Suite Of Simulations Research Articles

On the correlations of galaxy peculiar velocities and their covariance

ABSTRACT Measurements of the peculiar velocities of large samples of galaxies enable new tests of the standard cosmological model, including determination of the growth rate of cosmic structure that encodes gravitational physics. With the size of such samples now approaching hundreds of thousands of galaxies, complex statistical analysis techniques and models are required to extract cosmological information. In this paper, we summarize how correlation functions between galaxy velocities, and with the surrounding large-scale structure, may be utilized to test cosmological models. We present new determinations of the analytical covariance between such correlation functions, which may be useful for cosmological likelihood analyses. The statistical model we use to determine these covariances includes the sample selection functions, observational noise, curved-sky effects, and redshift-space distortions. By comparing these covariance determinations with corresponding estimates from large suites of cosmological simulations, we demonstrate that these analytical models recover the key features of the covariance between different statistics and separations, and produce similar measurements of the growth rate of structure.

Feedback-driven anisotropy in the circumgalactic medium for quenching galaxies in the simba simulations

ABSTRACT We use the simba galaxy formation simulation suite to explore anisotropies in the properties of circumgalactic gas that result from accretion and feedback processes. We particularly focus on the impact of bipolar active galactic nuclei (AGNs) jet feedback as implemented in simba , which quenches galaxies and has a dramatic effect on large-scale gas properties. We show that jet feedback at low redshifts is most common in the stellar mass range (1–5) × 1010 M⊙, so we focus on galaxies with active jets in this mass range. In comparison to runs without jet feedback, jets cause lower densities and higher temperatures along the galaxy minor axis (simba jet direction) at radii $\gtrsim 0.5r_{200c}-4r_{200c}$ and beyond. This effect is less apparent at higher or lower stellar masses, and is strongest within green valley galaxies. The metallicity also shows strong anisotropy out to large scales, driven by star formation feedback. We find substantially stronger anisotropy at $\lesssim 0.5r_{200c}$, but this also exists in runs with no explicit feedback, suggesting that it is due to anisotropic accretion. Finally, we explore anisotropy in the bulk radial motion of the gas, finding that both star formation and AGN wind feedback contribute to pushing the gas outwards along the minor axis at $\lesssim 1$ Mpc, but AGN jet feedback further causes bulk outflow along the minor axis out to several Mpc, which drives quenching via gas starvation. These results provide observational signatures for the operation of AGN feedback in galaxy quenching.

GJ 357 d: Potentially Habitable World or Agent of Chaos?

Multiplanet systems provide important laboratories for exploring dynamical interactions within the range of known exoplanetary system architectures. One such system is GJ 357, consisting of a low-mass host star and three orbiting planets, the outermost (planet d) of which does not transit but lies within the habitable zone (HZ) of the host star. The minimum mass of planet d causes its nature to be unknown, both in terms of whether it is truly terrestrial and if it is a candidate for harboring surface liquid water. Here, we use three sectors of photometry from the Transiting Exoplanet Survey Satellite to show that planets c and d do not transit the host star, and therefore may have masses higher than the derived minimum masses. We present the results for a suite of dynamical simulations that inject an Earth-mass planet within the HZ of the system for three different orbital and mass configurations of planet d. These results show that planet d, rather than being a potentially habitable planet, is likely a source of significant orbital instability for other potential terrestrial planets within the HZ. We find that relatively small eccentricities of planet d cause a majority of the HZ to be unstable for an Earth-mass planet. These results highlight the importance of dynamical stability for systems that are prioritized in the context of planetary habitability.

Beyond the 3rd moment: a practical study of using lensing convergence CDFs for cosmology with DES Y3

ABSTRACT Widefield surveys probe clustered scalar fields – such as galaxy counts, lensing potential, etc. – which are sensitive to different cosmological and astrophysical processes. Constraining such processes depends on the statistics that summarize the field. We explore the cumulative distribution function (CDF) as a summary of the galaxy lensing convergence field. Using a suite of N-body light-cone simulations, we show the CDFs’ constraining power is modestly better than the second and third moments, as CDFs approximately capture information from all moments. We study the practical aspects of applying CDFs to data, using the Dark Energy Survey (DES Y3) data as an example, and compute the impact of different systematics on the CDFs. The contributions from the point spread function and reduced shear approximation are $\lesssim 1~{{\ \rm per\ cent}}$ of the total signal. Source clustering effects and baryon imprints contribute 1–10 per cent. Enforcing scale cuts to limit systematics-driven biases in parameter constraints degrade these constraints a noticeable amount, and this degradation is similar for the CDFs and the moments. We detect correlations between the observed convergence field and the shape noise field at 13σ. The non-Gaussian correlations in the noise field must be modelled accurately to use the CDFs, or other statistics sensitive to all moments, as a rigorous cosmology tool.

