arXiv subjects

Naoki Yoshida

Publications and source records attributed to Naoki Yoshida.

At least 91 records · Page 5Linked to original sources

arXiv2021

Cosmological inference from the emulator based halo model II: Joint analysis of galaxy-galaxy weak lensing and galaxy clustering from HSC-Y1 and SDSS

We present high-fidelity cosmology results from a blinded joint analysis of galaxy-galaxy weak lensing ($Δ\!Σ$) and projected galaxy clustering ($w_{\rm p}$) measured from the Hyper Suprime-Cam Year-1 (HSC-Y1) data and spectroscopic Sloan Digital Sky Survey (SDSS) galaxy catalogs in the redshift range $0.15<z<0.7$. We define luminosity-limited samples of SDSS galaxies to serve as the tracers of $w_{\rm p}$ in three spectroscopic redshift bins, and as the lens samples for $Δ\!Σ$. For the $Δ\!Σ$ measurements, we select a single sample of 4 million source galaxies over 140 deg$^2$ from HSC-Y1 with photometric redshifts (photo-$z$) greater than 0.75, enabling a better handle of photo-$z$ errors by comparing the $Δ\!Σ$ amplitudes for the three lens redshift bins. For cosmological parameter inference, we use an input galaxy-halo connection model built on the {\tt Dark Emulator} package with a halo occupation distribution that includes nuisance parameters to marginalize over modeling uncertainties. We model the $Δ\!Σ$ and $w_{\rm p}$ measurements on scales from $R\simeq 3$ and $2\,h^{-1}{\rm Mpc}$, respectively, up to $30\,h^{-1}{\rm Mpc}$ assuming a flat $Λ$CDM cosmology. With various tests using mock catalogs described in Miyatake et al. (2021), we show that any bias in the clustering amplitude $S_8\equiv σ_8(Ω_{\rm m}/0.3)^{0.5}$ due to uncertainties in the galaxy-halo connection is less than $\sim50$\% of the statistical uncertainty of $S_8$, {\it unless} the assembly bias effect is unexpectedly large. Our best-fit models have $S_8=0.795^{+0.049}_{-0.042}$ (mode and 68\% credible interval) for the flat $Λ$CDM model; we find tighter constraints on the quantity $S_8(α=0.17)\equivσ_8(Ω_{\rm m}/0.3)^{0.17} =0.745^{+0.039}_{-0.031}$. (abriged)

astro-ph.CO↗

arXiv2021

Deep learning reconstruction of three-dimensional galaxy distributions with intensity mapping observations

Line intensity mapping is emerging as a novel method that can measure the collective intensity fluctuations of atomic/molecular line emission from distant galaxies. Several observational programs with various wavelengths are ongoing and planned, but there remains a critical problem of line confusion; emission lines originating from galaxies at different redshifts are confused at the same observed wavelength. We devise a generative adversarial network that extracts designated emission line signals from noisy three-dimensional data. Our novel network architecture allows two input data, in which the same underlying large-scale structure is traced by two emission lines of H$\rm α$ and [OIII], so that the network learns the relative contributions at each wavelength and is trained to decompose the respective signals. After being trained with a large number of realistic mock catalogs, the network is able to reconstruct the three-dimensional distribution of emission-line galaxies at $z = 1.3-2.4$. Bright galaxies are identified with a precision of 84%, and the cross-correlation coefficients between the true and reconstructed intensity maps are as high as 0.8. Our deep-learning method can be readily applied to data from planned space-borne and ground-based experiments.

astro-ph.GA↗

arXiv2021

A 400 Trillion-Grid Vlasov Simulation on Fugaku Supercomputer: Large-Scale Distribution of Cosmic Relic Neutrinos in a Six-dimensional Phase Space

