Search

We present a comprehensive analysis of the shape of dark matter (DM) halos in a sample of 25 Milky Way-like galaxies in TNG50 simulation. Using an Enclosed Volume Iterative Method (EVIM), we infer an oblate-to-triaxial shape for the DM halo with the median $T \simeq 0.24 $. We group DM halos in 3 different categories. Simple halos (32% of population) establish principal axes whose ordering in magnitude does not change with radius and whose orientations are almost fixed throughout the halo. Twisted halos (32% of population), experience levels of gradual rotations throughout their radial profiles. Finally, stretched halos (36% of population) demonstrate a stretching in their principal axes lengths where the ordering of different eigenvalues change with radius. Subsequently, the halo experiences a "rotation" of $\sim$90 deg where the stretching occurs. Visualizing the 3D ellipsoid of each halo, for the first time, we report signs of re-orienting ellipsoid in twisted and stretched halos. We examine the impact of baryonic physics on DM halo shape through a comparison to dark matter only (DMO) simulations. This suggests a triaxial (prolate) halo. We analyze the impact of substructure on DM halo shape in both hydro and DMO simulations and confirm that their impacts are subdominant. We study the distribution of satellites in our sample. In simple and twisted halos, the angle of satellites' angular momentum with galaxy's angular momentum grows with radius. However, stretched halos show a flat distribution of angles. Overlaying our theoretical outcome on the observational results presented in the literature establishes a fair agreement., Comment: 32 pages, 28 plots, Accepted for publication in ApJ

We contrast the gas kinematics and dark matter contents of $z=2$ star-forming galaxies (SFGs) from state-of-the-art cosmological simulations within the $\Lambda$CDM framework to observations. To this end, we create realistic mock observations of massive SFGs ($M_*>4\times10^{10} M_{\odot}$, SFR $>50~M_{\odot}$ yr$^{-1}$) from the TNG50 simulation of the IllustrisTNG suite, resembling near-infrared, adaptive-optics assisted integral-field observations from the ground. Using observational line fitting and modeling techniques, we analyse in detail the kinematics of seven TNG50 galaxies from five different projections per galaxy, and compare them to observations of twelve massive SFGs by Genzel et al. (2020). The simulated galaxies show clear signs of disc rotation but mostly exhibit more asymmetric rotation curves, partly due to large intrinsic radial and vertical velocity components. At identical inclination angle, their one-dimensional velocity profiles can vary along different lines of sight by up to $\Delta v=200$ km s$^{-1}$. From dynamical modelling we infer rotation speeds and velocity dispersions that are broadly consistent with observational results. We find low central dark matter fractions compatible with observations ($f_{\rm DM}^v(

We use the IllustrisTNG simulations to show how the fractions of quenched galaxies vary across different environments and cosmic time, and to quantify the role AGN feedback and preprocessing play in quenching group and cluster satellites. At $z=0$, we select galaxies with $M_* = 10^{9-12} M_{\odot}$ residing within ($\leq R_{200c}$) groups and clusters of total host mass $M_{200c}=10^{13-15.2} M_{\odot}$. TNG predicts a quenched fraction of $\sim70-90\%$ (on average) for centrals and satellites $\gtrsim 10^{10.5} M_{\odot}$, regardless of host mass, cosmic time ($0\leq z\leq0.5$), clustercentric distance and time since infall in the $z=0$ host. Low-mass centrals ($\lesssim 10^{10} M_{\odot}$), instead, are rarely quenched unless they become members of groups ($10^{13-14} M_{\odot}$) or clusters ($\geq10^{14} M_{\odot}$), where the quenched fraction rises to $\sim80\%$. The fraction of low-mass passive galaxies is higher closer to the host center and for more massive hosts. The population of low-mass satellites accreted $\gtrsim$4-6 Gyr ago in massive hosts is almost entirely passive, thus suggesting an upper limit for the time needed for environmental quenching to occur. In fact, $\sim30\%$ of group and cluster satellites that are quenched at $z=0$ were already quenched before falling into their current host, and the bulk of them quenched as early as 4 to 10 billion years ago. For low-mass galaxies ($\lesssim10^{10-10.5}M_{\odot}$), this is due to preprocessing, whereby current satellites may have been members of other hosts, and hence have undergone environmental processes, before falling into their final host, this mechanism being more common and more effective for the purposes of quenching for satellites found today in more massive hosts. On the other hand, massive galaxies quench on their own and because of AGN feedback, regardless of whether they are centrals or satellites., Comment: 21 Pages, 11 figures, 2 tables, 1 appendix, published in MNRAS

Understanding the variability of galaxy star formation histories (SFHs) across a range of timescales provides insight into the underlying physical processes that regulate star formation within galaxies. We compile the SFHs of galaxies at $z=0$ from an extensive set of models, ranging from cosmological hydrodynamical simulations (Illustris, IllustrisTNG, Mufasa, Simba, EAGLE), zoom simulations (FIRE-2, g14, and Marvel/Justice League), semi-analytic models (Santa Cruz SAM) and empirical models (UniverseMachine), and quantify the variability of these SFHs on different timescales using the power spectral density (PSD) formalism. We find that the PSDs are well described by broken power-laws, and variability on long timescales ($\gtrsim1$ Gyr) accounts for most of the power in galaxy SFHs. Most hydrodynamical models show increased variability on shorter timescales ($\lesssim300$ Myr) with decreasing stellar mass. Quenching can induce $\sim0.4-1$ dex of additional power on timescales $>1$ Gyr. The dark matter accretion histories of galaxies have remarkably self-similar PSDs and are coherent with the in-situ star formation on timescales $>3$ Gyr. There is considerable diversity among the different models in their (i) power due to SFR variability at a given timescale, (ii) amount of correlation with adjacent timescales (PSD slope), (iii) evolution of median PSDs with stellar mass, and (iv) presence and locations of breaks in the PSDs. The PSD framework is a useful space to study the SFHs of galaxies since model predictions vary widely. Observational constraints in this space will help constrain the relative strengths of the physical processes responsible for this variability., Comment: 31 pages, 17 figures (+ appendix). Resubmitted to MNRAS after responding to referee's comments. Comments are welcome!

Physical and chemical properties of the interstellar medium (ISM) at sub-galactic ($\sim$kpc) scales play an indispensable role in controlling the ability of gas to form stars. As part of the SMAUG (Simulating Multiscale Astrophysics to Understand Galaxies) project, in this paper, we use the TNG50 cosmological simulation to explore the physical parameter space of 8 resolved ISM properties in star-forming regions to constrain the areas of this hyperspace over which most star-forming environments exist. We deconstruct our simulated galaxies spanning a wide range of mass (M$_\star = 10^{7-11}$ M$_\odot$) and redshift ($0 \leq z \leq 3$) into kpc-sized regions, and statistically analyze the gas/stellar surface densities, gas metallicity, vertical stellar velocity dispersion, epicyclic frequency and dark-matter volumetric density representative of each region in the context of their star formation activity and galactic environment (radial galactocentric location). By examining the star formation rate (SFR) weighted distributions of these properties, we show that stars primarily form in two spatially distinct environmental regimes, which are brought about by an underlying bi-component radial SFR surface density profile in galaxies. We examine how the relative prominence of these two regimes depends on host galaxy mass and cosmic time. We also compare our findings with those from integral field spectroscopy observations and achieve a good overall agreement. Further, using dimensionality reduction, we characterise the aforementioned hyperspace to reveal a high-degree of multicollinearity in relationships amongst ISM properties that drive the distribution of star formation at kpc-scales. Based on this, we show that a reduced 3D representation underpinned by a multi-variate radius relationship is sufficient to capture most of the variance in the original 8D space., Comment: Updated intro, minor text and abstract modifications. More information on the SMAUG project and the other first result papers can be found here: https://www.simonsfoundation.org/flatiron/center-for-computational-astrophysics/galaxy-formation/smaug/papersplash1

Semi-analytic models (SAMs) are a promising means of tracking the physical processes associated with galaxy formation, but many of their approximations have not been rigorously tested. As part of the SMAUG (Simulating Multiscale Astrophysics to Understand Galaxies) project, we compare predictions from the FIRE-2 hydrodynamical "zoom-in" simulations to those from the Santa Cruz SAM run on the same halo merger trees, with an emphasis on the global mass flow cycle. Our study includes 13 halos spanning low-mass dwarfs (M_vir~10^10 M_sun at z=0), intermediate-mass dwarfs (M_vir~10^11 M_sun) and Milky Way-mass galaxies (M_vir~10^12 M_sun). The SAM and FIRE-2 predictions agree relatively well with each other in terms of stellar and interstellar mass, but differ dramatically on circumgalactic mass (the SAM is lower than FIRE-2 by ~3 orders of magnitude for dwarfs). Strikingly, the SAM predicts higher gas accretion rates for dwarfs compared to FIRE-2 by factors of ~10-100, and this is compensated for with higher mass outflow rates in the SAM. We argue that the most severe model discrepancies are caused by the lack of preventative stellar feedback and the assumptions for halo gas cooling and recycling in the SAM. As a first step towards resolving these model tensions, we present a simple yet promising new preventative stellar feedback model in which the energy carried by supernova-driven winds is allowed to heat some fraction of gas outside of halos to at least the virial temperature such that accretion is suppressed., Comment: Accepted to ApJ with minor revisions. More information on the SMAUG project and the other first result papers can be found here: https://www.simonsfoundation.org/flatiron/center-for-computational-astrophysics/galaxy-formation/smaug/papersplash1