The FLAMINGO project: revisiting the S8 tension and the role of baryonic physics

ABSTRACT A number of recent studies have found evidence for a tension between observations of large-scale structure (LSS) and the predictions of the standard model of cosmology with the cosmological parameters fit to the cosmic microwave background (CMB). The origin of this ‘S8 tension’ remains unclear, but possibilities include new physics beyond the standard model, unaccounted for systematic errors in the observational measurements and/or uncertainties in the role that baryons play. Here, we carefully examine the latter possibility using the new FLAMINGO suite of large-volume cosmological hydrodynamical simulations. We project the simulations onto observable harmonic space and compare with observational measurements of the power and cross-power spectra of cosmic shear, CMB lensing, and the thermal Sunyaev-Zel’dovich (tSZ) effect. We explore the dependence of the predictions on box size and resolution and cosmological parameters, including the neutrino mass, and the efficiency and nature of baryonic ‘feedback’. Despite the wide range of astrophysical behaviours simulated, we find that baryonic effects are not sufficiently large to remove the S8 tension. Consistent with recent studies, we find the CMB lensing power spectrum is in excellent agreement with the standard model, while the cosmic shear power spectrum, tSZ effect power spectrum, and the cross-spectra between shear, CMB lensing, and the tSZ effect are all in varying degrees of tension with the CMB-specified standard model. These results suggest that some mechanism is required to slow the growth of fluctuations at late times and/or on non-linear scales, but that it is unlikely that baryon physics is driving this modification.

An emulator-based halo model in modified gravity – I. The halo concentration–mass relation and density profile

ABSTRACT In this series of papers, we present an emulator-based halo model for the non-linear clustering of galaxies in modified gravity cosmologies. In the first paper, we present emulators for the following halo properties: the halo mass function, concentration–mass relation and halo-matter cross-correlation function. The emulators are trained on data extracted from the forge and bridge suites of N-body simulations, respectively, for two modified gravity (MG) theories: f(R) gravity, and the DGP model, varying three standard cosmological parameters Ωm0, H0, σ8, and one MG parameter, either $\bar{f}_{R0}$ or rc. Our halo property emulators achieve an accuracy of ${\lesssim}1\ \hbox{per cent}$ on independent test data sets. We demonstrate that the emulators can be combined with a galaxy–halo connection prescription to accurately predict the galaxy–galaxy and galaxy–matter correlation functions using the halo model framework.

Identifying the possible ex-situ origin of the globular clusters of the Milky Way: a kinematic study

ABSTRACT This is the second paper in a series, which studies the likelihood that some globular clusters (GCs) of the Milky Way (MW) could have originated from a dwarf satellite galaxy (DSG). Using a large suite of three-body simulations we determine the present-day orbital properties of 154 GCs that could have escaped from 41 MW DSGs over the past $8\ \mathrm{Gyr}$. For the MW we considered two sets of static and dynamic models which account for the sustained growth of the MW since its birth. We focus on the Magellanic clouds and Sagittarius. We compare the apogalactic distance, eccentricity, and orbital inclination of the MW GCs with those of runaway GCs from DSGs, to constrain their possible ex-situ origin. We observe a positive correlation between a DSG mass and the dispersion of its runaway GCs in the orbital parameter space of (Rap, e). We provide tables of the identified MW GCs and their likely associated progenitors. In total, we find 29 (19 per cent) MW GCs which could be kinematically associated with MW DSGs. We report, for the first time, 6 and 10 new associations with the Large Magellanic Cloud and the Sagittarius, respectively. For the Sagittarius we predict a concentration of runaway GCs at large apogalactic distances of $R_\mathrm{ap}\approx 275-375 \ \mathrm{kpc}$, e ≈ 0.8, and a relative inclination of Δθ ≈ 20°. So far, there has not been any observed GCs with such orbital elements. Complemented with photometric and spectroscopic observations, and cosmological simulations, the findings from the present study could conclusively settle the debate over the in-situ versus ex-situ origin of the MW GCs.

