Simulation design is the choice of locations in parameter space at which simulations are to be run and is the first step in building an emulator capable of quickly providing estimates of simulation results for arbitrary locations in the parameter space. We introduce an alteration to the "OALHS" design used by Heitmann et al. (2006) that reduces the number of simulation runs required to achieve a fixed accuracy in our case study by a factor of two. We also compare interpolation procedures for emulators and find that interpolation via Gaussian Process models and via the much-easier-to-implement polynomial interpolation have comparable accuracy. A very simple emulation-building procedure consisting of a design sampled from the parameter prior distribution, combined with interpolation via polynomials also performs well. Although our primary motivation is efficient emulators of non-linear cosmological N-body simulations, in an appendix we describe an emulator for the CMB temperature power spectrum publicly available as computer code.
The interpretation of cosmological observables requires the use of increasingly sophisticated theoretical models. Since these models are becoming computationally very expensive and display non-trivial uncertainties, the use of standard Bayesian algorithms for cosmological inferences, such as MCMC, might become inadequate. Here, we propose a new approach to parameter estimation based on an iterative Gaussian emulation of the target likelihood function. This requires a minimal number of likelihood evaluations and naturally accommodates for stochasticity in theoretical models. We apply the algori…
We develop a set of machine-learning based cosmological emulators, to obtain fast model predictions for the $C(\ell)$ angular power spectrum coefficients characterising tomographic observations of galaxy clustering and weak gravitational lensing from multi-band photometric surveys (and their cross-correlation). A set of neural networks are trained to map cosmological parameters into the coefficients, achieving a speed-up $\mathcal{O}(10^3)$ in computing the required statistics for a given set of cosmological parameters, with respect to standard Boltzmann solvers, with an accuracy better than $…
We describe our nonlinear emulation (i.e., interpolation) framework that combines the halo occupation distribution (HOD) galaxy bias model with $N$-body simulations of nonlinear structure formation, designed to accurately predict the projected clustering and galaxy-galaxy lensing signals from luminous red galaxies (LRGs) in the redshift range $0.16 < z < 0.36$ on comoving scales $0.6 < r_p < 30$ \hMpc. The interpolation accuracy is $\lesssim 1-2$ per cent across the entire physically plausible range of parameters for all scales considered. We correctly recover the true value of the cosmologica…
We present MGLenS, a large series of modified gravity lensing simulations tailored for cosmic shear data analyses and forecasts in which cosmological and modified gravity parameters are varied simultaneously. Based on the FORGE and BRIDGE $N$-body simulation suites presented in companion papers, we construct 500,000 deg$^2$ of mock Stage-IV lensing data, sampling a pair of 4-dimensional volumes designed for the training of emulators. We validate the accuracy of MGLenS with inference analyses based on the lensing power spectrum exploiting our implementation of $f(R)$ and nDGP theoretical predic…
Cosmological Field Emulation and Parameter Inference with Diffusion Models
Harvard University（哈佛大学）；The NSF AI Institute for Artificial Intelligence and Fundamental Interactions（NSF人工智能与基础相互作用人工智能研究所）；Massachusetts Institute of Technology（麻省理工学院）
Design and optimization of neural networks for multifidelity cosmological emulation
University of California, Riverside（加州大学河滨分校）；Department of Physics & Astronomy, University of California, Riverside（加州大学河滨分校物理与天文系）；Department of Physics, University of Michigan（密歇根大学物理系）；Leinweber Center for Theoretical Physics（莱因韦伯理论物理中心）；The University of Texas at Austin（德克萨斯大学奥斯汀分校）
Modern cosmological surveys are delivering datasets characterized by unprecedented quality and statistical completeness; this trend is expected to continue into the future as new ground- and space-based surveys come online. In order to maximally extract cosmological information from these observations, matching theoretical predictions are needed. At low redshifts, the surveys probe the nonlinear regime of structure formation where cosmological simulations are the primary means of obtaining the required information. The computational cost of sufficiently resolved large-volume simulations makes …