We report a Vlasov simulation of cosmic relic neutrinos combined with N-body simulation of cold dark matter in the context of large-scale structure formation in the Universe performed on Fugaku supercomputer. Gravitational dynamics of the neutrinos is followed, for the first time, by directly integrating the Vlasov equation in a six-dimensional phase space. Our largest simulation combines the Vlasov simulation on 400 trillion grids and 330 billion-body calculations in a self-consistent manner, and reproduces accurately the nonlinear dynamics of neutrinos in the Universe. The novel high-order Vlasov solver is optimized by combining an array of state-of-the-art numerical schemes and fully utilizing the SIMD instructions on the A64FX processors. Time-To-Solution of our simulation is an order of magnitude shorter than the largest N-body simulations. The performance scales excellently with up to 147,456 nodes (7 million CPU cores) on Fugaku; the weak and strong scaling efficiencies are 82% - 96% and 82% - 93%, respectively.

astro-ph.CO↗

arXiv2021

Internal r-process abundance spread of M15 and a single stellar population model

The member stars in globular cluster M15 show a substantial spread in the abundances of r-process elements. We argue that a rare and prolific r-process event enriched the natal cloud of M15 in an inhomogeneous manner. To critically examine the possibility, we perform cosmological galaxy formation simulations and study the physical conditions for the inhomogeneous enrichment. We explore a large parameter space of the merger event time and the site. Our simulations reproduce the large r-process abundance spread if a neutron-star merger occurs at \sim 100 pc away from the formation site of the cluster and in a limited time range of a few tens million years before the formation. Interestingly, a bimodal feature is found in the Eu abundance distribution in some cases, similarly to that inferred from recent observations. M15 member stars do not show clear correlation between the abundances of Eu and light elements such as Na that is expected in models with two stellar populations. We thus argue that a majority of the stars in M15 are formed in a single burst. The ratio of heavy to light r-process element abundance [Eu/Y] \sim 1.0 is consistent with that of the so-called r-II stars, suggesting that a lanthanide-rich r-process event dominantly enriched M15.

astro-ph.GA↗

arXiv2021

Disc fragmentation and oligarchic growth of protostellar systems in low-metallicity gas clouds

We study low-metallicity star formation with a set of high-resolution hydrodynamics simulations for various gas metallicities over a wide range $0$--$10^{-3} \ {\rm Z}_{\bigodot}$. Our simulations follow non-equilibrium chemistry and radiative cooling by adopting realistic elemental abundance and dust size distribution. We examine the condition for the fragmentation of collapsing clouds (cloud fragmentation; CF) and of accretion discs (disc fragmentation; DF). We find that CF is suppressed due to rapid gas heating accompanied with molecular hydrogen formation even with efficient dust cooling for metallicities $\gtrsim 10^{-5} \ {\rm Z}_{\bigodot}$. Instead, DF occurs in almost all runs regardless of metallicity. We also find that, in the accretion discs, the growth of the protostellar systems is overall oligarchic. The primary protostar grows through the accretion of gas, and secondary protostars form through the interaction of spiral arms or the break-up of a rapidly rotating protostar. Despite vigorous fragmentation, a large fraction of secondary protostars are destroyed through mergers or tidal disruption events. For a few hundred years after the first adiabatic core formation, only several protostars survive in the disc, and the total mass of protostars is $0.52$--$3.8 \ {\rm M}_{\bigodot}$.

astro-ph.GA↗

arXiv2021

Fragmentation of ring galaxies and transformation to clumpy galaxies

We study the fragmentation of collisional ring galaxies (CRGs) using a linear perturbation analysis that computes the physical conditions of gravitational instability, as determined by the balance of self-gravity of the ring against pressure and Coriolis forces. We adopt our formalism to simulations of CRGs and show that the analysis can accurately characterise the stability and onset of fragmentation, although the linear theory appears to under-predict the number of fragments of an unstable CRG by a factor of 2. In addition, since the orthodox `density-wave' model is inapplicable to such self-gravitating rings, we devise a simple approach that describes the rings propagating as material waves. We find that the toy model can predict whether the simulated CRGs fragment or not using information from their pre-collision states. We also apply our instability analysis to a CRG discovered at a high redshift, $z=2.19$. We find that a quite high velocity dispersion is required for the stability of the ring, and therefore the CRG should be unstable to ring fragmentation. CRGs are rarely observed at high redshifts, and this may be because CRGs are usually too faint. Since the fragmentation can induce active star formation and make the ring bright enough to observe, the instability could explain this rarity. An unstable CRG fragments into massive clumps retaining the initial disc rotation, and thus it would evolve into a clumpy galaxy with a low surface density in an inter-clump region.

