Reconstructing Cosmological Initial Conditions from Late-Time Structure with Convolutional Neural Networks

Abstract

We present a method to reconstruct the initial linear-regime matter density field from the late-time non-linearly evolved density field in which we channel the output of standard first-order reconstruction to a convolutional neural network (CNN). Our method shows dramatic improvement over the reconstruction of either component alone. We show why CNNs are not well-suited for reconstructing the initial density directly from the late-time density: CNNs are local models, but the relationship between initial and late-time density is not local. Our method leverages standard reconstruction as a preprocessing step, which inverts bulk gravitational flows sourced over very large scales, transforming the residual reconstruction problem from long-range to local and making it ideally suited for a CNN. We develop additional techniques to account for redshift distortions, which warp the density fields measured by galaxy surveys. Our method improves the range of scales of high-fidelity reconstruction by a factor of 2 in wavenumber above standard reconstruction, corresponding to a factor of 8 increase in the number of well-reconstructed modes. In addition, our method almost completely eliminates the anisotropy caused by redshift distortions. As galaxy surveys continue to map the Universe in increasingly greater detail, our results demonstrate the opportunity offered by CNNs to untangle the non-linear clustering at intermediate scales more accurately than ever before.

Published in Monthly Notices of the Royal Astronomical Society

References in corpus (8)

• Improving Cosmological Distance Measurements by Reconstruction of the Baryon Acoustic Peak
• AbacusSummit: A Massive Set of High-Accuracy, High-Resolution $N$-Body Simulations
• Reconstructing Baryon Oscillations: A Lagrangian Theory Perspective
• The $\text{Abacus}$ Cosmological $N$-body Code
• Reconstruction within the Zeldovich approximation
• Galaxy Bias and its Effects on the Baryon Acoustic Oscillations Measurements
• Efficient Reconstruction of Linear Baryon Acoustic Oscillations in Galaxy Surveys
• New method for initial density reconstruction

Cited by in corpus (15)

• Cosmology inference at the field level from biased tracers in redshift-space
• Effective cosmic density field reconstruction with convolutional neural network
• Improving constraints on primordial non-Gaussianity using neural network based reconstruction
• Neural network reconstruction of density and velocity fields from the 2MASS Redshift Survey
• Learning the Universe: Learning to Optimize Cosmic Initial Conditions with Non-Differentiable Structure Formation Models
• Neural Network Reconstruction of Non-Gaussian Initial Conditions from Dark Matter Halos
• ParamANN: A Neural Network to Estimate Cosmological Parameters for $\Lambda$CDM Universe Using Hubble Measurements
• Lagrangian displacement field estimators in cosmology
• Reconstruction of full sky CMB $\mathbf{E}$ and $\mathbf{B}$ modes spectra removing $\mathbf{E}$-to-$\mathbf{B}$ leakage from partial sky using deep learning
• Map Reconstruction of radio observations with Conditional Invertible Neural Networks
• Lagrangian space remapping and the angular momentum reconstruction from cosmic structures
• From Redshift to Real Space: Combining Linear Theory With Neural Networks
• Fast Sampling of Cosmological Initial Conditions with Gaussian Neural Posterior Estimation
• On the Connection between Field-Level Inference and $n$-point Correlation Functions
• Searching optimal scales for reconstructing cosmological initial conditions using convolutional neural networks