arXiv subjects

Francisco Villaescusa-Navarro

Publications and source records attributed to Francisco Villaescusa-Navarro.

At least 163 records · Page 9Linked to original sources

arXiv2020

Super-resolution emulator of cosmological simulations using deep physical models

We present an extension of our recently developed Wasserstein optimized model to emulate accurate high-resolution features from computationally cheaper low-resolution cosmological simulations. Our deep physical modelling technique relies on restricted neural networks to perform a mapping of the distribution of the low-resolution cosmic density field to the space of the high-resolution small-scale structures. We constrain our network using a single triplet of high-resolution initial conditions and the corresponding low- and high-resolution evolved dark matter simulations from the Quijote suite of simulations. We exploit the information content of the high-resolution initial conditions as a well constructed prior distribution from which the network emulates the small-scale structures. Once fitted, our physical model yields emulated high-resolution simulations at low computational cost, while also providing some insights about how the large-scale modes affect the small-scale structure in real space.

astro-ph.CO↗

arXiv2020

Constraining $M_ν$ with the Bispectrum I: Breaking Parameter Degeneracies

Massive neutrinos suppress the growth of structure below their free-streaming scale and leave an imprint on large-scale structure. Measuring this imprint allows us to constrain the sum of neutrino masses, $M_ν$, a key parameter in particle physics beyond the Standard Model. However, degeneracies among cosmological parameters, especially between $M_ν$ and $σ_8$, limit the constraining power of standard two-point clustering statistics. In this work, we investigate whether we can break these degeneracies and constrain $\smnu$ with the next higher-order correlation function --- the bispectrum. We first examine the redshift-space halo bispectrum of $800$ $N$-body simulations from the HADES suite and demonstrate that the bispectrum helps break the $M_ν$--$σ_8$ degeneracy. Then using 22,000 $N$-body simulations of the Quijote suite, we quantify for the first time the full information content of the redshift-space halo bispectrum down to nonlinear scales using a Fisher matrix forecast of $\{Ω_m$, $Ω_b$, $h$, $n_s$, $σ_8$, $M_ν\}$. For $k_{\rm max}{=}0.5~h/{\rm Mpc}$, the bispectrum provides $Ω_m$, $Ω_b$, $h$, $n_s$, and $σ_8$ constraints 1.9, 2.6, 3.1, 3.6, and 2.6 times tighter than the power spectrum. For $M_ν$, the bispectrum improves the 1$σ$ constraint from 0.2968 to 0.0572 eV --- over 5 times tighter than the power spectrum. Even with priors from {\em Planck}, the bispectrum improves $M_ν$ constraints by a factor of 1.8. Although we reserve marginalizing over a more complete set of bias parameters to the next paper of the series, these constraints are derived for a $(1~h^{-1}{\rm Gpc})^3$ box, a substantially smaller volume than upcoming surveys. Thus, our results demonstrate that the bispectrum offers significant improvements over the power spectrum, especially for constraining $M_ν$.

astro-ph.CO↗

arXiv2020

Using the Marked Power Spectrum to Detect the Signature of Neutrinos in Large-Scale Structure

Cosmological neutrinos have their greatest influence in voids: these are the regions with the highest neutrino to dark matter density ratios. The marked power spectrum can be used to emphasize low density regions over high density regions, and therefore is potentially much more sensitive than the power spectrum to the effects of neutrino masses. Using 22,000 N-body simulations from the Quijote suite, we quantify the information content in the marked power spectrum of the matter field, and show that it outperforms the standard power spectrum by setting constraints improved by a factor larger than 2 on all cosmological parameters. The combination of marked and standard power spectrum allows to place a 4.3σ constraint on the minimum sum of the neutrino masses with a volume equal to 1 (Gpc/h)^3 and without CMB priors. Combinations of different marked power spectra yield a 6σ constraint within the same conditions.

astro-ph.CO↗

arXiv2019

Fisher for complements: Extracting cosmology and neutrino mass from the counts-in-cells PDF