astro-ph.GA↗

arXiv2021

Noise reduction for weak lensing mass mapping: An application of generative adversarial networks to Subaru Hyper Suprime-Cam first-year data

We propose a deep-learning approach based on generative adversarial networks (GANs) to reduce noise in weak lensing mass maps under realistic conditions. We apply image-to-image translation using conditional GANs to the mass map obtained from the first-year data of Subaru Hyper Suprime-Cam (HSC) survey. We train the conditional GANs by using 25000 mock HSC catalogues that directly incorporate a variety of observational effects. We study the non-Gaussian information in denoised maps using one-point probability distribution functions (PDFs) and also perform matching analysis for positive peaks and massive clusters. An ensemble learning technique with our GANs is successfully applied to reproduce the PDFs of the lensing convergence. About $60\%$ of the peaks in the denoised maps with height greater than $5σ$ have counterparts of massive clusters within a separation of 6 arcmin. We show that PDFs in the denoised maps are not compromised by details of multiplicative biases and photometric redshift distributions, nor by shape measurement errors, and that the PDFs show stronger cosmological dependence compared to the noisy counterpart. We apply our denoising method to a part of the first-year HSC data to show that the observed mass distribution is statistically consistent with the prediction from the standard $Λ$CDM model.

astro-ph.CO↗

arXiv2021

s-process Enirchment of Ultrafaint Dwarf Galaxies

We study the production of barium (Ba) and strontium (Sr) in ultrafaint dwarf galaxies (UFDs). Both r- and s- processes produce these elements, and one can infer the contribution of the r-process from the characteristic r-process abundance pattern, whereas the s-process contribution remains largely unknown. We show that the current s-process yield from asymptotic giant branch (AGB) stars is not sufficient to explain the Ba and Sr abundances observed in UFDs. Production of these elements would need to be efficient from the beginning of star formation in the galaxies. The discrepancy of nearly or more than 1 dex is not reconciled even if we consider s-process in super-AGB stars. We consider a possible resolution by assuming rotating massive stars (RMSs) and electron-capture supernovae (ECSNe) as additional contributors. We find that the RMSs could be the origin of Ba in UFDs if ~ 10 per cent of massive stars are rotating at 300 km s^{-1}. As for ECSNe, we argue that their fraction is less than 2 per cent of core-collapse supernova. It narrows the progenitor mass-range to < 0.1 Msun at -3 < [Fe/H] < -2. We also explore another resolution by modifying the stellar initial mass function (IMF) in UFDs and find a top-light IMF model that reproduces the observed level of Ba-enrichment. Future observations that determine or tightly constrain the europium and nitrogen abundances are crucial to identify the origin of Ba and Sr in UFDs.

astro-ph.GA↗

arXiv2021

Formation of an extended stellar halo around an ultra-faint dwarf galaxy following one of the earliest mergers from galactic building blocks

Ultra-faint dwarf galaxies (UFDs) are promising observable proxies to building blocks of galaxies formed in the early Universe. We study the formation and evolution of UFDs using cosmological hydrodynamic simulations. In particular, we show that a major merger of two building block galaxies with 3,900 Msun and 7,500 Msun at the cosmic age of 510 Myr results in a system with an extended stellar distribution consistent with the de Vaucouleurs profile. The simulated galaxy has an average stellar metallicity of [Fe/H]=-2.7 and features a metallicity gradient. These results closely resemble the properties of a recently discovered UFD, Tucana II, which is extremely metal-poor and has a spatially extended stellar halo with the more distant stars being more metal-poor. Our simulation suggests that the extended stellar halo of Tucana II may have been formed through a past major merger. Future observational searches for spatially extended structures around other UFDs, combined with further theoretical studies, will provide tangible measures of the evolutionary history of the ancient, surviving satellite galaxies.