The past decade has seen significant progress in understanding galaxy formation and evolution using large-scale cosmological simulations. While these simulations produce galaxies in overall good agreement with observations, they employ different sub-grid models for galaxies and supermassive black holes (BHs). We investigate the impact of the sub-grid models on the BH mass properties of the Illustris, TNG100, TNG300, Horizon-AGN, EAGLE, and SIMBA simulations, focusing on the M_BH-M_star relation and the BH mass function. All simulations predict tight M_BH-M_star relations, and struggle to produce the lowest (M_BH< 10^7.5 Msun) in galaxies of M_star~10^10.5-10^11.5 Msun. While the time evolution of the mean M_BH-M_star relation is mild (<1 dex in BH mass for 010^9 Msun in most of the simulations. The BH mass function is dominated by efficiently accreting BHs (log10 f_Edd >-2$) at high redshifts, and transitions progressively from the high-mass to the low-mass end to be governed by inactive BHs. The transition time and the contribution of active BHs are different among the simulations, and can be used to evaluate models against observations., Comment: Published version - MNRAS

The standard cosmological model ($\Lambda$CDM) predicts the existence of the cosmic web: a distribution of matter into sheets and filaments connecting massive halos. However, observational evidence has been elusive due to the low surface brightness of the filaments. Recent deep MUSE/VLT data and upcoming observations offer a promising avenue for Ly$\alpha$ detection, motivating the development of modern theoretical predictions. We use hydrodynamical cosmological simulations run with the AREPO code to investigate the potential detectability of large-scale filaments, excluding contributions from the halos embedded in them. We focus on filaments connecting massive ($M_{200c}\sim(1-3)\times10^{12} M_\odot$) halos at z=3, and compare different simulation resolutions, feedback levels, and mock-image pixel sizes. We find increasing simulation resolution does not substantially improve detectability notwithstanding the intrinsic enhancement of internal filament structure. By contrast, for a MUSE integration of 31 hours, including feedback increases the detectable area by a factor of $\simeq$5.5 on average compared with simulations without feedback, implying that even the non-bound components of the filaments have substantial sensitivity to feedback. Degrading the image resolution from the native MUSE scale of (0.2")$^2$ per pixel to (5.3")$^2$ apertures has the strongest effect, increasing the detectable area by a median factor of $\simeq$200 and is most effective when the size of the pixel roughly matches the width of the filament. Finally, we find the majority of Ly$\alpha$ emission is due to electron impact collisional excitations, as opposed to radiative recombination.

We present an analysis of the angular momentum content of the circumgalactic medium (CGM) using TNG100, one of the flagship runs of the IllustrisTNG project. We focus on Milky Way-mass halos ($\sim 10^{12} \; M_{\odot}$) at $z=0$ but also analyze other masses and redshifts up to $z=5$. We find that the CGM angular momentum properties are strongly correlated with the stellar angular momentum of the corresponding galaxy: the CGM surrounding high-angular momentum galaxies has a systematically higher angular momentum and is better aligned to the rotational axis of the galaxy itself than the CGM surrounding low-angular momentum galaxies. Both the hot and cold phases of the CGM show this dichotomy, though it is stronger for colder gas. The CGM of high-angular momentum galaxies is characterized by a large wedge of cold gas with rotational velocities at least $\sim1/2$ of the halo's virial velocity, extending out to $\sim 1/2$ of the virial radius, and by biconical polar regions dominated by radial velocities suggestive of galactic fountains; both of these features are absent from the CGM of low-angular momentum galaxies. These conclusions are general to halo masses $\lesssim 10^{12} \; M_{\odot}$ and for $z \lesssim 2$, but they do not apply for more massive halos or at the highest redshift studied. By comparing simulations run with alterations to the fiducial feedback model, we identify the better alignment of the CGM to high-angular momentum galaxies as a feedback-independent effect and the galactic winds as a dominant influence on the CGM's angular momentum., Comment: Accepted to ApJ. 16 pages, 12 figures

We study the relationship between supermassive black hole (BH) feedback, BH luminosity and the kinematics of stars and gas for galaxies in IllustrisTNG. We use a sample of galaxies with mock MaNGA observations to identify kinematic misalignment at $z=0$ (difference in rotation of stars and gas), for which we follow the evolutionary history of BH activity and gas properties over the last 8 Gyrs. Misaligned low mass galaxies typically have boosted BH luminosity, BH growth and have had more energy injected into the gas over the last 8 Gyr in comparison to aligned galaxies. These properties likely lead to outflows and gas loss, in agreement with active low mass galaxies in observations. We show that splitting on BH luminosity at $z=0$ produces statistically consistent distributions of kinematic misalignment at $z=0$, however, splitting on the maximum BH luminosity over the last 8 Gyrs does not. While instantaneous correlation at $z=0$ is difficult due to misalignment persisting on longer timescales, the relationship between BH activity and misalignment is clear. High mass quenched galaxies with misalignment typically have similar BH luminosities, show no overall gas loss, and have typically lower gas phase metallicity over the last 8 Gyrs in comparison to those aligned; suggesting external origin., Comment: 5 pages, 3 Figures, Submitted to MNRAS letters. For first paper in series, see arXiv:1910.10744

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., Comment: 5 pages, 2 figures. Accepted to the Second Workshop on Machine Learning and the Physical Sciences (NeurIPS 2019)

We study black hole - host galaxy correlations, and the relation between the over-massiveness (the distance from the average $M_{BH}-\sigma$ relation) of super-massive black holes (SMBHs) and star formation histories of their host galaxies in the Illustris and TNG100 simulations. We find that both simulations are able to produce black hole scaling relations in general agreement with observations at $z=0$, but with noticeable discrepancies. Both simulations show an offset from the observations for the $M_{BH}-\sigma$ relation, and the relation between $M_{BH}$ and the Sersic index. The relation between $M_{BH}$ and stellar mass $M_*$ is tighter than the observations, especially for TNG100. For massive galaxies in both simulations, the hosts of over-massive SMBHs (those above the mean $M_{BH}-\sigma$ relation) tend to have larger Sersic indices and lower baryon conversion efficiency, suggesting a multidimensional link between SMBHs and properties of their hosts. In Illustris, the hosts of over-massive SMBHs have formed earlier and have lower present-day star formation rates, in qualitative agreement with the observations for massive galaxies with $\sigma>100 \rm km/s$. For low-mass galaxies, such a correlation still holds in Illustris but does not exist in the observed data. For TNG100, the correlation between SMBH over-massiveness and star formation history is much weaker. The hosts of over-massive SMBHs generally have consistently larger star formation rates throughout history. These galaxies have higher stellar mass as well, due to the strong $M_{BH}-M_*$ correlation. Our findings show that simulated SMBH scaling relations and correlations are sensitive to features in the modeling of SMBHs., Comment: 18 pages, 11 figures, submitted to ApJ

Supermassive black hole feedback is thought to be responsible for the lack of star formation, or quiescence, in a significant fraction of galaxies. We explore how observable correlations between the specific star formation rate (sSFR), stellar mass (M$_{\rm{star}}$), and black hole mass (M$_{\rm{BH}}$) are sensitive to the physics of black hole feedback in a galaxy formation model. We use the IllustrisTNG simulation suite, specifically the TNG100 simulation and ten model variations that alter the parameters of the black hole model. Focusing on central galaxies at $z = 0$ with M$_{\rm{star}} > 10^{10}$ M$_{\odot}$, we find that the sSFR of galaxies in IllustrisTNG decreases once the energy from black hole kinetic winds at low accretion rates becomes larger than the gravitational binding energy of gas within the galaxy stellar radius. This occurs at a particular M$_{\rm{BH}}$ threshold above which galaxies are found to sharply transition from being mostly star-forming to mostly quiescent. As a result of this behavior, the fraction of quiescent galaxies as a function of M$_{\rm{star}}$ is sensitive to both the normalization of the M$_{\rm{BH}}$-M$_{\rm{star}}$ relation and the M$_{\rm{BH}}$ threshold for quiescence in IllustrisTNG. Finally, we compare these model results to observations of 91 central galaxies with dynamical M$_{\rm{BH}}$ measurements with the caveat that this sample is not representative of the whole galaxy population. While IllustrisTNG reproduces the observed trend that quiescent galaxies host more massive black holes, the observations exhibit a broader scatter in M$_{\rm{BH}}$ at a given M$_{\rm{star}}$ and show a smoother decline in sSFR with M$_{\rm{BH}}$., Comment: 20 pages, submitted to MNRAS