We comprehensively analyse the cosmology dependence of counts-in-cell statistics. We focus on the shape of the one-point probability distribution function (PDF) of the matter density field at mildly nonlinear scales. Based on large-deviation statistics, we parametrise the cosmology dependence of the matter PDF in terms of the linear power spectrum, the growth factor, the spherical collapse dynamics, and the nonlinear variance. We extend our formalism to include massive neutrinos, finding that the total matter PDF is highly sensitive to the total neutrino mass $M_ν$ and can disentangle it from the clustering amplitude $σ_8$. Using more than a million PDFs extracted from the Quijote simulations, we determine the response of the matter PDF to changing parameters in the $νΛ$CDM model and successfully cross-validate the theoretical model and the simulation measurements. We present the first $νΛ$CDM Fisher forecast for the matter PDF at multiple scales and redshifts, and its combination with the matter power spectrum. We establish that the matter PDF and the matter power spectrum are highly complementary at mildly nonlinear scales. The matter PDF is particularly powerful for constraining the matter density $Ω_m$, clustering amplitude $σ_8$ and the total neutrino mass $M_ν$. Adding the mildly nonlinear matter PDF to the mildly nonlinear matter power spectrum improves constraints on $Ω_m$ by a factor of 5 and $σ_8$ by a factor of 2 when considering the three lowest redshifts. In our joint analysis of the matter PDF and matter power spectrum at three redshifts, the total neutrino mass is constrained to better than 0.01 eV with a total volume of 6 (Gpc/h)$^3$. We discuss how density-split statistics can be used to translate those encouraging results for the matter PDF into realistic observables in galaxy surveys.

astro-ph.CO↗

arXiv2019

BE-HaPPY: Bias Emulator for Halo Power Spectrum including massive neutrinos

We study the clustering properties of dark matter halos in real- and redshift-space in cosmologies with massless and massive neutrinos through a large set of state-of-the-art N-body simulations. We provide quick and easy-to-use prescriptions for the halo bias on linear and mildly non-linear scales, both in real and redshift space, which are valid also for massive neutrinos cosmologies. Finally we present a halo bias emulator,$\textbf{BE-HaPPY}$, calibrated on the N-body simulations, which is fast enough to be used in the standard Markov Chain Monte Carlo approach to cosmological inference. For a fiducial standard $Λ$CDM cosmology $\textbf{BE-HaPPY}$ provides percent or sub-percent accuracy on the scales of interest (linear and well into the mildly non-linear regime), meeting therefore for the halo-bias the accuracy requirements for the analysis of next-generation large--scale structure surveys.

astro-ph.CO↗

arXiv2019

From Dark Matter to Galaxies with Convolutional Neural Networks

Cosmological simulations play an important role in the interpretation of astronomical data, in particular in comparing observed data to our theoretical expectations. However, to compare data with these simulations, the simulations in principle need to include gravity, magneto-hydrodyanmics, radiative transfer, etc. These ideal large-volume simulations (gravo-magneto-hydrodynamical) are incredibly computationally expensive which can cost tens of millions of CPU hours to run. In this paper, we propose a deep learning approach to map from the dark-matter-only simulation (computationally cheaper) to the galaxy distribution (from the much costlier cosmological simulation). The main challenge of this task is the high sparsity in the target galaxy distribution: space is mainly empty. We propose a cascade architecture composed of a classification filter followed by a regression procedure. We show that our result outperforms a state-of-the-art model used in the astronomical community, and provides a good trade-off between computational cost and prediction accuracy.

astro-ph.CO↗

arXiv2019

Inflation and Early Dark Energy with a Stage II Hydrogen Intensity Mapping Experiment

This white paper envisions a revolutionary post-DESI, post-LSST dark energy program based on intensity mapping of the redshifted 21cm emission line from neutral hydrogen at radio frequencies. The proposed intensity mapping survey has the unique capability to quadruple the volume of the Universe surveyed by optical programs, provide a percent-level measurement of the expansion history to $z \sim 6$, open a window to explore physics beyond the concordance $Λ$CDM model, and to significantly improve the precision on standard cosmological parameters. In addition, characterization of dark energy and new physics will be powerfully enhanced by cross-correlations with optical surveys and cosmic microwave background measurements. The rich dataset obtained by the proposed intensity mapping instrument will be simultaneously useful in exploring the time-domain physics of fast radio transients and pulsars, potentially in live "multi-messenger" coincidence with other observatories. The core dark energy/inflation science advances enabled by this program are the following: (i) Measure the expansion history of the universe over $z=0.3-6$ with a single instrument, extending the range deep into the pre-acceleration era, providing an unexplored window for new physics; (ii) Measure the growth rate of structure in the universe over the same redshift range; (iii) Observe, or constrain, the presence of inflationary relics in the primordial power spectrum, improving existing constraints by an order of magnitude; (iv) Observe, or constrain, primordial non-Gaussianity with unprecedented precision, improving constraints on several key numbers by an order of magnitude. Detailed mapping of the enormous, and still largely unexplored, volume of cosmic space will thus provide unprecedented information on fundamental questions of the vacuum energy and early-universe physics.

