power spectra, increments and structure functions. These indicators can be reproduced by our method during up to about a third of the turnover time scale. We also consider recently developed scattering transforms statistics, able to efficiently characterise non-Gaussian structures. This reveals more significant discrepancy, which can however be bridged by bootstrapping. Finally, the number of Fourier transforms necessary for one synthesis scales logarithmically in the resolution, compared to linearly for time-resolved simulations. We have designed a multiscale turbulence synthesis (MuScaTS) method to efficiently short-circuit costly numerical simulations to produce realistic instantaneous fields.

astro-ph.GA↗

arXiv2025

Generative Models of 21cm EoR Lightcones with 3D Scattering Transforms

The 21cm signal from the Epoch of Reionization (EoR) is observed as a three-dimensional data set known as a lightcone, consisting of a redshift (frequency) axis and two spatial sky plane axes. When observed by radio interferometers, this EoR signal is strongly obscured by foregrounds that are several orders of magnitude stronger. Due to its inherently non-Gaussian nature, the EoR signal requires robust statistical tools to accurately separate it from these foreground contaminants, but current foreground separation techniques focus primarily on recovering the EoR power spectrum, often neglecting valuable non-Gaussian information. Recent developments in astrophysics, particularly in the context of the Galactic interstellar medium, have demonstrated the efficacy of scattering transforms - novel summary statistics for highly non-Gaussian processes - for component separation tasks. Motivated by these advances, we extend the scattering transform formalism from two-dimensional data sets to three-dimensional EoR lightcones. To this end, we introduce a 3D wavelet set from the tensor product of 2D isotropic wavelets in the sky plane domain and 1D wavelets in the redshift domain. As generative models form the basis of component separation, our focus here is on building and validating generative models that can be used for component separation in future projects. To achieve this, we construct maximum entropy generative models to synthesise EoR lightcones, and statistically validate the generative model by quantitatively comparing the synthesised EoR lightcones with the single target lightcone used to construct them, using independent statistics such as the power spectrum and Minkowski Functionals. The synthesised lightcones agree well with the target lightcone both statistically and visually, opening up the possibility of developing for component separation methods using 3D scattering transforms.

astro-ph.CO↗

arXiv2024

Comparing the morphology of molecular clouds without supervision

Molecular clouds show complex structures reflecting their non-linear dynamics. Many studies investigating the bridge between their morphology and physical properties have shown the value of non-Gaussian higher-order statistics in capturing physical information. Yet, as this bridge is usually characterized in the supervised world of simulations, transferring it to observations can be hazardous, especially when the discrepancy between simulations and observations remains unknown. In this paper, we aim to identify relevant summary statistics, directly from the observation data. To do so, we developed a test to compare the informative power of two sets of summary statistics for a given unlabeled dataset. Contrary to supervised approaches, this test does not require knowledge of any class label or parameter associated with the data. Instead, it evaluates and compares the degeneracy levels of the summary statistics based on a notion of statistical compatibility. We applied this test to column density maps of 14 nearby molecular clouds observed by Herschel and iteratively compared different sets of typical summary statistics. We show that a standard Gaussian description of these clouds is highly degenerate but can be substantially improved when being estimated on the logarithm of the maps. This illustrates that low-order statistics, when properly used, remain very powerful. We further show that such descriptions still exhibit a small quantity of degeneracies, some of which are lifted by the higher-order statistics provided by reduced wavelet scattering transforms. These degeneracies quantitatively differ between observations and state-of-the-art simulations of dense cloud collapse, and they are not present for logFBM models. Finally, we show how to cooperatively use the summary statistics identified to build a morphological distance, which is evaluated visually and gives convincing results.

astro-ph.GA↗

arXiv2024

Scattering transforms on the sphere, application to large scale structure modelling

astro-ph.IM↗

arXiv2024

Generative models of astrophysical fields with scattering transforms on the sphere

Scattering transforms are a new type of summary statistics recently developed for the study of highly non-Gaussian processes, which have been shown to be very promising for astrophysical studies. In particular, they allow one to build generative models of complex non-linear fields from a limited amount of data, and have also been used as the basis of new statistical component separation algorithms. In the context of upcoming cosmological surveys, such as LiteBIRD for the cosmic microwave background polarization or Rubin-LSST and Euclid for study of the large scale structures of the Universe, the extension of these tools to spherical data is necessary. We develop scattering transforms on the sphere and focus on the construction of maximum-entropy generative models of several astrophysical fields. We construct, from a single target field, generative models of homogeneous astrophysical and cosmological fields, whose samples are quantitatively compared to the target fields using common statistics (power spectrum, pixel probability density function and Minkowski functionals). Our sampled fields agree well with the target fields, both statistically and visually. These generative models therefore open up a wide range of new applications for future astrophysical and cosmological studies; particularly those for which very little simulated data is available. We make our code available to the community so that this work can be easily reproduced and developed further.