In this work we extend our recently developed multi-fidelity emulation technique to the simulated Lyman-$α$ forest flux power spectrum. Multi-fidelity emulation allows interpolation of simulation outputs between cosmological parameters using many cheap low-fidelity simulations and a few expensive high-fidelity simulations. Using a test suite of small box (30 Mpc/h) simulations, we show that multi-fidelity emulation is able to reproduce the Lyman-$α$ forest flux power spectrum well, achieving an average accuracy when compared to a test suite of $0.8\%$. We further show that it has a substantial…
We construct an emulator for the halo mass function over group and cluster mass scales for a range of cosmologies, including the effects of dynamical dark energy and massive neutrinos. The emulator is based on the recently completed Mira-Titan Universe suite of cosmological $N$-body simulations. The main set of simulations spans 111 cosmological models with 2.1 Gpc boxes. We extract halo catalogs in the redshift range $z=[0.0, 2.0]$ and for masses $M_{200\mathrm{c}}\geq 10^{13}M_\odot/h$. The emulator covers an 8-dimensional hypercube spanned by {$Ω_\mathrm{m}h^2$, $Ω_\mathrm{b}h^2$, $Ω_νh^2$,…
We compare and validate COLA (COmoving Lagrangian Acceleration) simulations against existing emulators in the literature, namely Bacco and Euclid Emulator 2. Our analysis focuses on the non-linear response function, i.e., the ratio between the non-linear dark matter power spectrum in a given cosmology with respect to a pre-defined reference cosmology, which is chosen to be the Euclid Emulator 2 reference cosmology in this paper. We vary three cosmological parameters, the total matter density, the amplitude of the primordial scalar perturbations and the spectral index. By comparing the COLA non…
Three-dimensional Conditional Diffusion Models for Cosmological 21 cm Lightcone Emulation
Center for Relativistic Astrophysics, School of Physics, Georgia Institute of Technology, Atlanta, GA 30332, USA（相对论天体物理中心，物理学院，佐治亚理工学院，亚特兰大，GA 30332，USA）
We present methods for interpolating between the 1-D flux power spectrum of the Lyman-$α$ forest, as output by cosmological hydrodynamic simulations. Interpolation is necessary for cosmological parameter estimation due to the limited number of simulations possible. We construct an emulator for the Lyman-$α$ forest flux power spectrum from $21$ small simulations using Latin hypercube sampling and Gaussian process interpolation. We show that this emulator has a typical accuracy of 1.5% and a worst-case accuracy of 4%, which compares well to the current statistical error of 3 - 5% at $z < 3$ from…
Swiss Data Science Center, ETH Zurich（苏黎世联邦理工学院瑞士数据科学中心）；ETH Zurich（苏黎世联邦理工学院）
Weak gravitational lensing mass maps play a crucial role in understanding the evolution of structures in the universe and our ability to constrain cosmological models. The prediction of these mass maps is based on expensive N-body simulations, which can create a computational bottleneck for cosmological analyses. Modern deep generative models, such as Generative Adversarial Networks (GAN), have demonstrated their potential to achieve this goal. Most existing GAN approaches produce simulations for a fixed value of the cosmological parameters, which limits their practical applicability. We propo…
We perform an ensemble of $N$-body simulations with $2048^3$ particles for 101 flat $w$CDM cosmological models sampled based on a maximin-distance Sliced Latin Hypercube Design. By using the halo catalogs extracted at multiple redshifts in the range of $z=[0,1.48]$, we develop Dark Emulator, which enables fast and accurate computations of the halo mass function, halo-matter cross-correlation, and halo auto-correlation as a function of halo masses, redshift, separations and cosmological models, based on the Principal Component Analysis and the Gaussian Process Regression for the large-dimension…
We present methods for emulating the matter power spectrum by combining information from cosmological $N$-body simulations at different resolutions. An emulator allows estimation of simulation output by interpolating across the parameter space of a limited number of simulations. We present the first implementation in cosmology of multi-fidelity emulation, where many low-resolution simulations are combined with a few high-resolution simulations to achieve an increased emulation accuracy. The power spectrum's dependence on cosmology is learned from the low-resolution simulations, which are in tu…