astro-ph.CO↗

arXiv2019

Packed Ultra-wideband Mapping Array (PUMA): A Radio Telescope for Cosmology and Transients

PUMA is a proposal for an ultra-wideband, low-resolution and transit interferometric radio telescope operating at $200-1100\,\mathrm{MHz}$. Its design is driven by six science goals which span three science themes: the physics of dark energy (measuring the expansion history and growth of the universe up to $z=6$), the physics of inflation (constraining primordial non-Gaussianity and primordial features) and the transient radio sky (detecting one million fast radio bursts and following up SKA-discovered pulsars). We propose two array configurations composed of hexagonally close-packed 6m dish arrangements with 50% fill factor. The initial 5,000 element 'petite array' is scientifically compelling, and can act as a demonstrator and a stepping stone to the full 32,000 element 'full array'. Viewed as a 21cm intensity mapping telescope, the program has the noise equivalent of a traditional spectroscopic galaxy survey comprised of 0.6 and 2.5 billion galaxies at a comoving wavenumber of $k=0.5\,h\mathrm{Mpc}^{-1}$ spanning the redshift range $z = 0.3 - 6$ for the petite and full configurations, respectively. At redshifts beyond $z=2$, the 21cm technique is a uniquely powerful way of mapping the universe, while the low-redshift range will allow for numerous cross-correlations with existing and upcoming surveys. This program is enabled by the development of ultra-wideband radio feeds, cost-effective dish construction methods, commodity radio-frequency electronics driven by the telecommunication industry and the emergence of sufficient computing power to facilitate real-time signal processing that exploits the full potential of massive radio arrays. The project has an estimated construction cost of 55 and 330 million FY19 USD for the petite and full array configurations. Including R&D, design, operations and science analysis, the cost rises to 125 and 600 million FY19 USD, respectively.

astro-ph.IM↗

arXiv2019

Research and Development for HI Intensity Mapping

Development of the hardware, data analysis, and simulation techniques for large compact radio arrays dedicated to mapping the 21 cm line of neutral hydrogen gas has proven to be more difficult than imagined twenty years ago when such telescopes were first proposed. Despite tremendous technical and methodological advances, there are several outstanding questions on how to optimally calibrate and analyze such data. On the positive side, it has become clear that the outstanding issues are purely technical in nature and can be solved with sufficient development activity. Such activity will enable science across redshifts, from early galaxy evolution in the pre-reionization era to dark energy evolution at low redshift.

astro-ph.IM↗

arXiv2019

Weighing neutrinos with the halo environment

Nonlinear objects like halos and voids exhibit a scale-dependent bias on linear scales in massive neutrino cosmologies. The shape of this scale-dependent bias is a unique signature of the neutrino masses, but the amplitude of the signal is generally small, of the order of $f_ν$, the contribution of neutrinos to the total matter content ($\lesssim 1\%$). In this paper, we demonstrate for the first time how the strength of this signal can be substantially enhanced by using information about the halo environment at a range of scales. This enhancement is achieved by using certain combinations of the large scale Cold Dark Matter and total matter environments of halos, both of which are measurable from galaxy clustering and weak lensing surveys.