astro-ph.IM↗

arXiv2023

Wavelet Based Statistics for Enhanced 21cm EoR Parameter Constraints

We propose a new approach to improve the precision of astrophysical parameter constraints for the 21cm signal from the epoch of reionisation (EoR). Our method introduces new sets of summary statistics, hereafter `evolution compressed' statistics, which quantify the spectral evolution of the 2D spatial statistics computed at fixed redshift. We defined such compressed statistics for power spectrum (PS), wavelet scattering transforms (WST), and wavelet moments (WM), which also characterise non-Gaussian features. To compare these different statistics, along with the 3D power spectrum, we estimated their Fisher information on three cosmological parameters from an ensemble of simulations of 21cm EoR data, both in noiseless and noisy scenarios using Square Kilometre Array (SKA) noise levels equivalent to 100 and 1000 hours of observations. We also compare wavelet statistics, in particular WST, built from standard directional Morlet wavelets, as well as from a set of isotropic wavelets derived from the binning window function of the 2D power spectrum. For the noiseless case, the compressed wavelet statistics give constraints that are up to five times more precise than those obtained from the 3D isotropic power spectrum. At the same time, for 100h SKA noise, from which it is difficult to extract non-Gaussian features, compressed wavelet statistics still give over 30\% tighter constraints. We find that the wavelet statistics with wavelets derived from the power-spectrum binning window function provide the tightest constraints of all the statistics, with the WSTs seemingly performing better than the WMs, in particular when working with noisy data. The findings of this study demonstrate that evolution-compressed statistics extract more information than usual 3D isotropic power-spectra approaches and that our wavelet-based statistics can consistently outmatch power-spectrum-based statistics.

astro-ph.CO↗

arXiv2023

Scattering Spectra Models for Physics

Physicists routinely need probabilistic models for a number of tasks such as parameter inference or the generation of new realizations of a field. Establishing such models for highly non-Gaussian fields is a challenge, especially when the number of samples is limited. In this paper, we introduce scattering spectra models for stationary fields and we show that they provide accurate and robust statistical descriptions of a wide range of fields encountered in physics. These models are based on covariances of scattering coefficients, i.e. wavelet decomposition of a field coupled with a point-wise modulus. After introducing useful dimension reductions taking advantage of the regularity of a field under rotation and scaling, we validate these models on various multi-scale physical fields and demonstrate that they reproduce standard statistics, including spatial moments up to 4th order. These scattering spectra provide us with a low-dimensional structured representation that captures key properties encountered in a wide range of physical fields. These generic models can be used for data exploration, classification, parameter inference, symmetry detection, and component separation.

physics.data-an↗

arXiv2023

Unearthing InSights into Mars: Unsupervised Source Separation with Limited Data

Source separation involves the ill-posed problem of retrieving a set of source signals that have been observed through a mixing operator. Solving this problem requires prior knowledge, which is commonly incorporated by imposing regularity conditions on the source signals, or implicitly learned through supervised or unsupervised methods from existing data. While data-driven methods have shown great promise in source separation, they often require large amounts of data, which rarely exists in planetary space missions. To address this challenge, we propose an unsupervised source separation scheme for domains with limited data access that involves solving an optimization problem in the wavelet scattering covariance representation space$\unicode{x2014}$an interpretable, low-dimensional representation of stationary processes. We present a real-data example in which we remove transient, thermally-induced microtilts$\unicode{x2014}$known as glitches$\unicode{x2014}$from data recorded by a seismometer during NASA's InSight mission on Mars. Thanks to the wavelet scattering covariances' ability to capture non-Gaussian properties of stochastic processes, we are able to separate glitches using only a few glitch-free data snippets.

cs.LG↗

arXiv2023

Multi-scale clustering and source separation of InSight mission seismic data

Unsupervised source separation involves unraveling an unknown set of source signals recorded through a mixing operator, with limited prior knowledge about the sources, and only access to a dataset of signal mixtures. This problem is inherently ill-posed and is further challenged by the variety of timescales exhibited by sources in time series data from planetary space missions. As such, a systematic multi-scale unsupervised approach is needed to identify and separate sources at different timescales. Existing methods typically rely on a preselected window size that determines their operating timescale, limiting their capacity to handle multi-scale sources. To address this issue, we propose an unsupervised multi-scale clustering and source separation framework by leveraging wavelet scattering spectra that provide a low-dimensional representation of stochastic processes, capable of distinguishing between different non-Gaussian stochastic processes. Nested within this representation space, we develop a factorial variational autoencoder that is trained to probabilistically cluster sources at different timescales. To perform source separation, we use samples from clusters at multiple timescales obtained via the factorial variational autoencoder as prior information and formulate an optimization problem in the wavelet scattering spectra representation space. When applied to the entire seismic dataset recorded during the NASA InSight mission on Mars, containing sources varying greatly in timescale, our approach disentangles such different sources, e.g., minute-long transient one-sided pulses (known as "glitches") and structured ambient noises resulting from atmospheric activities that typically last for tens of minutes, and provides an opportunity to conduct further investigations into the isolated sources.