Motivated by the recently discovered kinematic "Hubble sequence" shown by the stellar orbit-circularity distribution of 260 CALIFA galaxies, we make use of a comparable galaxy sample at z = 0 with a stellar mass range from 5E9 to 5E11 solar masses, selected from the IllustrisTNG simulation and study their stellar orbit compositions in relation to a number of other fundamental galaxy properties.We find that the TNG100 simulation broadly reproduces the observed fractions of different orbital components and their stellar mass dependencies. In particular, the mean mass dependencies of the luminosity fractions for the kinematically warm and hot orbits are well reproduced within model uncertainties of the observed galaxies. The simulation also largely reproduces the observed peak and trough features at a stellar mass of 1-2E10 solar masses, in the mean distributions of the cold- and hot-orbit fractions, respectively, indicating fewer cooler orbits and more hotter orbits in both more- and less-massive galaxies beyond such a mass range. Several marginal disagreements are seen between the simulation and observations: the average cold-orbit (counter-rotating) fractions of the simulated galaxies below (above) a stellar mass of 6E10 solar masses, are systematically higher than the observational data by < 10% (absolute orbital fraction); the simulation also seems to produce more scatter for the cold-orbit fraction and less so for the non-cold orbits at any given galaxy mass. Possible causes that stem from the adopted heating mechanisms are discussed., Comment: 12 pages, 6 figures, 3 tables, submitted to MNRAS. Comments are welcome!

Using the IllustrisTNG simulations, we investigate the connection between galaxy morphology and star formation in central galaxies with stellar masses in the range $10^9-10^{11.5}~\mathrm{M}_{\odot}$. We quantify galaxy morphology by a kinematical decomposition of the stellar component into a spheroidal and a disc component (spheroid-to-total ratio, S/T) and by the concentration of the stellar mass density profile ($C_{82}$). S/T is correlated with stellar mass and star-formation activity, while $C_{82}$ correlates only with stellar mass. Overall, we find good agreement with observational estimates for both S/T and $C_{82}$. Low and high mass galaxies are dominated by random stellar motion, while only intermediate-mass galaxies ($M_{\star}\approx10^{10}-10^{10.5}~\mathrm{M}_{\odot}$) are dominated by ordered rotation. Whereas higher-mass galaxies are typical spheroids with high concentrations, lower-mass galaxies have low concentration, pointing to different formation channels. Although we find a correlation between S/T and star-formation activity, in the TNG model galaxies do not necessarily change their morphology when they transition through the green valley or when they cease their star formation, this depending on galaxy stellar mass and morphological estimator. Instead, the morphology (S/T and $C_{82}$) is generally set during the star-forming phase of galaxies. The apparent correlation between S/T and star formation arises because earlier-forming galaxies had, on average, a higher S/T at a given stellar mass. Furthermore, we show that mergers drive in-situ bulge formation in intermediate-mass galaxies and are responsible for the recent spheroidal mass assembly in the massive galaxies with $M_{\star}>10^{11}~\mathrm{M}_{\odot}$. In particular, these massive galaxies assemble about half of the spheroidal mass while star-forming and the other half through mergers while quiescent., Comment: 20 pages with 14 figures (+ appendix). Accepted by MNRAS

Stars in galaxies form from the cold rotationally supported gaseous disks that settle at the center of dark matter halos. In the simplest models, such angular momentum is acquired early on at the time of collapse of the halo and preserved thereafter, implying a well-aligned spin for the stellar and gaseous component. Observations however have shown the presence of gaseous disks in counterrotation with the stars. We use the Illustris numerical simulations to study the origin of such counterrotation in low mass galaxies ($M_\star = 2 \times 10^9$ - $5 \times 10^{10}\; \rm M_\odot$), a sample where mergers have not played a significant role. Only ${\sim}1\%$ of our sample shows a counterrotating gaseous disk at $z=0$. These counterrotating disks arise in galaxies that have had a significant episode of gas removal followed by the acquisition of new gas with misaligned angular momentum. In our simulations, we identify two main channels responsible for the gas loss: a strong feedback burst and gas stripping during a fly-by passage through a more massive group environment. Once settled, counterrotation can be long-lived with several galaxies in our sample displaying misaligned components consistently for more than $2$ Gyr. As a result, no major correlation with the present day environment or structural properties might remain, except for a slight preference for early type morphologies and a lower than average gas content at a given stellar mass., Comment: 16 pages, 11 figures. Submitted to ApJ. Comments welcome

In recent years, machine learning (ML) methods have remarkably improved how cosmologists can interpret data. The next decade will bring new opportunities for data-driven cosmological discovery, but will also present new challenges for adopting ML methodologies and understanding the results. ML could transform our field, but this transformation will require the astronomy community to both foster and promote interdisciplinary research endeavors., Comment: Submitted to the Astro2020 call for science white papers

We present a new cosmological, magnetohydrodynamical simulation for galaxy formation: TNG50, the third and final installment of the IllustrisTNG project. TNG50 evolves 2x2160^3 dark-matter particles and gas cells in a volume 50 comoving Mpc across. It hence reaches a numerical resolution typical of zoom-in simulations, with a baryonic element mass of 8.5x10^4 Msun and an average cell size of 70-140 parsecs in the star-forming regions of galaxies. Simultaneously, TNG50 samples ~700 (6,500) galaxies with stellar masses above 10^10 (10^8) Msun at z=1. Here we investigate the structural and kinematical evolution of star-forming galaxies across cosmic time (0 < z < 6). We quantify their sizes, disk heights, 3D shapes, and degree of rotational vs. dispersion-supported motions as traced by rest-frame V-band light (i.e. roughly stellar mass) and by Halpha light (i.e. star-forming and dense gas). The unprecedented resolution of TNG50 enables us to model galaxies with sub-kpc half-light radii and with <300-pc disk heights. Coupled with the large-volume statistics, we characterize a diverse, redshift- and mass-dependent structural and kinematical morphological mix of galaxies all the way to early epochs. Our model predicts that for star-forming galaxies the fraction of disk-like morphologies, based on 3D stellar shapes, increases with both cosmic time and galaxy stellar mass. Gas kinematics reveal that the vast majority of 10^9-11.5 Msun star-forming galaxies are rotationally-supported disks for most cosmic epochs (Vmax/sigma>2-3, z<5), being dynamically hotter at earlier epochs (z>1.5). Despite large velocity dispersion at high redshift, cold and dense gas in galaxies predominantly arranges in disky or elongated shapes at all times and masses; these gaseous components exhibit rotationally-dominated motions far exceeding the collisionless stellar bodies., Comment: MNRAS. Highlights: Figures 9, 14, 15. See companion paper by Nelson et al. 2019b. Visuals at http://www.tng-project.org

We present the new TNG50 cosmological, magnetohydrodynamical simulation -- the third and final volume of the IllustrisTNG project. This simulation occupies a unique combination of large volume and high resolution, with a 50 Mpc box sampled by 2160^3 gas cells (baryon mass of 8x10^4 Msun). The median spatial resolution of star-forming ISM gas is ~100-140 parsecs. This resolution approaches or exceeds that of modern 'zoom' simulations of individual massive galaxies, while the volume contains ~20,000 resolved galaxies with M*>10^7 Msun. Herein we show first results from TNG50, focusing on galactic outflows driven by supernovae as well as supermassive black hole feedback. We find that the outflow mass loading is a non-monotonic function of galaxy stellar mass, turning over and rising rapidly above 10^10.5 Msun due to the action of the central black hole. Outflow velocity increases with stellar mass, and at fixed mass is faster at higher redshift. The TNG model can produce high velocity, multi-phase outflows which include cool, dense components. These outflows reach speeds in excess of 3000 km/s out to 20 kpc with an ejective, BH-driven origin. Critically, we show how the relative simplicity of model inputs (and scalings) at the injection scale produces complex behavior at galactic and halo scales. For example, despite isotropic wind launching, outflows exhibit natural collimation and an emergent bipolarity. Furthermore, galaxies above the star-forming main sequence drive faster outflows, although this correlation inverts at high mass with the onset of quenching, whereby low luminosity, slowly accreting, massive black holes drive the strongest outflows., Comment: MNRAS, see also companion paper by Pillepich et al. (2019b). Visualizations, movies, and an image gallery of paper figures available on the TNG50 website: www.tng-project.org