astro-ph.GA↗

arXiv2021

Formation of massive stars under protostellar radiation feedback: Very metal-poor stars

We study the formation of very metal-poor stars under protostellar radiative feedback effect. We use cosmological simulations to identify low-mass dark matter halos and star-forming gas clouds within them. We then follow protostar formation and the subsequent long-term mass accretion phase of over one million years using two-dimensional radiation-hydrodynamics simulations. We show that the critical physical process that sets the final mass is formation and expansion of a bipolar HII region. The process is similar to the formation of massive primordial stars, but radiation pressure exerted on dust grains also contributes to halting the accretion flow in the low-metallicity case. We find that the net feedback effect in the case with metallicity $Z = 10^{-2}~Z_{\odot}$ is stronger than in the case with $Z \sim 1~Z_{\odot}$. With decreasing metallicity, the radiation pressure effect becomes weaker, but photoionization heating of the circumstellar gas is more efficient owing to the reduced dust attenuation. In the case with $Z = 10^{-2}~Z_{\odot}$, the central star grows as massive as 200 solar-masses, similarly to the case of primordial star formation. We conclude that metal-poor stars with a few hundred solar masses can be formed by gas accretion despite the strong radiative feedback.

astro-ph.GA↗

arXiv2021

Atmospheric Escape of Close-in Giants around Hot Stars: Far-Ultraviolet Radiation and Photoelectric Heating Effect

Atmospheric escape is an important process that controls the long-term evolution of close-in planets. We perform radiation hydrodynamics simulations of photo-evaporation of exoplanets' atmospheres to study the effect of photoelectric heating by far-ultraviolet (FUV) radiation. Specifically, we consider a close-in hot Jupiter around a hot A-star. Hot main-sequence stars emit not only extreme ultraviolet radiation but also FUV radiation, and thus can drive strong atmospheric escape by photoelectric heating. We show that the planetary atmosphere escapes at a rate as large as $\dot{M}\sim10^{14}\, \mathrm{g}~{\rm sec}^{-1}$ if the atmosphere contains a small amount of dust grains with the level of ten percent of the local interstellar medium. Close-in planets around hot stars can lose a significant fraction of the atmosphere during the long-term evolution. We quantify the amount of dust necessary for causing photoevaporation. The dust-to-gas mass ratio of $10^{-4}$ is sufficient to drive stronger atmospheric escape by FUV photoelectric heating than in the case with only extreme ultraviolet radiation. We also explore the metallicity dependence of the FUV-driven escape. The mass-loss rate increases with increasing the atmosphere's metallicity because of the enhanced photoelectric heating, but the stellar FUV flux decreases with increasing stellar metallicity. We derive an accurate estimate for the mass-loss rate as a function of FUV flux and metallicity, and of the planet's characteristics. The FUV driven atmospheric escape may be a key process to understand and explain the so-called sub-Jovian desert.

astro-ph.EP↗

arXiv2021

Radiation hydrodynamics simulations of protoplanetary disks: Stellar mass dependence of the disk photoevaporation rate

Recent multi-wavelength observations suggest that inner parts of protoplanetary disks (PPDs) have shorter lifetimes for heavier host stars. Since PPDs around high-mass stars are irradiated by strong ultra-violet radiation, photoevaporation may provide an explanation for the observed trend. We perform radiation hydrodynamics simulations of photoevaporation of PPDs for a wide range of host star mass of $M_* =0.5$-$7.0 M_{\odot}$. We derive disk mass-loss rate $\dot{M}$, which has strong stellar dependence as $\dot{M} \approx 7.30\times10^{-9}(M_{*}/M_{\odot})^{2}M_{\odot}\rm{yr}^{-1}$. The absolute value of $\dot{M}$ scales with the adopted far-ultraviolet and X-ray luminosities. We derive the surface mass-loss rates and provide polynomial function fits to them. We also develop a semi-analytic model that well reproduces the derived mass-loss rates. The estimated inner disk lifetime decreases as the host star mass increases, in agreement with the observational trend. We thus argue that photoevaporation is a major physical mechanism for PPD dispersal for a wide range of the stellar mass and can account for the observed stellar mass dependence of the inner disk lifetime.