cs.LG↗

arXiv2023

Separation of dust emission from the Cosmic Infrared Background in Herschel observations with Wavelet Phase Harmonics

The low brightness dust emission at high Galactic latitude is of interest to study the interplay between physical processes in shaping the structure of the interstellar medium (ISM), as well as to statistically characterize dust emission as a foreground to the Cosmic Microwave Background (CMB). Progress in this avenue of research have been hampered by the difficulty of separating the dust emission from the Cosmic Infrared Background (CIB). We demonstrate that dust and CIB may be effectively separated based on their different structure on the sky and use the separation to characterize the structure of diffuse dust emission on angular scales where CIB is a significant component in terms of power. We use scattering transform statistics, the Wavelet Phase Harmonics (WPH), to perform a statistical component separation using Herschel SPIRE observations. This component separation is done only from observational data using non-Gaussian properties as a lever arm, and is done at a single 250 microns frequency. This method, that we validate on mock data, gives us access to non-Gaussian statistics of the interstellar dust and an output dust map essentially free from CIB contamination. Our statistical modelling characterizes the non-Gaussian structure of the diffuse ISM down to the smallest scales observed by Herschel. We recover the power-law shape of the dust power spectrum up to a wavenumber of 2 arcmin$^{-1}$ where the dust signal represents 2 percent of the total power. The output dust map reveals coherent structures at the smallest scales which were hidden by the CIB anisotropies. It opens new observational perspectives on the formation of structure in the diffuse ISM which we discuss with reference to past work. We have succeeded to perform a statistical separation from observational data only at a single frequency by using non-Gaussian statistics.

astro-ph.GA↗

arXiv2022

Generative Models of Multi-channel Data from a Single Example -- Application to Dust Emission

The quest for primordial $B$-modes in the cosmic microwave background has emphasized the need for refined models of the Galactic dust foreground. Here, we aim at building a realistic statistical model of the multi-frequency dust emission from a single example. We introduce a generic methodology relying on microcanonical gradient descent models conditioned by an extended family of wavelet phase harmonic (WPH) statistics. To tackle the multi-channel aspect of the data, we define cross-WPH statistics, quantifying non-Gaussian correlations between maps. Our data-driven methodology could apply to various contexts, and we have updated the software PyWPH, on which this work relies, accordingly. Applying this to dust emission maps built from a magnetohydrodynamics simulation, we construct and assess two generative models of: 1) a $(I, E, B)$ multi-observable input, 2) a $\{I_ν\}_ν$ multi-frequency input. The samples exhibit consistent features compared to the original maps. A statistical analysis of 1) shows that the power spectra, distributions of pixels, and Minkowski functionals are captured to a good extent. We analyze 2) by fitting the spectral energy distribution (SED) of both the synthetic and original maps with a modified blackbody (MBB) law. The maps are equally well fitted, and a comparison of the MBB parameters shows that our model succeeds in capturing the spatial variations of the SED from the data. Besides the perspectives of this work for dust emission modeling, the introduction of cross-WPH statistics opens a new avenue to characterize non-Gaussian interactions across different maps, which we believe will be fruitful for astrophysics.

astro-ph.CO↗

arXiv2022

Wavelet Moments for Cosmological Parameter Estimation

Extracting non-Gaussian information from the non-linear regime of structure formation is key to fully exploiting the rich data from upcoming cosmological surveys probing the large-scale structure of the universe. However, due to theoretical and computational complexities, this remains one of the main challenges in analyzing observational data. We present a set of summary statistics for cosmological matter fields based on 3D wavelets to tackle this challenge. These statistics are computed as the spatial average of the complex modulus of the 3D wavelet transform raised to a power $q$ and are therefore known as invariant wavelet moments. The 3D wavelets are constructed to be radially band-limited and separable on a spherical polar grid and come in three types: isotropic, oriented, and harmonic. In the Fisher forecast framework, we evaluate the performance of these summary statistics on matter fields from the Quijote suite, where they are shown to reach state-of-the-art parameter constraints on the base $Λ$CDM parameters, as well as the sum of neutrino masses. We show that we can improve constraints by a factor 5 to 10 in all parameters with respect to the power spectrum baseline.

astro-ph.CO↗