We select galaxies from the IllustrisTNG hydrodynamical simulations ($M_*>10^9~\rm M_\odot$ at $0\le z\le2$) and characterize the shapes and evolutions of their UVJ and star-formation rate -- stellar mass (SFR-$M_*$) diagrams. We quantify the systematic uncertainties related to different criteria to classify star-forming vs. quiescent galaxies, different SFR estimates, and by accounting for the star formation measured within different physical apertures. The TNG model returns the observed features of the UVJ diagram at $z\leq2$, with a clear separation between two classes of galaxies. It also returns a tight star-forming main sequence (MS) for $M_*<10^{10.5}\,\rm M_\odot$ with a $\sim0.3$ dex scatter at $z\sim0$ in our fiducial choices. If a UVJ-based cut is adopted, the TNG MS exhibits a downwardly bending at stellar masses of about $10^{10.5-10.7}~\rm M_\odot$. Moreover, the model predicts that $\sim80\,(50)$ per cent of $10^{10.5-11}~\rm M_\odot$ galaxies at $z=0~(z=2)$ are quiescent and the numbers of quenched galaxies at intermediate redshifts and high masses are in better agreement with observational estimates than previous models. However, shorter SFR-averaging timescales imply higher normalizations and scatter of the MS, while smaller apertures lead to underestimating the galaxy SFRs: overall we estimate the inspected systematic uncertainties to sum up to about $0.2-0.3$ dex in the locus of the MS and to about 15 percentage points in the quenched fractions. While TNG color distributions are clearly bimodal, this is not the case for the SFR logarithmic distributions in bins of stellar mass (SFR$\geq 10^{-3}~\rm M_\odot$yr$^{-1}$). Finally, the slope and $z=0$ normalization of the TNG MS are consistent with observational findings; however, the locus of the TNG MS remains lower by about $0.2-0.5$ dex at $0.75\le z<2$ than the available observational estimates taken at face value., Comment: 24 pages, 4 tables, 11 figures. Accepted for publication on MNRAS

We present the full public release of all data from the TNG50, TNG100 and TNG300 simulations of the IllustrisTNG project. IllustrisTNG is a suite of large volume, cosmological, gravo-magnetohydrodynamical simulations run with the moving-mesh code Arepo. TNG includes a comprehensive model for galaxy formation physics, and each TNG simulation self-consistently solves for the coupled evolution of dark matter, cosmic gas, luminous stars, and supermassive blackholes from early time to the present day, z=0. Each of the flagship runs -- TNG50, TNG100, and TNG300 -- are accompanied by lower-resolution and dark-matter only counterparts, and we discuss scientific and numerical cautions and caveats relevant when using TNG. Full volume snapshots are available at 100 redshifts; halo and subhalo catalogs at each snapshot and merger trees are also released. The data volume now directly accessible online is ~1.1 PB, including 2,000 full volume snapshots and ~110,000 high time-resolution subbox snapshots. Data access and analysis examples are available in IDL, Python, and Matlab. We describe improvements and new functionality in the web-based API, including on-demand visualization and analysis of galaxies and halos, exploratory plotting of scaling relations and other relationships between galactic and halo properties, and a new JupyterLab interface. This provides an online, browser-based, near-native data analysis platform which supports user computation with fully local access to TNG data, alleviating the need to download large simulated datasets., Comment: TNG50 joins TNG100 and TNG300 in public release (1 Feb 2021) at http://www.tng-project.org/data

In this study we explore the complex multi-phase gas of the circumgalactic medium (CGM) surrounding galaxies. We propose and implement a novel, super-Lagrangian 'CGM zoom' scheme in the moving-mesh code AREPO, which focuses more resolution into the CGM and intentionally lowers resolution in the dense ISM. We run two cosmological simulations of the same galaxy halo, once with a simple 'no feedback' model, and separately with a more comprehensive physical model including galactic-scale outflows as in the Illustris simulation. Our chosen halo has a total mass of ~10^12 Msun at z ~ 2, and we achieve a median gas mass (spatial) resolution of ~2,200 solar masses (~95 parsecs) in the CGM, six-hundred (fourteen) times better than in the Illustris-1 simulation, a higher spatial resolution than any cosmological simulation at this mass scale to date. We explore the primary channel(s) of cold-phase CGM gas production in this regime. We find that winds substantially enhance the amount of cold gas in the halo, also evidenced in the covering fractions of HI and the equivalent widths of MgII out to large radii, in better agreement with observations than the case without galactic winds. Using a tracer particle analysis to follow the thermodynamic history of gas, we demonstrate how the majority of this cold, dense gas arises due to rapid cooling of the wind material interacting with the hot halo, and how large amounts of cold, ~10^4 K gas can be produced and persist in galactic halos with Tvir ~ 10^6 K. At the resolutions presently considered, the quantitative properties of the CGM we explore are not appreciably affected by the refinement scheme., Comment: MNRAS submitted, comments welcome. High-res version at http://www.mpa-garching.mpg.de/~dnelson/papers/Suresh19_zooming2.pdf

We test a method to reduce unwanted sample variance when predicting Lyman-$\alpha$ (ly$\alpha$) forest power spectra from cosmological hydrodynamical simulations. Sample variance arises due to sparse sampling of modes on large scales and propagates to small scales through non-linear gravitational evolution. To tackle this, we generate initial conditions in which the density perturbation amplitudes are {\it fixed} to the ensemble average power spectrum -- and are generated in {\it pairs} with exactly opposite phases. We run $50$ such simulations ($25$ pairs) and compare their performance against $50$ standard simulations by measuring the ly$\alpha$ 1D and 3D power spectra at redshifts $z=2$, 3, and 4. Both ensembles use periodic boxes of $40$ Mpc/h containing $512^3$ particles each of dark matter and gas. As a typical example of improvement, for wavenumbers $k=0.25$ h/Mpc at $z=3$, we find estimates of the 1D and 3D power spectra converge $34$ and $12$ times faster in a paired-fixed ensemble compared with a standard ensemble. We conclude that, by reducing the computational time required to achieve fixed accuracy on predicted power spectra, the method frees up resources for exploration of varying thermal and cosmological parameters -- ultimately allowing the improved precision and accuracy of statistical inference., Comment: 15 pages, 3 figures

We use IllustrisTNG, a suite of gravity and MHD simulations, to study the demographics and properties of jellyfish galaxies in the full cosmological context. By jellyfish galaxies, we mean satellites orbiting in massive groups and clusters that exhibit highly asymmetric distributions of gas and gas tails. We use the TNG100 run and select galaxies at redshifts $z\le0.6$ with stellar mass exceeding $10^{9.5}{\rm M_\odot}$ and with host halo masses of $10^{13}-10^{14.6}\,{\rm M_\odot}$. Among more than about 6000 (2600) galaxies with stars (and some gas), we identify 800 jellyfish galaxies by visually inspecting their gas and stellar mass maps in random projections. About $31\%$ of cluster satellites are found with signatures of ram-pressure stripping and gaseous tails stemming from the main luminous bodies. This is a lower limit, since the random orientation entails a loss of about $30\%$ of galaxies that in an optimal projection would otherwise be identified as jellyfish. The connection with ram-pressure stripping is further confirmed by a series of findings: jellyfish galaxies are more frequent at intermediate and large cluster-centric distances ($r/R_{\rm 200c}\gtrsim 0.25$); they move through the ICM with larger bulk velocities and Mach numbers than the general cluster population, typically orbiting supersonically and experiencing larger ram pressures. Furthermore, the gaseous tails usually extend in opposite directions to the galaxy trajectory, with no relation between tail orientation and the host's center. The frequency of jellyfish galaxies shows a very weak dependence on redshift $(0\le z\le0.6)$ but larger fractions of disturbed gaseous morphologies occur in more massive hosts and at smaller satellite masses. Finally, jellyfish galaxies are late infallers ($< 2.5-3$ Gyrs ago, at $z=0$) and the emergence of gaseous tails correlates well with the presence of bow shocks in the ICM., Comment: 25 pages, 15 figures, Accepted for publication on MNRAS after minor revisions

We have generated synthetic images of $\sim$27,000 galaxies from the IllustrisTNG and the original Illustris hydrodynamic cosmological simulations, designed to match Pan-STARRS observations of $\log_{10}(M_{\ast}/{\rm M}_{\odot}) \approx 9.8$-$11.3$ galaxies at $z \approx 0.05$. Most of our synthetic images were created with the SKIRT radiative transfer code, including the effects of dust attenuation and scattering, and performing the radiative transfer directly on the Voronoi mesh used by the simulations themselves. We have analysed both our synthetic and real Pan-STARRS images with the newly developed $\tt{statmorph}$ code, which calculates non-parametric morphological diagnostics -- including the Gini-$M_{20}$ and concentration-asymmetry-smoothness (CAS) statistics -- and performs two-dimensional S\'ersic fits. Overall, we find that the optical morphologies of IllustrisTNG galaxies are in good agreement with observations, and represent a substantial improvement compared to the original Illustris simulation. In particular, the locus of the Gini-$M_{20}$ diagram is consistent with that inferred from observations, while the median trends with stellar mass of all the morphological, size and shape parameters considered in this work lie within the $\sim$1$\sigma$ scatter of the observational trends. However, the IllustrisTNG model has some difficulty with more stringent tests, such as producing a strong morphology-colour relation. This results in a somewhat higher fraction of red discs and blue spheroids compared to observations. Similarly, the morphology-size relation is problematic: while observations show that discs tend to be larger than spheroids at a fixed stellar mass, such a trend is not present in IllustrisTNG., Comment: 21 pages, 11 figures. Published in MNRAS. Code available at https://statmorph.readthedocs.io

