Ewald Puchwein

Shaping the galaxy stellar mass function with supernova- and AGN-driven winds

Cosmological hydrodynamical simulations of galaxy formation in representative regions of the Universe typically need to resort to subresolution models to follow some of the feedback processes crucial for galaxy formation. Here, we show that an energy-driven outflow model in which the wind velocity decreases and the wind mass loading increases in low-mass galaxies, as suggested by observations, can produce a good match to the low-mass end of the observed galaxy stellar mass function. The high-mass end can be recovered simultaneously if feedback from active galactic nuclei (AGN) and a correction for diffuse stellar light plausibly missed in observations are included. At the same time, our model is in good agreement with the stellar mass functions at redshifts z=1 and z=2, and with the observed redshift evolution of the cosmic star formation rate density. In addition, it accurately reproduces the observed gas to stellar mass ratios and specific star formation rates of galaxies as a function of their stellar mass. This agreement with a diverse set of data marks significant progress in hydrodynamically modelling the formation of a representative galaxy population. It also suggests that the mass flux in real galactic winds should strongly increase towards low-mass galaxies. Without this assumption, an overproduction of galaxies at the faint-end of the galaxy luminosity function seems inevitable in our models.

Lyman α emitters gone missing: evidence for late reionization?

We combine high resolution hydrodynamical simulations with an intermediate resolution, dark matter only simulation and an analytical model for the growth of ionized regions to estimate the large scale distribution and redshift evolution of the visibility of Lyα emission in 6<=z<=8 galaxies. The inhomogeneous distribution of neutral hydrogen during the reionization process results in significant fluctuations in the Lyα transmissivity on large scales. The transmissivity depends not only on the ionized fraction of the intergalactic medium by volume and the amplitude of the local ionizing background, but is also rather sensitive to the evolution of the relative velocity shift of the Lyα emission line due to resonant scattering. We reproduce a decline in the space density of Lyα emitting galaxies as rapid as observed with a rather rapidly evolving neutral fraction between z=6-8, and a typical Lyα line velocity offset of 100 km/s redward of systemic at z=6 which decreases toward higher redshift. The new (02/2015) Planck results indicate such a recent end to reionization is no longer disfavoured by constraints from the cosmic microwave background.

Modified-Gravity-gadget: a new code for cosmological hydrodynamical simulations of modified gravity models

We present a new massively parallel code for N-body and cosmological hydrodynamical simulations of modified gravity models. The code employs a multigrid-accelerated Newton-Gauss-Seidel relaxation solver on an adaptive mesh to efficiently solve for perturbations in the scalar degree of freedom of the modified gravity model. As this new algorithm is implemented as a module for the P-Gadget3 code, it can at the same time follow the baryonic physics included in P-Gadget3, such as hydrodynamics, radiative cooling and star formation. We demonstrate that the code works reliably by applying it to simple test problems that can be solved analytically, as well as by comparing cosmological simulations to results from the literature. Using the new code, we perform the first non-radiative and radiative cosmological hydrodynamical simulations of an f(R)-gravity model. We also discuss the impact of AGN feedback on the matter power spectrum, as well as degeneracies between the influence of baryonic processes and modifications of gravity.

Cosmic Degeneracies I: Joint N-body Simulations of Modified Gravity and Massive Neutrinos

We present the first suite of cosmological N-body simulation s that simultaneously include the effects of two different and theoretically indepe ndent extensions of the standard �CDM cosmological scenario ‐ namely an f(R) theory of Modified Gravity (MG) and a cosmological background of massive neutrinos ‐ with the aim to investigate their possible observational degeneracies. We focus on three basic statistics of the large-scale matter distribution, more specifically the nonlinear matter power spectr um, the halo mass function, and the halo bias. Our results show that while these two extended models separately determine very prominent and potentially detectable features in all t he three statistics, when we allow them to be simultaneously at work these features are strongly suppressed. In particular, when an f(R) gravity model with fR0 = 1 × 10 4 is combined with a total neutrino mass of

The photoheating of the intergalactic medium in synthesis models of the UV background

We compare cosmological hydrodynamical simulations combined with the homogeneous metagalactic UV background (UVB) of Haardt & Madau (2012) (HM2012) to observations of the Lyman-alpha forest that are sensitive to the thermal and ionization state of the intergalactic medium (IGM). The transition from optically thick to thin photoheating predicted by the simple one-zone, radiative transfer model implemented by HM2012 predicts a thermal history that is in remarkably good agreement with the observed rise of the IGM temperature at z~3 if we account for the expected evolution of the volume filling factor of HeIII. Our simulations indicate that there may be, however, some tension between the observed peak in the temperature evolution and the rather slow evolution of the HeII opacities suggested by recent Hubble Space Telescope/COS measurements. The HM2012 UVB also underpredicts the metagalactic hydrogen photoionization rate required by our simulations to match the observed opacity of the forest at z>4 and z<2.

