[en] Cosmological probes pose an inverse problem where the measurement result is obtained through observations, and the objective is to infer values of model parameters that characterize the underlying physical system—our universe, from these observations and theoretical forward-modeling. The only way to accurately forward-model physical behavior on small scales is via expensive numerical simulations, which are further “emulated” due to their high cost. Emulators are commonly built with a set of simulations covering the parameter space with Latin hypercube sampling and an interpolation procedure; the aim is to establish an approximately constant prediction error across the hypercube. In this paper, we provide a description of a novel statistical framework for obtaining accurate parameter constraints. The proposed framework uses multi-output Gaussian process emulators that are adaptively constructed using Bayesian optimization methods with the goal of maintaining a low emulation error in the region of the hypercube preferred by the observational data. In this paper, we compare several approaches for constructing multi-output emulators that enable us to take possible inter-output correlations into account while maintaining the efficiency needed for inference. Using a Lyα forest flux power spectrum, we demonstrate that our adaptive approach requires considerably fewer—by a factor of a few in the Lyα P(k) case considered here—simulations compared to the emulation based on Latin hypercube sampling, and that the method is more robust in reconstructing parameters and their Bayesian credible intervals.

[en] Cosmological hydrodynamic simulations can accurately predict the properties of the intergalactic medium (IGM), but only under the condition of retaining the high spatial resolution necessary to resolve density fluctuations in the IGM. This resolution constraint prohibits simulating large volumes, such as those probed by BOSS and future surveys, like DESI and 4MOST. To overcome this limitation, we present “Iteratively Matched Statistics” (IMS), a novel method to accurately model the Ly α forest with collisionless N -body simulations, where the relevant density fluctuations are unresolved. We use a small-box, high-resolution hydrodynamic simulation to obtain the probability distribution function (PDF) and the power spectrum of the real-space Ly α forest flux. These two statistics are iteratively mapped onto a pseudo-flux field of an N -body simulation, which we construct from the matter density. We demonstrate that our method can reproduce the PDF, line of sight and 3D power spectra of the Ly α forest with good accuracy (7%, 4%, and 7% respectively). We quantify the performance of the commonly used Gaussian smoothing technique and show that it has significantly lower accuracy (20%–80%), especially for N -body simulations with achievable mean inter-particle separations in large-volume simulations. In addition, we show that IMS produces reasonable and smooth spectra, making it a powerful tool for modeling the IGM in large cosmological volumes and for producing realistic “mock” skies for Ly α forest surveys.

[en] The ultraviolet background (UVB) emitted by quasars and galaxies governs the ionization and thermal state of the intergalactic medium (IGM), regulates the formation of high-redshift galaxies, and is thus a key quantity for modeling cosmic reionization. The vast majority of cosmological hydrodynamical simulations implement the UVB via a set of spatially uniform photoionization and photoheating rates derived from UVB synthesis models. We show that simulations using canonical UVB rates reionize and, perhaps more importantly, spuriously heat the IGM, much earlier ($z\sim <!--\; ???\; -->15$) than they should. This problem arises because at $z><!--\; >\; -->6$, where observational constraints are nonexistent, the UVB amplitude is far too high. We introduce a new methodology to remedy this issue, and we generate self-consistent photoionization and photoheating rates to model any chosen reionization history. Following this approach, we run a suite of hydrodynamical simulations of different reionization scenarios and explore the impact of the timing of reionization and its concomitant heat injection on the thermal state of the IGM. We present a comprehensive study of the pressure smoothing scale of IGM gas, illustrating its dependence on the details of both hydrogen and helium reionization, and argue that it plays a fundamental role in interpreting Lyα forest statistics and the thermal evolution of the IGM. The premature IGM heating we have uncovered implies that previous work has likely dramatically overestimated the impact of photoionization feedback on galaxy formation, which sets the minimum halo mass able to form stars at high redshifts. We make our new UVB photoionization and photoheating rates publicly available for use in future simulations.