We study the connection between active galactic nuclei (AGN) and their host galaxies through cosmic time in the large-scale cosmological IllustrisTNG simulations. We first compare BH properties, i.e. the hard X-ray BH luminosity function, AGN galaxy occupation fraction, and distribution of Eddington ratios, to available observational constraints. The simulations produce a population of BHs in good agreement with observations, but we note an excess of faint AGN in hard X-ray (L_x ~ 10^{43-44} erg/s), and a lower number of bright AGN (L_x>10^{44} erg/s), a conclusion that varies quantitatively but not qualitatively with BH luminosity estimation method. The lower Eddington ratios of the 10^{9} Msun BHs compared to observations suggest that AGN feedback may be too efficient in this regime. We study galaxy star formation activity and structural properties, and design sample-dependent criteria to identify different galaxy types (star-forming/quiescent, extended/compact) that we apply both to the simulations and observations from the candels fields. We analyze how the simulated and observed galaxies populate the specific star formation rate - stellar mass surface density diagram. A large fraction of the z=0 M_{star}>10^{11} Msun quiescent galaxies first experienced a compaction phase (i.e. reduction of galaxy size) while still forming stars, and then a quenching event. We measure the dependence of AGN fraction on galaxies' locations in this diagram. After correcting the simulations with a redshift and AGN luminosity-dependent model for AGN obscuration, we find good qualitative and quantitative agreement with observations. The AGN fraction is the highest among compact star-forming galaxies (16-20% at z~1.5-2), and the lowest among compact quiescent galaxies (6-10% at z~1.5-2)., Comment: 35 pages, 22 figures, accepted for publication in MNRAS

A tightly correlated star formation rate-stellar mass relation of star forming galaxies, or star-forming sequence (SFS), is a key feature in galaxy property-space that is predicted by modern galaxy formation models. We present a flexible data-driven approach for identifying this SFS over a wide range of star formation rates and stellar masses using Gaussian mixture modeling (GMM). Using this method, we present a consistent comparison of the $z=0$ SFSs of central galaxies in the Illustris, EAGLE, and Mufasa hydrodynamic simulations and the Santa Cruz semi-analytic model (SC-SAM), alongside data from the Sloan Digital Sky Survey. We find, surprisingly, that the amplitude of the SFS varies by up to ${\sim} 0.7\,\mathrm{dex}$ (factor of ${\sim} 5$) among the simulations with power-law slopes range from $0.7$ to $1.2$. In addition to the SFS, our GMM method also identifies sub-components in the star formation rate-stellar mass relation corresponding to star-burst, transitioning, and quiescent sub-populations. The hydrodynamic simulations are similarly dominated by SFS and quiescent sub-populations unlike the SC-SAM, which predicts substantial fractions of transitioning and star-burst galaxies at stellar masses above and below $10^{10} M_\odot$, respectively. All of the simulations also produce an abundance of low-mass quiescent central galaxies in apparent tension with observations. These results illustrate that, even among models that well reproduce many observables of the galaxy population, the $z=0$ SFS and other sub-populations still show marked differences that can provide strong constraints on galaxy formation models., Comment: 28 pages, 15 figures

Intensity mapping surveys will provide access to a coarse view of the cosmic large-scale structure in unprecedented large volumes at high redshifts. Given the large fractions of the sky that can be efficiently scanned using emission from cosmic neutral hydrogen (HI), intensity mapping is ideally suited to probe a wide range of density environments and hence to constrain cosmology and fundamental physics. To efficiently extract information from 21cm intensities beyond average, one needs non-Gaussian statistics that capture large deviations from mean HI density. Counts-in-cells statistics are ideally suited for this purpose, as the statistics of matter densities in spheres can be predicted accurately on scales where their variance is below unity. We use a large state-of-the-art magneto-hydrodynamic simulation from the IllustrisTNG project to determine the relation between neutral hydrogen and matter densities in cells. We demonstrate how our theoretical knowledge about the matter PDF for a given cosmology can be used to extract a parametrisation-independent HI bias function from a measured neutral hydrogen PDF. When combining the predicted matter PDFs with a simple bias fit to the simulation, we obtain a prediction for neutral hydrogen PDFs at a few percent accuracy at scale R=5 Mpc/h from redshift z=5 to z=1. Furthermore, we find a density-dependent HI clustering signal that is consistent with theoretical expectations and could allow for joint constraints of HI bias and the amplitude of matter fluctuations or the growth of structure.

We study the chaotic-like behavior of cosmological simulations by quantifying how minute perturbations grow over time and manifest as macroscopic differences in galaxy properties. When we run pairs of 'shadow' simulations that are identical except for random minute initial displacements to particle positions (e.g. of order 1e-7pc), the results diverge from each other at the individual galaxy level (while the statistical properties of the ensemble of galaxies are unchanged). After cosmological times, the global properties of pairs of 'shadow' galaxies that are matched between the simulations differ from each other generally at a level of ~2-25%, depending on the considered physical quantity. We perform these experiments using cosmological volumes of (25-50Mpc/h)^3 evolved either purely with dark matter, or with baryons and star-formation but no feedback, or using the full feedback model of the IllustrisTNG project. The runs cover four resolution levels spanning a factor of 512 in mass. We find that without feedback the differences between shadow galaxies generally become smaller as the resolution increases, but with the IllustrisTNG model the results are mostly converging towards a 'floor'. This hints at the role of feedback in setting the chaotic properties of galaxy formation. Importantly, we compare the macroscopic differences between shadow galaxies to the overall scatter in various galaxy scaling relations, and conclude that for the star formation-mass and the Tully-Fisher relations the butterfly effect in our simulations contributes significantly to the overall scatter. We find that our results are robust to whether random numbers are used in the sub-grid models or not. We discuss the implications for galaxy formation theory in general and for cosmological simulations in particular., Comment: Key figures: 10 & 11. Accepted for publication in ApJ. This final version includes a new verification of the conclusions in simulations that completely avoid the usage of random numbers

The initial conditions of cosmological simulations are commonly drawn from a Gaussian ensemble. The limited number of modes inside a simulation volume gives rise to statistical fluctuations known as \textit{sample variance}, limiting the accuracy of simulation predictions. Fixed fields offer an alternative initialization strategy; they have the same power spectrum as standard Gaussian fields but without intrinsic amplitude scatter at linear order. Paired fixed fields consists of two fixed fields with opposite phases that cancel phase correlations which otherwise induce second-order scatter in the non-linear power spectrum. We study the statistical properties of those fields for 19 different quantities at different redshifts through a large set of 600 N-body and 506 state-of-the-art magneto-hydrodynamic simulations covering a wide range of scales, mass and spatial resolutions. We find that paired fixed simulations do not introduce a bias on any of the examined quantities. We quantify the statistical improvement brought by these simulations, over standard ones, on different power spectra such as matter, halos, CDM, gas, stars, black-holes and magnetic fields, finding that they can reduce their variance by factors as large as $10^6$. We quantify the improvement achieved by fixing and by pairing, showing that sample variance in some quantities can be highly suppressed by pairing after fixing. Paired fixed simulations do not change the scatter in quantities such as the probability distribution function of matter density, or the halo, void or stellar mass functions. We argue that procedures aiming at reducing the sample variance of those quantities are unlikely to work. Our results show that paired fixed simulations do not affect either mean relations or scatter of galaxy properties, and suggest that the information embedded in 1-pt statistics is highly complementary to that in clustering., Comment: 24 pages, 12 figures

The galaxy Malin 1 contains the largest stellar disk known but the formation mechanism of this structure has been elusive. In this paper, we report a Malin 1 analogue in the 100 Mpc IllustrisTNG simulation and describe its formation history. At redshift zero, this massive galaxy, having a maximum circular velocity $V_{\rm max}$ of 430 ${\rm km\ s^{-1}}$, contains a 100 kpc gas/stellar disk with morphology similar to Malin 1. The simulated galaxy reproduces well many observed features of Malin 1's vast disk, including its stellar ages, metallicities, and gas rotation curve. We trace the extended disk back in time and find that a large fraction of the cold gas at redshift zero originated from the cooling of hot halo gas, triggered by the merger of a pair of intruding galaxies. Our finding provides a novel way to form large galaxy disks as extreme as Malin 1 within the current galaxy formation framework., Comment: 5 page, 5 figures; resubmitted to MNRAS Letters after first referee report; Comments welcome. Animations of gas/stars can be found here at https://youtu.be/Ibl4Cyybw2Q and https://youtu.be/3aIpCu6e9dQ, or downloaded at http://www.tng-project.org/movies/tng/TNG_Malin_1_gas.mpeg and http://www.tng-project.org/movies/tng/TNG_Malin_1_star.mpeg