astro-ph.CO↗

arXiv2019

Atomic and molecular gas in IllustrisTNG galaxies at low redshift

We have recently developed a post-processing framework to estimate the abundance of atomic and molecular hydrogen (HI and H2, respectively) in galaxies in large-volume cosmological simulations. Here we compare the HI and H2 content of IllustrisTNG galaxies to observations. We mostly restrict this comparison to $z \approx 0$ and consider six observational metrics: the overall abundance of HI and H2, their mass functions, gas fractions as a function of stellar mass, the correlation between H2 and star formation rate, the spatial distribution of gas, and the correlation between gas content and morphology. We find generally good agreement between simulations and observations, particularly for the gas fractions and the HI mass-size relation. The H2 mass correlates with star formation rate as expected, revealing an almost constant depletion time that evolves up to z = 2 as observed. However, we also discover a number of tensions with varying degrees of significance, including an overestimate of the total neutral gas abundance at z = 0 by about a factor of two and a possible excess of satellites with no or very little neutral gas. These conclusions are robust to the modelling of the HI/H2 transition. In terms of their neutral gas properties, the IllustrisTNG simulations represent an enormous improvement over the original Illustris run. All data used in this paper are publicly available as part of the IllustrisTNG data release.

astro-ph.GA↗

arXiv2019

HIGAN: Cosmic Neutral Hydrogen with Generative Adversarial Networks

One of the most promising ways to observe the Universe is by detecting the 21cm emission from cosmic neutral hydrogen (HI) through radio-telescopes. Those observations can shed light on fundamental astrophysical questions only if accurate theoretical predictions are available. In order to maximize the scientific return of these surveys, those predictions need to include different observables and be precise on non-linear scales. Currently, one of the best ways to achieve this is via cosmological hydrodynamic simulations; however, the computational cost of these simulations is high -- tens of millions of CPU hours. In this work, we use Wasserstein Generative Adversarial Networks (WGANs) to generate new high-resolution ($35~h^{-1}{\rm kpc}$) 3D realizations of cosmic HI at $z=5$. We do so by sampling from a 100-dimension manifold, learned by the generator, that characterizes the fully non-linear abundance and clustering of cosmic HI from the state-of-the-art simulation IllustrisTNG. We show that different statistical properties of the produced samples -- 1D PDF, power spectrum, bispectrum, and void size function -- match very well those of IllustrisTNG, and outperform state-of-the-art models such as Halo Occupation Distributions (HODs). Our WGAN samples reproduce the abundance of HI across 9 orders of magnitude, from the Ly$α$ forest to Damped Lyman Absorbers. WGAN can produce new samples orders of magnitude faster than hydrodynamic simulations.

astro-ph.CO↗

arXiv2019

From Dark Matter to Galaxies with Convolutional Networks

Cosmological surveys aim at answering fundamental questions about our Universe, including the nature of dark matter or the reason of unexpected accelerated expansion of the Universe. In order to answer these questions, two important ingredients are needed: 1) data from observations and 2) a theoretical model that allows fast comparison between observation and theory. Most of the cosmological surveys observe galaxies, which are very difficult to model theoretically due to the complicated physics involved in their formation and evolution; modeling realistic galaxies over cosmological volumes requires running computationally expensive hydrodynamic simulations that can cost millions of CPU hours. In this paper, we propose to use deep learning to establish a mapping between the 3D galaxy distribution in hydrodynamic simulations and its underlying dark matter distribution. One of the major challenges in this pursuit is the very high sparsity in the predicted galaxy distribution. To this end, we develop a two-phase convolutional neural network architecture to generate fast galaxy catalogues, and compare our results against a standard cosmological technique. We find that our proposed approach either outperforms or is competitive with traditional cosmological techniques. Compared to the common methods used in cosmology, our approach also provides a nice trade-off between time-consumption (comparable to fastest benchmark in the literature) and the quality and accuracy of the predicted simulation. In combination with current and upcoming data from cosmological observations, our method has the potential to answer fundamental questions about our Universe with the highest accuracy.

astro-ph.CO↗

arXiv2019

Cosmic voids: a novel probe to shed light on our Universe

Cosmic voids, the less dense patches of the Universe, are promising laboratories to extract cosmological information. Thanks to their unique low density character, voids are extremely sensitive to diffuse components such as neutrinos and dark energy, and represent ideal environments to study modifications of gravity, where the effects of such modifications are expected to be more prominent. Robust void-related observables, including for example redshift-space distortions (RSD) and weak lensing around voids, are a promising way to chase and test new physics. Cosmological analysis of the large-scale structure of the Universe predominantly relies on the high density regions. Current and upcoming surveys are designed to optimize the extraction of cosmological information from these zones, but leave voids under-exploited. A dense, large area spectroscopic survey with imaging capabilities is ideal to exploit the power of voids fully. Besides helping illuminate the nature of dark energy, modified gravity, and neutrinos, this survey will give access to a detailed map of under-dense regions, providing an unprecedented opportunity to observe and study a so far under-explored galaxy population.