Scaling relations and mass bias in hydrodynamical f (R) gravity simulations of galaxy clusters

We investigate the impact of chameleon-type f(R) gravity models on the properties of galaxy clusters and groups. Our f(R) simulations follow for the first time also the hydrodynamics of the intracluster and intragroup medium. This allows us to assess how f(R) gravity alters the X-ray scaling relations of clusters and how hydrostatic and dynamical mass estimates are biased when modifications of gravity are ignored in their determination. We find that velocity dispersions and intracluster medium temperatures are both increased by up to 1/3 in f(R) gravity in low-mass halos, while the difference disappears in massive objects. The mass scale of the transition depends on the background value f_R0 of the scalar degree of freedom. These changes in temperature and velocity dispersion alter the mass-temperature and X-ray luminosity-temperature scaling relations and bias dynamical and hydrostatic mass estimates that do not explicitly account for modified gravity towards higher values. Recently, a relative enhancement of X-ray compared to weak lensing masses was found by the Planck Collaboration (2013). We demonstrate that an explanation for this offset may be provided by modified gravity and the associated bias effects, which interestingly are of the required size. Finally, we find that the abundance of subhalos at fixed cluster mass is only weakly affected by f(R) gravity.

nIFTy galaxy cluster simulations – I. Dark matter and non-radiative models

We have simulated the formation of a galaxy cluster in a Ʌ cold dark matter universe using 13 different codes modelling only gravity and non-radiative hydrodynamics (RAMSES, ART, AREPO, HYDRA and nine incarnations of GADGET). This range of codes includes particle-based, moving and fixed mesh codes as well as both Eulerian and Lagrangian fluid schemes. The various GADGET implementations span classic and modern smoothed particle hydrodynamics (SPH) schemes. The goal of this comparison is to assess the reliability of cosmological hydrodynamical simulations of clusters in the simplest astrophysically relevant case, that in which the gas is assumed to be non-radiative. We compare images of the cluster at z = 0, global properties such as mass and radial profiles of various dynamical and thermodynamical quantities. The underlying gravitational framework can be aligned very accurately for all the codes allowing a detailed investigation of the differences that develop due to the various gas physics implementations employed. As expected, the mesh-based codes RAMSES, ART and AREPO form extended entropy cores in the gas with rising central gas temperatures. Those codes employing classic SPH schemes show falling entropy profiles all the way into the very centre with correspondingly rising density profiles and central temperature inversions. We show that methods with modern SPH schemes that allow entropy mixing span the range between these two extremes and the latest SPH variants produce gas entropy profiles that are essentially indistinguishable from those obtained with grid-based methods.

Modified Gravity N-body Code Comparison Project

Self-consistent N-body simulations of modified gravity models are a key ingredient to obtain rigorous constraints on deviations from general relativity using large-scale structure observations. This paper provides the first detailed comparison of the results of different N-body codes for the f (R), Dvali-Gabadadze-Porrati and Symmetron models, starting from the same initial conditions. We find that the fractional deviation of the matter power spectrum from Lambda cold dark matter agrees to better than 1 per cent up to k ˜ 5-10 h Mpc-1 between the different codes. These codes are thus able to meet the stringent accuracy requirements of upcoming observational surveys. All codes are also in good agreement in their results for the velocity divergence power spectrum, halo abundances and halo profiles. We also test the quasi-static limit, which is employed in most modified gravity N-body codes, for the Symmetron model for which the most significant non-static effects among the models considered are expected. We conclude that this limit is a very good approximation for all of the observables considered here.

Intracluster stars in simulations with active galactic nucleus feedback: Intracluster stars in simulations

We use a set of high-resolution hydrodynamical simulations of clusters of galaxies to study the build-up of the intracluster light (ICL), an interesting and likely significant component of their total stellar mass. Our sample of groups and clusters includes active galactic nucleus (AGN) feedback and is of high enough resolution to accurately resolve galaxy populations down to the smallest galaxies that are expected to significantly contribute to the stellar mass budget. We describe and test four different methods to identify the ICL in cluster simulations, thereby allowing us to assess the reliability of the measurements. For all of the methods, we consistently find a very significant ICL stellar fraction (∼45 per cent) which exceeds the values typically inferred from observations. However, we show that this result is robust with respect to numerical resolution and integration accuracy, remarkably insensitive to changes in the star formation model, and almost independent of halo mass. It is also almost invariant when black hole growth is included, even though AGN feedback successfully prevents excessive overcooling in clusters and leads to a drastically improved agreement of the simulated cluster galaxy population with observations. In particular, the luminosities of central cluster galaxies and the ages of their stellar populations are much more realistic when including AGN. In the light of these findings, it appears challenging to construct a simulation model that simultaneously matches the cluster galaxy population and at the same time produces a low ICL component. We find that intracluster stars are preferentially stripped in a clusters densest region from massive galaxies that fall into the forming cluster at z > 1. Surprisingly, some of the intracluster stars also form in the intracluster medium inside cold gas clouds that are stripped out of infalling galaxies.

The Sherwood simulation suite: overview and data comparisons with the Lyman α forest at redshifts 2 ≤ z ≤ 5

We introduce a new set of large-scale, high-resolution hydrodynamical simulations of the intergalactic medium: the Sherwood simulation suite. These are performed in volumes of 103–1603h−3 comoving Mpc3, span almost four orders of magnitude in mass resolution with up to 17.2 billion particles, and employ a variety of physics variations including warm dark matter and galactic outflows. We undertake a detailed comparison of the simulations to high-resolution, high signal-to-noise observations of the Ly α forest over the redshift range 2 ≤ z ≤ 5. The simulations are in very good agreement with the observational data, lending further support to the paradigm that the Ly α forest is a natural consequence of the web-like distribution of matter arising in Λcold dark matter cosmological models. Only a small number of minor discrepancies remain with respect to the observational data. Saturated Ly α absorption lines with column densities NHI>1014.5cm−2 at 2 4. Finally, the temperature of intergalactic gas in the simulations may be slightly too low at z = 2.7 and a flatter temperature–density relation is required at z = 2.4, consistent with the expected effects of non-equilibrium ionization during He ii reionization.