[Abridged] We study the abundance and clustering properties of HI at redshifts $z\leqslant5$ using TNG100, a large state-of-the-art magneto-hydrodynamic simulation of a 75 Mpc/h box size. We show that most of the HI lies within dark matter halos and quantify the average HI mass hosted by halos of mass M at redshift z. We find that only halos with circular velocities larger than $\simeq$ 30 km/s contain HI. While the density profiles of HI exhibit a large halo-to-halo scatter, the mean profiles are universal across mass and redshift. The HI in low-mass halos is mostly located in the central galaxy, while in massive halos is concentrated in the satellites. We show that the HI and matter density probability distribution functions differ significantly. Our results point out that for small halos the HI bulk velocity goes in the same direction and has the same magnitude as the halo peculiar velocity, while in large halos differences show up. We find that halo HI velocity dispersion follows a power-law with halo mass. We find a complicated HI bias, with HI becoming non-linear already at $k=0.3$ h/Mpc at $z\gtrsim3$. Our simulation reproduces the DLAs bias value from observations. We find that the clustering of HI can be accurately reproduced by perturbative methods. We identify a new secondary bias, by showing that the clustering of halos depends not only on mass but also on HI content. We compute the amplitude of the HI shot-noise and find that it is small at all redshifts. We study the clustering of HI in redshift-space, and show that linear theory can explain the ratio between the monopoles in redshift- and real-space down to small scales at high redshift. We find that the amplitude of the Fingers-of-God effect is larger for HI than for matter. We point out that accurate 21 cm maps can be created from N-body or approximate simulations rather than full hydrodynamic simulations., Comment: 41 pages. 30 figures. Analysis tools publicly available at https://github.com/franciscovillaescusa/Pylians/tree/master/HI_Illustris

The similarity between the distributions of spins for galaxies ($\lambda_{\rm g}$) and for dark-matter haloes ($\lambda_{\rm h}$), indicated both by simulations and observations, is naively interpreted as a one-to-one correlation between the spins of a galaxy and its host halo. This is used to predict galaxy sizes in semi-analytic models via $R_{\rm e}\simeq\lambda_{\rm h} R_{\rm v}$, with $R_{\rm e}$ the half-mass radius of the galaxy and $R_{\rm v}$ the halo radius. Utilizing two different suites of zoom-in cosmological simulations, we find that $\lambda_{\rm g}$ and $\lambda_{\rm h}$ are in fact only barely correlated, especially at $z\geq 1$. A general smearing of this correlation is expected based on the different spin histories, where the more recently accreted baryons through streams gain and then lose significant angular momentum compared to the gradually accumulated dark matter. Expecting the spins of baryons and dark matter to be correlated at accretion into $R_{\rm v}$, the null correlation at the end reflects an anti-correlation between $\lambda_{\rm g}/\lambda_{\rm h}$ and $\lambda_{\rm h}$, which can partly arise from mergers and a compact star-forming phase that many galaxies undergo. On the other hand, the halo and galaxy spin vectors tend to be aligned, with a median $\cos\theta=0.6$-0.7 between galaxy and halo, consistent with instreaming within a preferred plane. The galaxy spin is better correlated with the spin of the inner halo, but this largely reflects the effect of the baryons on the halo. Following the null spin correlation, $\lambda_{\rm h}$ is not a useful proxy for $R_{\rm e}$. While our simulations reproduce a general relation of the sort $R_{\rm e}=AR_{\rm vir}$, in agreement with observational estimates, the relation becomes tighter with $A=0.02(c/10)^{-0.7}$, where $c$ is the halo concentration, which in turn introduces a dependence on mass and redshift., Comment: 20 pages, 16 figures, submitted to MNRAS

We use the IllustrisTNG (TNG) cosmological simulations to provide theoretical expectations for the dark matter mass fractions (DMFs) and circular velocity profiles of galaxies. TNG predicts flat circular velocity curves for $z = 0$ Milky Way (MW)-like galaxies beyond a few kpc from the galaxy centre, in better agreement with observational constraints than its predecessor, Illustris. TNG also predicts an enhancement of the dark matter mass within the 3D stellar half-mass radius ($r_\mathrm{half}$; $M_\mathrm{200c} = 10^{10}-10^{13}\mathrm{M}_{\odot}$, $z \le2$) compared to its dark matter only and Illustris counterparts. This enhancement leads TNG present-day galaxies to be dominated by dark matter within their inner regions, with $f_\mathrm{DM}(

We use the IllustrisTNG simulations to investigate the evolution of the mass-metallicity relation (MZR) for star-forming cluster galaxies as a function of the formation history of their cluster host. The simulations predict an enhancement in the gas-phase metallicities of star-forming cluster galaxies (10^9< M_star<10^10 M_sun) at z<1.0 in comparisons to field galaxies. This is qualitatively consistent with observations. We find that the metallicity enhancement of cluster galaxies appears prior to their infall into the central cluster potential, indicating for the first time a systematic "chemical pre-processing" signature for {\it infalling} cluster galaxies. Namely, galaxies which will fall into a cluster by z=0 show a ~0.05 dex enhancement in the MZR compared to field galaxies at z<0.5. Based on the inflow rate of gas into cluster galaxies and its metallicity, we identify that the accretion of pre-enriched gas is the key driver of the chemical evolution of such galaxies, particularly in the stellar mass range (10^9< M_star<10^10 M_sun). We see signatures of an environmental dependence of the ambient/inflowing gas metallicity which extends well outside the nominal virial radius of clusters. Our results motivate future observations looking for pre-enrichment signatures in dense environments., Comment: 5 pages, 4 figures, accepted for publication in MNRAS Letters

We explore the abundance, spatial distribution, and physical properties of the OVI, OVII, and OVIII ions of oxygen in circumgalactic and intergalactic media (the CGM, IGM, and WHIM). We use the TNG100 and TNG300 large volume cosmological magneto-hydrodynamical simulations. Modeling the ionization states of simulated oxygen, we find good agreement with observations of the low-redshift OVI column density distribution function (CDDF), and present its evolution for all three ions from z=0 to z=4. Producing mock quasar absorption line spectral surveys, we show that the IllustrisTNG simulations are fully consistent with constraints on the OVI content of the CGM from COS-Halos and other low redshift observations, producing columns as high as observed. We measure the total amount of mass and average column densities of each ion using hundreds of thousands of simulated galaxies spanning 10^11 < Mhalo/Msun < 10^15 corresponding to 10^9 < M*/Msun < 10^12 in stellar mass. Stacked radial profiles of OVI are computed in 3D number density and 2D projected column, decomposing into the 1-halo and 2-halo terms. Relating halo OVI to properties of the central galaxy, we find a correlation between the (g-r) color of a galaxy and the total amount of OVI in its CGM. In comparison to the COS-Halos finding, this leads to a dichotomy of columns around star-forming versus passive galaxies at fixed stellar (or halo) mass. We demonstrate that this correlation is a direct result of blackhole feedback associated with quenching and represents a causal consequence of galactic-scale baryonic feedback impacting the physical state of the circumgalactic medium., Comment: MNRAS. For an interactive OVI/OVIII visualization, see the 2D explorer at http://www.tng-project.org (fig 5 updated in v2)

The fundamental metallicity relation (FMR) is a postulated correlation between galaxy stellar mass, star formation rate (SFR), and gas-phase metallicity. At its core, this relation posits that offsets from the mass-metallicity relation (MZR) at a fixed stellar mass are correlated with galactic SFR. In this Letter, we quantify the timescale with which galactic SFRs and metallicities evolve using hydrodynamical simulations. We find that Illustris and IllustrisTNG predict that galaxy offsets from the star formation main sequence and MZR evolve over similar timescales, are often anti-correlated in their evolution, evolve with the halo dynamical time, and produce a pronounced FMR. In fact, for a FMR to exist, the metallicity and SFR must evolve in an anti-correlated sense which requires that they evolve with similar time variability. In contrast to Illustris and IllustrisTNG, we speculate that the SFR and metallicity evolution tracks may become decoupled in galaxy formation models dominated by globally-bursty SFR histories, which could weaken the FMR residual correlation strength. This opens the possibility of discriminating between bursty and non-bursty feedback models based on the strength and persistence of the FMR -- especially at high redshift., Comment: 5 pages; submitted to MNRAS Letters; comments welcome. The IllustrisTNG project website can be found at http://www.tng-project.org/

The coevolution of galaxies and their metal content serves as an important test for galaxy feedback models. We analyze the distribution and evolution of metals within the IllustrisTNG simulation suite with a focus on the gas-phase mass-metallicity relation (MZR). We find that the IllustrisTNG model broadly reproduces the slope and normalization evolution of the MZR across the redshift range $02$) evolution fit by $\mathrm{d\;log(Z)}/\mathrm{dz} \approx - 0.064$. Our simulations indicate that the metal retention efficiency of the interstellar medium (ISM) is low: a majority of gas-phase metals ($\sim$ 85 per cent at $z=0$) live outside of the ISM, either in an extended gas disk, the circumgalactic medium, or outside the halo. Nevertheless, the redshift evolution in the simulated MZR normalization is driven by the higher gas fractions of high redshift galaxies, not by changes to the metal retention efficiency. The scatter in the simulated MZR contains a clear correlation with the gas-mass or star formation rate of the system, in agreement with the observed fundamental metallicity relation. The scatter in the MZR is driven by a competition between periods of enrichment- and accretion-dominated metallicity evolution. We expect that while the normalization of the MZR declines with redshift, the strength of the correlation between metallicity and gas-mass at fixed stellar mass is not a strong function of redshift. Our results indicate that the "regulator" style models are best suited for simultaneously explaining the shape, redshift evolution, and existence of correlated scatter with gas fraction about the MZR., Comment: 21 pages, 13 figures, submitted to MNRAS; The IllustrisTNG project website can be found at http://www.tng-project.org/