[en] Many galaxy formation models predict alignments between galaxy spin and the cosmic web (i.e., directions of filaments and sheets), leading to an intrinsic alignment between galaxies that creates a systematic error in weak-lensing measurements. These effects are often predicted to be stronger at high redshifts (z ≳ 1) that are inaccessible to massive galaxy surveys on foreseeable instrumentation, but IGM tomography of the Lyα forest from closely spaced quasars and galaxies is starting to measure the z ∼ 2–3 cosmic web with requisite fidelity. Using mock surveys from hydrodynamical simulations, we examine the utility of this technique, in conjunction with coeval galaxy samples, to measure alignment between galaxies and the cosmic web at z ∼ 2.5. We show that IGM tomography surveys with ≲5 h ⁻¹ Mpc sightline spacing can accurately recover the eigenvectors of the tidal tensor, which we use to define the directions of the cosmic web. For galaxy spins and shapes, we use a model parameterized by the alignment strength, $\mathrm{\Delta <!--\; ??\; -->}⟨<!--\; ???\; -->\cos <!--\; ???\; -->\mathrm{\theta <!--\; ??\; -->}⟩<!--\; ???\; -->$, with respect to the tidal tensor eigenvectors from the underlying density field, and also consider observational effects such as errors in the galaxy position angle, inclination, and redshift. Measurements using the upcoming ∼1 deg² CLAMATO tomographic survey and 600 coeval zCOSMOS-Deep galaxies should place 3σ limits on extreme alignment models with $\mathrm{\Delta <!--\; ??\; -->}⟨<!--\; ???\; -->\cos <!--\; ???\; -->\mathrm{\theta <!--\; ??\; -->}⟩<!--\; ???\; -->\sim <!--\; ???\; -->0.1$, but much larger surveys encompassing >10,000 galaxies, such as Subaru PFS, will be required to constrain models with $\mathrm{\Delta <!--\; ??\; -->}⟨<!--\; ???\; -->\cos <!--\; ???\; -->\mathrm{\theta <!--\; ??\; -->}⟩<!--\; ???\; -->\sim <!--\; ???\; -->0.03$. These measurements will constrain models of galaxy–cosmic web alignment and test tidal torque theory at z ∼ 2, improving our understanding of the physics of intrinsic alignments.

[en] Sky surveys are the largest data generators in astronomy, making automated tools for extracting meaningful scientific information an absolute necessity. We show that, without the need for labels, self-supervised learning recovers representations of sky survey images that are semantically useful for a variety of scientific tasks. These representations can be directly used as features, or fine-tuned, to outperform supervised methods trained only on labeled data. We apply a contrastive learning framework on multiband galaxy photometry from the Sloan Digital Sky Survey (SDSS), to learn image representations. We then use them for galaxy morphology classification and fine-tune them for photometric redshift estimation, using labels from the Galaxy Zoo 2 data set and SDSS spectroscopy. In both downstream tasks, using the same learned representations, we outperform the supervised state-of-the-art results, and we show that our approach can achieve the accuracy of supervised models while using 2–4 times fewer labels for training. The codes, trained models, and data can be found at https://portal.nersc.gov/project/dasrepo/self-supervised-learning-sdss.

[en] We present a new N-body and gas dynamics code, called Nyx, for large-scale cosmological simulations. Nyx follows the temporal evolution of a system of discrete dark matter particles gravitationally coupled to an inviscid ideal fluid in an expanding universe. The gas is advanced in an Eulerian framework with block-structured adaptive mesh refinement; a particle-mesh scheme using the same grid hierarchy is used to solve for self-gravity and advance the particles. Computational results demonstrating the validation of Nyx on standard cosmological test problems, and the scaling behavior of Nyx to 50,000 cores, are presented.