astro-ph.CO↗

arXiv2019

Neutrino Mass from Cosmology: Probing Physics Beyond the Standard Model

Recent advances in cosmic observations have brought us to the verge of discovery of the absolute scale of neutrino masses. Nonzero neutrino masses are known evidence of new physics beyond the Standard Model. Our understanding of the clustering of matter in the presence of massive neutrinos has significantly improved over the past decade, yielding cosmological constraints that are tighter than any laboratory experiment, and which will improve significantly over the next decade, resulting in a guaranteed detection of the absolute neutrino mass scale.

astro-ph.CO↗

arXiv2019

Measuring the EoR Power Spectrum Without Measuring the EoR Power Spectrum

The large-scale structure of the Universe should soon be measured at high redshift during the Epoch of Reionization (EoR) through line-intensity mapping. A number of ongoing and planned surveys are using the 21 cm line to trace neutral hydrogen fluctuations in the intergalactic medium (IGM) during the EoR. These may be fruitfully combined with separate efforts to measure large-scale emission fluctuations from galactic lines such as [CII], CO, H-$α$, and Ly-$α$ during the same epoch. The large scale power spectrum of each line encodes important information about reionization, with the 21 cm power spectrum providing a relatively direct tracer of the ionization history. Here we show that the large scale 21 cm power spectrum can be extracted using only cross-power spectra between the 21 cm fluctuations and each of two separate line-intensity mapping data cubes. This technique is more robust to residual foregrounds than the usual 21 cm auto-power spectrum measurements and so can help in verifying auto-spectrum detections. We characterize the accuracy of this method using numerical simulations and find that the large-scale 21 cm power spectrum can be inferred to an accuracy of within 5% for most of the EoR, reaching 0.6% accuracy on a scale of $k\sim0.1\,\text{Mpc}^{-1}$ at $\left< x_i \right> = 0.36$ ($z = 8.34$ in our model). An extension from two to $N$ additional lines would provide $N(N-1)/2$ cross-checks on the large-scale 21 cm power spectrum. This work strongly motivates redundant line-intensity mapping surveys probing the same cosmological volumes.

astro-ph.CO↗

arXiv2019

First detection of scale-dependent linear halo bias in $N$-body simulations with massive neutrinos

Using $N$-body simulations with massive neutrino density perturbations, we detect the scale-dependent linear halo bias with high significance. This is the first time that this effect is detected in simulations containing neutrino density perturbations on all scales, confirming the same finding from separate universe simulations. The scale dependence is the result of the additional scale in the system, i.e. the massive neutrino free-streaming length, and it persists even if the bias is defined with respect to the cold dark matter plus baryon (instead of total matter) power spectrum. The separate universe approach provides a good model for the scale-dependent linear bias, and the effect is approximately $0.25f_ν$ and $0.43f_ν$ for halos with bias of 1.7 and 3.5, respectively. While the size of the effect is small, it is ${\it not}$ insignificant in terms of $f_ν$ and should therefore be included to accurately constrain neutrino mass from clustering statistics of biased tracers. More importantly, this feature is a distinct signature of free-streaming particles and cannot be mimicked by other components of the standard cosmological model.

astro-ph.CO↗

arXiv2018

Dipole Distortions in the Intergalactic Medium

Baryonic feedback can significantly modify the spatial distribution of matter on small scales and create a bulk relative velocity between the dominant cold dark matter and the hot gas. We study the consequences of such bulk motions using two high resolution hydrodynamic simulations, one with no feedback and one with very strong feedback. We find that relative velocities of order $100\ \kms$ are produced in the strong feedback simulation whereas it is much smaller when there is no feedback. Such relative motions induce dipole distortions to the gas, which we quantify by computing the dipole correlation function. We find halo coordinates and velocities are systematically changed in the direction of the relative velocity. Finally, we discuss potential to observe the relative velocity via large scale structure, Sunyaev-Zel'dovich and line emission measurements. Given the nonlinear nature of this effect, it should next be studied in simulations with different feedback implementations/strengths to determine the available model space.

astro-ph.CO↗