We study the population of supermassive black holes (SMBHs) and their effects on massive central galaxies in the IllustrisTNG cosmological hydrodynamical simulations of galaxy formation. The employed model for SMBH growth and feedback assumes a two-mode scenario in which the feedback from active galactic nuclei occurs through a kinetic, comparatively efficient mode at low accretion rates relative to the Eddington limit, and in the form of a thermal, less efficient mode at high accretion rates. We show that the quenching of massive central galaxies happens coincidently with kinetic-mode feedback, consistent with the notion that active supermassive black cause the low specific star formation rates observed in massive galaxies. However, major galaxy mergers are not responsible for initiating most of the quenching events in our model. Up to black hole masses of about $10^{8.5}\,{\rm M}_\odot$, the dominant growth channel for SMBHs is in the thermal mode. Higher mass black holes stay mainly in the kinetic mode and gas accretion is self-regulated via their feedback, which causes their Eddington ratios to drop, with SMBH mergers becoming the main channel for residual mass growth. As a consequence, the quasar luminosity function is dominated by rapidly accreting, moderately massive black holes in the thermal mode. We show that the associated growth history of SMBHs produces a low-redshift quasar luminosity function and a redshift zero black hole mass-stellar bulge mass relation in good agreement with observations, whereas the simulation tends to over-predict the high-redshift quasar luminosity function., Comment: 16 pages, 11 figures, submitted to MNRAS, the IllustrisTNG project website: www.tng-project.org, comments welcome

We analyze scaling relations and evolution histories of galaxy sizes in TNG100, part of the IllustrisTNG simulation suite. Observational qualitative trends of size with stellar mass, star-formation rate and redshift are reproduced, and a quantitative comparison of projected r-band sizes at 0~~10^{9.5}Msun, the evolution of the median main progenitor differs, with quenched galaxies hardly growing in median size before quenching, whereas main-sequence galaxies grow their median size continuously, thus opening a gap from the progenitors of quenched galaxies. This is partly because the main-sequence high-redshift progenitors of quenched z=0 galaxies are drawn from the lower end of the size distribution of the overall population of main-sequence high-redshift galaxies. (ii) Quenched galaxies with M_{*,z=0}>~10^{9.5}Msun experience a steep size growth on the size-mass plane after their quenching time, but with the exception of galaxies with M_{*,z=0}>~10^{11}Msun, the size growth after quenching is small in absolute terms, such that most of the size (and mass) growth of quenched galaxies (and its variation among them) occurs while they are still on the main-sequence. After they become quenched, the size growth rate of quenched galaxies as a function of time, as opposed to versus mass, is similar to that of main-sequence galaxies. Hence, the size gap is retained down to z=0., Comment: MNRAS, accepted. 22 pages, 14 figures. Key figures are 2, 5, and 9. www.tng-project.org

The distribution of metals in the intra-cluster medium encodes important information about the enrichment history and formation of galaxy clusters. Here we explore the metal content of clusters in IllustrisTNG - a new suite of galaxy formation simulations building on the Illustris project. Our cluster sample contains 20 objects in TNG100 - a ~(100 Mpc)^3 volume simulation with 2x1820^3 resolution elements, and 370 objects in TNG300 - a ~(300 Mpc)^3 volume simulation with 2x2500^3 resolution elements. The z=0 metallicity profiles agree with observations, and the enrichment history is consistent with observational data going beyond z~1, showing nearly no metallicity evolution. The abundance profiles vary only minimally within the cluster samples, especially in the outskirts with a relative scatter of ~15%. The average metallicity profile flattens towards the center, where we find a logarithmic slope of -0.1 compared to -0.5 in the outskirts. Cool core clusters have more centrally peaked metallicity profiles (~0.8 solar) compared to non-cool core systems (~0.5 solar), similar to observational trends. Si/Fe and O/Fe radial profiles follow positive gradients. The outer abundance profiles do not evolve below z~2, whereas the inner profiles flatten towards z=0. More than ~80% of the metals in the intra-cluster medium have been accreted from the proto-cluster environment, which has been enriched to ~0.1 solar already at z~2. We conclude that the intra-cluster metal distribution is uniform among our cluster sample, nearly time-invariant in the outskirts for more than 10 Gyr, and forms through a universal enrichment history., Comment: 21 pages, 16 figures. MNRAS accepted. The IllustrisTNG project website can be found at http://www.tng-project.org/

The IllustrisTNG project is a new suite of cosmological magneto-hydrodynamical simulations of galaxy formation performed with the Arepo code and updated models for feedback physics. Here we introduce the first two simulations of the series, TNG100 and TNG300, and quantify the stellar mass content of about 4000 massive galaxy groups and clusters ($10^{13} \leq M_{\rm 200c}/M_{\rm sun} \leq 10^{15}$) at recent times ($z \leq 1$). The richest clusters have half of their total stellar mass bound to satellite galaxies, with the other half being associated with the central galaxy and the diffuse intra-cluster light. The exact ICL fraction depends sensitively on the definition of a central galaxy's mass and varies in our most massive clusters between 20 to 40% of the total stellar mass. Haloes of $5\times 10^{14}M_{\rm sun}$ and above have more diffuse stellar mass outside 100 kpc than within 100 kpc, with power-law slopes of the radial mass density distribution as shallow as the dark matter's ( $-3.5 < \alpha_{\rm 3D} < -3$). Total halo mass is a very good predictor of stellar mass, and vice versa: at $z=0$, the 3D stellar mass measured within 30 kpc scales as $\propto (M_{\rm 500c})^{0.49}$ with a $\sim 0.12$ dex scatter. This is possibly too steep in comparison to the available observational constraints, even though the abundance of TNG less massive galaxies ($< 10^{11}M_{\rm sun}$ in stars) is in good agreement with the measured galaxy stellar mass functions at recent epochs. The 3D sizes of massive galaxies fall too on a tight ($\sim$0.16 dex scatter) power-law relation with halo mass, with $r^{\rm stars}_{\rm 0.5} \propto (M_{\rm 500c})^{0.53}$. Even more fundamentally, halo mass alone is a good predictor for the whole stellar mass profiles beyond the inner few kpc, and we show how on average these can be precisely recovered given a single mass measurement of the galaxy or its halo., Comment: Accepted by MNRAS, updated to match published version. Highlights: Figures 5, 9, 11. The IllustrisTNG website can be found at http://www.tng-project.org/

Hydrodynamical simulations of galaxy formation have now reached sufficient volume to make precision predictions for clustering on cosmologically relevant scales. Here we use our new IllustrisTNG simulations to study the non-linear correlation functions and power spectra of baryons, dark matter, galaxies and haloes over an exceptionally large range of scales. We find that baryonic effects increase the clustering of dark matter on small scales and damp the total matter power spectrum on scales up to k ~ 10 h/Mpc by 20%. The non-linear two-point correlation function of the stellar mass is close to a power-law over a wide range of scales and approximately invariant in time from very high redshift to the present. The two-point correlation function of the simulated galaxies agrees well with SDSS at its mean redshift z ~ 0.1, both as a function of stellar mass and when split according to galaxy colour, apart from a mild excess in the clustering of red galaxies in the stellar mass range 10^9-10^10 Msun/h^2. Given this agreement, the TNG simulations can make valuable theoretical predictions for the clustering bias of different galaxy samples. We find that the clustering length of the galaxy auto-correlation function depends strongly on stellar mass and redshift. Its power-law slope gamma is nearly invariant with stellar mass, but declines from gamma ~ 1.8 at redshift z=0 to gamma ~ 1.6 at redshift z ~ 1, beyond which the slope steepens again. We detect significant scale-dependencies in the bias of different observational tracers of large-scale structure, extending well into the range of the baryonic acoustic oscillations and causing nominal (yet fortunately correctable) shifts of the acoustic peaks of around ~5%., Comment: accepted by MNRAS, 21 figures, 24 pages, updated to match published version. For IllustrisTNG visualizations see http://www.tng-project.org