[en] Inferring model parameters from experimental data is a grand challenge in many sciences, including cosmology. This often relies critically on high fidelity numerical simulations, which are prohibitively computationally expensive. The application of deep learning techniques to generative modeling is renewing interest in using high dimensional density estimators as computationally inexpensive emulators of fully-fledged simulations. These generative models have the potential to make a dramatic shift in the field of scientific simulations, but for that shift to happen we need to study the performance of such generators in the precision regime needed for science applications. To this end, in this work we apply Generative Adversarial Networks to the problem of generating weak lensing convergence maps. We show that our generator network produces maps that are described by, with high statistical confidence, the same summary statistics as the fully simulated maps.

[en] We study quasar proximity zones in the redshift range $5.77⩽<!--\; ???\; -->z⩽<!--\; ???\; -->6.54$ by homogeneously analyzing 34 medium-resolution spectra, encompassing both archival and newly obtained data, and exploiting recently updated systemic redshift and magnitude measurements. Whereas previous studies found strong evolution of proximity zone sizes with redshift and argued that this provides evidence for a rapidly evolving intergalactic medium (IGM) neutral fraction during reionization, we measure a much shallower trend $\propto <!--\; ???\; -->{(1+z)}^{-<!--\; ???\; -->1.44}$. We compare our measured proximity zone sizes to predictions from hydrodynamical simulations post-processed with one-dimensional radiative transfer and find good agreement between observations and theory, irrespective of the ionization state of the ambient IGM. This insensitivity to IGM ionization state has been previously noted, and results from the fact that the definition of proximity zone size as the first drop of the smoothed quasar spectrum below the 10% flux transmission level probes locations where the ionizing radiation from the quasar is an order of magnitude larger than the expected ultraviolet ionizing background that sets the neutral fraction of the IGM. Our analysis also uncovered three objects with exceptionally small proximity zones (two have $R_p<<!--\; <\; -->1$ proper Mpc), which constitute outliers from the observed distribution and are challenging to explain with our radiative transfer simulations. We consider various explanations for their origin, such as strong absorption line systems associated with the quasar or patchy reionization, but find that the most compelling scenario is that these quasars have been shining for ≲10⁵ years.

[en] Measurements of the Lyα forest based on large numbers of quasar spectra from sky surveys such as SDSS/eBOSS accurately probe the distribution of matter on small scales and thus provide important constraints on several ingredients of the cosmological model. A main summary statistic derived from those measurements is the one-dimensional power spectrum, P_1D, of the Lyα absorption. However, model predictions for P_1D rely on expensive hydrodynamical simulations of the intergalactic medium, which was the limiting factor in previous analyses. Datasets from upcoming surveys such as DESI will push observational accuracy near the 1%-level and probe even smaller scales. This observational push mandates even more accurate simulations as well as more careful exploration of parameter space. In this work we evaluate the robustness and accuracy of simulations and the statistical framework used to constrain cosmological parameters. We present a comparison between the grid-based simulation code Nyx and SPH-based code Gadget in the context of P_1D. In addition, we perform resolution and box-size convergence tests using Nyx code. We use a Gaussian process emulation scheme to reduce the number of simulations required for exploration of parameter space without sacrificing the model accuracy. We demonstrate the ability to produce unbiased parameter constraints in an end-to-end inference test using mock eBOSS- and DESI-like data, and we advocate for the usage of adaptive sampling schemes as opposed to using a fixed Latin hypercube design. (paper)

[en] Muon tomography is a technique that uses cosmic ray muons to generate three dimensional images of volumes using information contained in the Coulomb scattering of the muons. Advantages of this technique are the ability of cosmic rays to penetrate significant overburden and the absence of any additional dose delivered to subjects under study above the natural cosmic ray flux. Disadvantages include the relatively long exposure times and poor position resolution and complex algorithms needed for reconstruction. Here we demonstrate a new method for obtaining improved position resolution and statistical precision for objects with spherical symmetry.