astro-ph.EP↗

arXiv2021

Capturing the inside-out quenching by black holes with far-infrared atomic line ratios

We propose to use relative strengths of far-infrared fine structure lines from galaxies to characterise early phases of the inside-out quenching by massive black holes (BHs). The BH feedback is thought to quench star formation by evacuating the ambient gas. In order to quantify the feedback effect on the gas density in the galactic centres, we utilise the outputs of IllustrisTNG and Illustris simulations, which implement different BH feedback models. We devise a physical model of H$_{\rm ~II}$ regions and compute the intensities of [O$_{\rm ~III}$] $52$ and $88~{\rm μm}$ lines. The line intensity ratio is sensitive to the local electron density, and thus can be used to measure the strength and physical extent of the BH quenching. If the BH feedback abruptly operates and expel the gas when it grows to a certain mass, as modelled in IllustrisTNG, the low-density gas yields relatively weak [O$_{\rm ~III}$] $52$ line with respect to $88~{\rm μm}$. In contrast, if the feedback strength and hence the local gas density are not strongly correlated with the BH mass, as in Illustris, the line ratio is not expected to vary significantly among galaxies with different evolutionary stages. We find these features are reproduced in the simulations. We also show that the line ratios are not sensitive to the aperture size for measurement, and thus observations do not need to resolve the galactic centres. We argue that the integrated line ratios can be used to capture the onset of the inside-out quenching by BHs.

astro-ph.GA↗

arXiv2021

Large-scale variation in reionization history caused by Baryon-dark matter streaming velocity

At cosmic recombination, there was supersonic relative motion between baryons and dark matter, which originated from the baryonic acoustic oscillations in the early universe. This motion has been considered to have a negligible impact on the late stage of cosmic reionization because the relative velocity quickly decreases. However, recent studies have suggested that the recombination in gas clouds smaller than the local Jeans mass ($\lesssim$ $10^8~M_\odot$) can affect the reionization history by boosting the number of ultraviolet photons required for ionizing the intergalactic medium. Motivated by this, we performed a series of radiation-hydrodynamic simulations to investigate whether the streaming motion can generate variation in the local reionization history by smoothing out clumpy small-scale structures and lowering the ionizing photon budget. We found that the streaming velocity can add a variation of $Δz_e$ $\sim$ $0.05$ $-$ $0.5$ in the end-of-reionization redshift, depending on the level of X-ray preheating and the time evolution of ionizing sources. The variation tends to be larger when the ionizing efficiency of galaxies decreases toward later times. Given the long spatial fluctuation scales of the streaming motion ($\gtrsim 100$ Mpc), it can help to explain the Ly$α$ opacity variation observed from quasars and leave large-scale imprints on the ionization field of the intergalactic medium during the reionization. The pre-reionization heating by X-ray sources is another critical factor that can suppress small-scale gas clumping and can diminish the variation in $z_e$ introduced by the streaming motion.

astro-ph.CO↗

arXiv2020

Deep Learning for Line Intensity Mapping Observations: Information Extraction from Noisy Maps

Line intensity mapping (LIM) is a promising observational method to probe large-scale fluctuations of line emission from distant galaxies. Data from wide-field LIM observations allow us to study the large-scale structure of the universe as well as galaxy populations and their evolution. A serious problem with LIM is contamination by foreground/background sources and various noise contributions. We develop conditional generative adversarial networks (cGANs) that extract designated signals and information from noisy maps. We train the cGANs using 30,000 mock observation maps with assuming a Gaussian noise matched to the expected noise level of NASA's SPHEREx mission. The trained cGANs successfully reconstruct Hα emission from galaxies at a target redshift from observed, noisy intensity maps. Intensity peaks with heights greater than 3.5 σ noise are located with 60 % precision. The one-point probability distribution and the power spectrum are accurately recovered even in the noise-dominated regime. However, the overall reconstruction performance depends on the pixel size and on the survey volume assumed for the training data. It is necessary to generate training mock data with a sufficiently large volume in order to reconstruct the intensity power spectrum at large angular scales. Our deep-learning approach can be readily applied to observational data with line confusion and with noise.