We introduce the first two simulations of the IllustrisTNG project, a next generation of cosmological magnetohydrodynamical simulations, focusing on the optical colors of galaxies. We explore TNG100, a rerun of the original Illustris box, and TNG300, which includes 2x2500^3 resolution elements in a volume twenty times larger. Here we present first results on the galaxy color bimodality at low redshift. Accounting for the attenuation of stellar light by dust, we compare the simulated (g-r) colors of 10^9 < M*/Msun < 10^12.5 galaxies to the observed distribution from the Sloan Digital Sky Survey (SDSS). We find a striking improvement with respect to the original Illustris simulation, as well as excellent quantitative agreement in comparison to the observations, with a sharp transition in median color from blue to red at a characteristic M* ~ 10^10.5 Msun. Investigating the build-up of the color-mass plane and the formation of the red sequence, we demonstrate that the primary driver of galaxy color transition in the TNG model is supermassive blackhole feedback in its low-accretion state. Across the entire population we measure a median color transition timescale dt_green of ~1.6 Gyr, a value which drops for increasingly massive galaxies. We find signatures of the physical process of quenching: at fixed stellar mass, redder galaxies have lower SFRs, gas fractions, and gas metallicities; their stellar populations are also older and their large-scale interstellar magnetic fields weaker than in bluer galaxies. Finally, we measure the amount of stellar mass growth on the red sequence. Galaxies with M* > 10^11 Msun which redden at z<1 accumulate on average ~25% of their final z=0 mass post-reddening; at the same time, ~18% of such massive galaxies acquire half or more of their final stellar mass while on the red sequence., Comment: The IllustrisTNG project website is http://www.tng-project.org

We introduce the IllustrisTNG project, a new suite of cosmological magnetohydrodynamical simulations performed with the moving-mesh code AREPO employing an updated Illustris galaxy formation model. Here we focus on the general properties of magnetic fields and the diffuse radio emission in galaxy clusters. Magnetic fields are prevalent in galaxies, and their build-up is closely linked to structure formation. We find that structure formation amplifies the initial seed fields ($10^{-14}$ comoving Gauss) to the values observed in low-redshift galaxies ($1-10\,\mu{\rm G}$). The magnetic field topology is closely connected to galaxy morphology such that irregular fields are hosted by early-type galaxies, while large-scale, ordered fields are present in disc galaxies. Using two simple models for the energy distribution of relativistic electrons we predict the diffuse radio emission of $280$ clusters with a baryonic mass resolution of $1.1\times 10^{7}\,{\rm M_{\odot}}$, and generate mock observations for VLA, LOFAR, ASKAP and SKA. Our simulated clusters show extended radio emission, whose detectability correlates with their virial mass. We reproduce the observed scaling relations between total radio power and X-ray emission, $M_{500}$, and the Sunyaev-Zel'dovich $Y_{\rm 500}$ parameter. The radio emission surface brightness profiles of our most massive clusters are in reasonable agreement with VLA measurements of Coma and Perseus. Finally, we discuss the fraction of detected extended radio haloes as a function of virial mass and source count functions for different instruments. Overall our results agree encouragingly well with observations, but a refined analysis requires a more sophisticated treatment of relativistic particles in large-scale galaxy formation simulations., Comment: 28 pages, 18 figures, 2 tables, 3 appendices. Added a new relativistic electron energy parametrization and text modifications to match the accepted version for publication in MNRAS. More information, images and movies of the IllustrisTNG project can be found at http://www.tng-project.org

The distribution of elements in galaxies provides a wealth of information about their production sites and their subsequent mixing into the interstellar medium. Here we investigate the distribution of elements within stars in the IllustrisTNG simulations. In particular, we analyze the abundance ratios of magnesium and europium in Milky Way-like galaxies from the TNG100 simulation (stellar masses ${\log} (M_\star / {\rm M}_\odot) \sim 9.7 - 11.2$). As abundances of magnesium and europium for individual stars in the Milky Way are observed across a variety of spatial locations and metallicities, comparison with the stellar abundances in our more than $850$ Milky Way-like galaxies provides stringent constraints on our chemical evolutionary methods. To this end we use the magnesium to iron ratio as a proxy for the effects of our SNII and SNIa metal return prescription, and a means to compare our simulated abundances to a wide variety of galactic observations. The europium to iron ratio tracks the rare ejecta from neutron star -- neutron star mergers, the assumed primary site of europium production in our models, which in turn is a sensitive probe of the effects of metal diffusion within the gas in our simulations. We find that europium abundances in Milky Way-like galaxies show no correlation with assembly history, present day galactic properties, and average galactic stellar population age. In general, we reproduce the europium to iron spread at low metallicities observed in the Milky Way, with the level of enhancement being sensitive to gas properties during redshifts $z \approx 2-4$. We show that while the overall normalization of [Eu/Fe] is susceptible to resolution and post-processing assumptions, the relatively large spread of [Eu/Fe] at low [Fe/H] when compared to that at high [Fe/H] is very robust., Comment: 18 pages, 14 figures, accepted to MNRAS

Observed galactic disks have specific angular momenta similar to expectations for typical dark matter halos in $\Lambda$CDM. Cosmological hydrodynamical simulations have recently reproduced this similarity in large galaxy samples by including strong galactic winds, but the exact mechanism that achieves this is not yet clear. Here we present an analysis of key aspects contributing to this relation: angular momentum selection and evolution of Lagrangian mass elements as they accrete onto dark matter halos, condense into Milky Way-scale galaxies, and join the $z=0$ stellar phase. We contrast this evolution in the Illustris simulation with that in a simulation without galactic winds, where the $z=0$ angular momentum is $\approx0.6$ dex lower. We find that winds induce differences between these simulations in several ways: increasing angular momentum, preventing angular momentum loss, and causing $z=0$ stars to sample the accretion-time angular momentum distribution of baryons in a biased way. In both simulations, gas loses on average $\approx0.4$ dex between accreting onto halos and first accreting onto central galaxies. In Illustris, this is followed by $\approx0.2$ dex gains in the `galactic wind fountain' and no further net evolution past the final accretion onto the galaxy. Without feedback, further losses of $\approx0.2$ dex occur in the gas phase inside the galaxies. An additional $\approx0.15$ dex difference arises from feedback preferentially selecting higher angular momentum gas at accretion by expelling gas that is poorly aligned. These and additional effects of similar magnitude are discussed, suggesting a complex origin of the similarity between the specific angular momenta of galactic disks and typical halos., Comment: Accepted to ApJ. 13 pages, 10 figures. Key figures are 1, 2, and 3

We introduce an updated physical model to simulate the formation and evolution of galaxies in cosmological, large-scale gravity+magnetohydrodynamical simulations with the moving mesh code AREPO. The overall framework builds upon the successes of the Illustris galaxy formation model, and includes prescriptions for star formation, stellar evolution, chemical enrichment, primordial and metal-line cooling of the gas, stellar feedback with galactic outflows, and black hole formation, growth and multi-mode feedback. In this paper we give a comprehensive description of the physical and numerical advances which form the core of the IllustrisTNG (The Next Generation) framework. We focus on the revised implementation of the galactic winds, of which we modify the directionality, velocity, thermal content, and energy scalings, and explore its effects on the galaxy population. As described in earlier works, the model also includes a new black hole driven kinetic feedback at low accretion rates, magnetohydrodynamics, and improvements to the numerical scheme. Using a suite of (25 Mpc $h^{-1}$)$^3$ cosmological boxes we assess the outcome of the new model at our fiducial resolution. The presence of a self-consistently amplified magnetic field is shown to have an important impact on the stellar content of $10^{12} M_{\rm sun}$ haloes and above. Finally, we demonstrate that the new galactic winds promise to solve key problems identified in Illustris in matching observational constraints and affecting the stellar content and sizes of the low mass end of the galaxy population., Comment: 31 pages, 16 figures; published in MNRAS (Figs. 6 and 7 revised + additional minor changes to match accepted version)

Using the Illustris simulation, we follow thousands of elliptical galaxies back in time to identify how the dichotomy between fast and slow rotating ellipticals (FRs and SRs) develops. Comparing to the $\textrm{ATLAS}^\textrm{3D}$ survey, we show that Illustris reproduces similar elliptical galaxy rotation properties, quantified by the degree of ordered rotation, $\lambda_\textrm{R}$. There is a clear segregation between low-mass ($M_{\rm *} < 10^{11} M_{\rm \odot}$) ellipticals, which form a smooth distribution of FRs, and high-mass galaxies ($M_{\rm *} > 10^{11.5} M_{\rm \odot}$), which are mostly SRs, in agreement with observations. We find that SRs are very gas poor, metal rich and red in colour, while FRs are generally more gas rich and still star forming. We suggest that ellipticals begin naturally as FRs and, as they grow in mass, lose their spin and become SRs. While at $z = 1$, the progenitors of SRs and FRs are nearly indistinguishable, their merger and star formation histories differ thereafter. We find that major mergers tend to disrupt galaxy spin, though in rare cases can lead to a spin-up. No major difference is found between the effects of gas-rich and gas-poor mergers and the amount of minor mergers seem to have little correlation with galaxy spin. In between major mergers, lower-mass ellipticals, which are mostly gas-rich, tend to recover their spin by accreting gas and stars. For galaxies with $M_{\rm *}$ above $\sim 10^{11} M_{\rm \odot}$, this trend reverses; galaxies only retain or steadily lose their spin. More frequent mergers, accompanied by an inability to regain spin, lead massive ellipticals to lose most of ordered rotation and transition from FRs to SRs., Comment: 26 pages, 21 figures, MNRAS accepted