Gaussian-Process Emulation of the Redshift-Space Halo Power Spectrum Monopole in Cosmologies with Massive Neutrinos
National Astronomical Observatories, Chinese Academy of Sciences（中国科学院国家天文台）；School of Astronomy and Space Science, University of Chinese Academy of Sciences（中国科学院大学天文与空间科学学院）；Institute for Frontiers in Astronomy and Astrophysics, Beijing Normal University（北京师范大学天文与前沿科学研究院）；Chinese Academy of Sciences South America Center for Astronomy (CASSACA), National Astronomical Observatories, Chinese Academy of Sciences（中国科学院南美天文中心）
SageNet: Fast Neural Network Emulation of the Stiff-amplified Gravitational Waves from Inflation
Zhejiang University（浙江大学）；Massachusetts Institute of Technology（麻省理工学院）；Guangxi University（广西大学）；Newton South High School（牛顿南高中）；Wuhan University（武汉大学）；Southern Methodist University（南方卫理公会大学）；The University of Texas at Austin（德克萨斯大学奥斯汀分校）
The Lyman-alpha forest provides strong constraints on both cosmological parameters and intergalactic medium astrophysics, which are forecast to improve further with the next generation of surveys including eBOSS and DESI. As is generic in cosmological inference, extracting this information requires a likelihood to be computed throughout a high-dimensional parameter space. Evaluating the likelihood requires a robust and accurate mapping between the parameters and observables, in this case the 1D flux power spectrum. Cosmological simulations enable such a mapping, but due to computational time c…
Studying the impact of systematic effects, optimizing survey strategies, assessing tensions between different probes and exploring synergies of different data sets require a large number of simulated likelihood analyses, each of which cost thousands of CPU hours. In this paper, we present a method to accelerate cosmological inference using emulators based on Gaussian process regression and neural networks. We iteratively acquire training samples in regions of high posterior probability which enables accurate emulation of data vectors even in high dimensional parameter spaces. We showcase the p…
We use the emulation framework CosmoPower to construct and publicly release neural network emulators of cosmological observables, including the Cosmic Microwave Background (CMB) temperature and polarization power spectra, matter power spectrum, distance-redshift relation, baryon acoustic oscillation (BAO) and redshift-space distortion (RSD) observables, and derived parameters. We train our emulators on Einstein-Boltzmann calculations obtained with high-precision numerical convergence settings, for a wide range of cosmological models including $Λ$CDM, $w$CDM, $Λ$CDM+$N_\mathrm{eff}$, and $Λ$CDM…
We implement a model for the two-point statistics of biased tracers that combines dark matter dynamics from $N$-body simulations with an analytic Lagrangian bias expansion. Using Aemulus, a suite of $N$-body simulations built for emulation of cosmological observables, we emulate the cosmology dependence of these nonlinear spectra from redshifts $z = 0$ to $z=2$. We quantify the accuracy of our emulation procedure, which is sub-per cent at $k=1\, h {\rm Mpc}^{-1}$ for the redshifts probed by upcoming surveys and improves at higher redshifts. We demonstrate its ability to describe the statistics…
Scientific analyses often rely on slow, but accurate forward models for observable data conditioned on known model parameters. While various emulation schemes exist to approximate these slow calculations, these approaches are only safe if the approximations are well understood and controlled. This workshop submission reviews and updates a previously published method, which has been used in cosmological simulations, to (1) train an emulator while simultaneously estimating posterior probabilities with MCMC and (2) explicitly propagate the emulation error into errors on the posterior probabilitie…
Testing a subset of viable cosmological models beyond General Relativity (GR), with implications for cosmic acceleration and the Dark Energy associated with it, is within the reach of Rubin Observatory Legacy Survey of Space and Time (LSST) and a part of its endeavor. Deviations from GR-w(z)CDM models can manifest in the growth rate of structure and lensing, as well as in screening effects on non-linear scales. We explore the constraining power of small-scale deviations predicted by the f(R) Hu-Sawicki Modified Gravity (MG) candidate, by emulating this model with COLA (COmoving Lagrangian Acce…