astro-ph.GA↗

arXiv2020

Cosmological Vlasov-Poisson Simulations of Structure Formation with Relic Neutrinos: Nonlinear Clustering and the Neutrino Mass

We present the results of cosmological simulations of large-scale structure formation with massive neutrinos. The phase-space distribution of the cosmic relic neutrinos is followed, for the first time, by directly integrating the six-dimensional Vlasov-Poisson equations. Our novel approach allows us to represent free streaming and clustering of neutrinos, and their gravitational interaction with cold dark matter accurately. We thus obtain solutions for the collisionless dynamics independent of conventional N-body methods. We perform a suite of hybrid N-body/Vlasov simulations with varying the neutrino mass, and systematically examine the dynamical effects of massive neutrinos on large-scale structure formation. Our simulations show characteristic large-scale clustering of the neutrinos and their coherent streaming motions relative to dark matter. The effective local neutrino "temperature" around massive galaxy clusters varies by several percent with respect to the cosmic mean; the neutrinos in clusters can be hotter or colder depending on the neutrino mass. We study a number of statistics of the large-scale structure and of dark matter halos in comparison with those obtained by N-body simulations and/or by perturbation theory. Our simulations mark an important milestone in numerical cosmology, and pave a new way to study cosmic structure formation with massive neutrinos.

astro-ph.CO↗

arXiv2020

The CO universe: Modelling CO emission and H$_{\rm 2}$ abundance in cosmological galaxy formation simulations

We devise a physical model of formation and distribution of molecular gas clouds in galaxies. We use the model to predict the intensities of rotational transition lines of carbon monoxide (CO) and the molecular hydrogen (H$_{\rm 2}$) abundance. Using the outputs of Illustris-TNG cosmological simulations, we populate molecular gas clouds of unresolved sizes in individual simulated galaxies, where the effect of the interstellar radiation field with dust attenuation is also taken into account. We then use the publicly available code DESPOTIC to compute the CO line luminosities and H$_{\rm 2}$ densities without assuming the CO-to-H$_{\rm 2}$ conversion factor ($α_{\rm CO}$). Our method allows us to study the spatial and kinematic structures traced by CO(1-0) and higher transition lines. We compare the CO luminosities and H$_{\rm 2}$ masses with recent observations of galaxies at low and high redshifts. Our model reproduces well the observed CO-luminosity function and the estimated H$_{\rm 2}$ mass in the local Universe. About ten per cent of molecules in the Universe reside in dwarf galaxies with stellar masses lower than $10^9~{\rm M_\odot}$, but the galaxies are generally `CO-dark' and have typically high $α_{\rm CO}$. Our model predicts generally lower CO line luminosities than observations at redshifts $z\gtrsim 1$--$2$. We argue that the difference can be explained by the highly turbulent structure suggested for the high-redshift star-forming galaxies.

astro-ph.GA↗

arXiv2020

Photometric classification of HSC transients using machine learning

The advancement of technology has resulted in a rapid increase in supernova (SN) discoveries. The Subaru/Hyper Suprime-Cam (HSC) transient survey, conducted from fall 2016 through spring 2017, yielded 1824 SN candidates. This gave rise to the need for fast type classification for spectroscopic follow-up and prompted us to develop a machine learning algorithm using a deep neural network (DNN) with highway layers. This machine is trained by actual observed cadence and filter combinations such that we can directly input the observed data array into the machine without any interpretation. We tested our model with a dataset from the LSST classification challenge (Deep Drilling Field). Our classifier scores an area under the curve (AUC) of 0.996 for binary classification (SN Ia or non-SN Ia) and 95.3% accuracy for three-class classification (SN Ia, SN Ibc, or SN II). Application of our binary classification to HSC transient data yields an AUC score of 0.925. With two weeks of HSC data since the first detection, this classifier achieves 78.1% accuracy for binary classification, and the accuracy increases to 84.2% with the full dataset. This paper discusses the potential use of machine learning for SN type classification purposes.

astro-ph.IM↗