Relative importance of forcings and feedbacks in the Holocene temperature conundrum

The pre-industrial Holocene provides the backdrop for the emergence of civilisations and the starting point of anthropogenic climate change. Several reconstructions show an early Holocene warming that was followed by cooling for several thousand years before Industrialisation. In contrast climate simulations show warming throughout the Holocene. Whilst reconstructed trends over ocean can be reconciled with warming through either seasonality or uncertainties, a consistent explanation for cooling trends over some land areas is missing. We present a suite of transient Holocene climate model simulations with a coupled general circulation model and show that a widespread mid- to late-Holocene cooling emerges over some regions of the northern hemisphere with the inclusion of anthropogenic land-use. This is mostly because in regions of prescribed late Holocene deforestation, the simulated early to mid-Holocene is characterised by a lower albedo than the late Holocene. Whilst this cooling through time can quantitatively explain some regional aspects of the reconstructions, the model-data agreement remains imperfect, and differences between reconstructions also hinders the evaluation. Moreover, model-dependency in the response of several feedbacks, particularly sea-ice, but potentially also clouds, means that it is difficult to uniquely attribute Holocene temperature evolution to specific factors. Future work should aim to derive a consensus-signal including uncertainties from the available proxy data which could be used to fingerprint simulations covering a range of plausible feedback strengths.

Constraining Circumgalactic Turbulence with QSO Absorption Line Measurements

Our knowledge of the circumgalactic medium (CGM) is mostly based on quasar absorption line measurements. These have uncovered a multiphase medium that is likely highly turbulent, but constraints of this turbulence are limited to measurements of the nonthermal width of absorption line components (b turb) and the line-of-sight velocity dispersion between components (σ LOS). Here we analyze a suite of CGM simulations to determine how well these indirect measures are related to the underlying CGM. Our simulations track the nonequilibrium evolution of all commonly observed ions and consist of two main types: small-scale simulations of regions of homogenous CGM turbulence and global simulations of inhomogenous turbulence throughout a galactic halo. From each simulation, we generate mock spectra of Si ii, Si iv, C iv, and O vi, which allow us to directly compare b turb and σ LOS to the true line-of-sight turbulence (σ 1D). In the small-scale simulations, b turb is only weakly correlated with σ 1D, likely because it measures random motions within individual warm CGM clouds, which do not sample the overall random motions. Meanwhile, σ LOS and σ 1D are strongly correlated with σ 1D ≈ σ LOS + 10 km s−1 in the densest regions we simulated, although the strength of this correlation depended weakly on the gas phase being probed. Our large-scale simulations also indicate that b turb and σ 1D are largely uncorrelated and that σ 1D ≈ σ LOS + 10 km s−1 on average, although it varies along individual sight lines. Moreover, the σ LOS distributions from our global simulations are similar to recent observations, suggesting that this quantity may provide useful constraints on circumgalactic turbulence regardless of the axis probed.

Characterizing the velocity-space signature of electron Landau damping

Plasma turbulence plays a critical role in the transport of energy from large-scale magnetic fields and plasma flows to small scales, where the dissipated turbulent energy ultimately leads to heating of the plasma species. A major goal of the broader heliophysics community is to identify the physical mechanisms responsible for the dissipation of the turbulence and to quantify the consequent rate of plasma heating. One of the mechanisms proposed to damp turbulent fluctuations in weakly collisional space and astrophysical plasmas is electron Landau damping. The velocity-space signature of electron energization by Landau damping can be identified using the recently developed field–particle correlation technique. Here, we perform a suite of gyrokinetic turbulence simulations with ion plasma beta values$\beta _i = 0.01, 0.1, 1$and$10$and use the field–particle correlation technique to characterize the features of the velocity-space signatures of electron Landau damping in turbulent plasma conditions consistent with those observed in the solar wind and planetary magnetospheres. We identify the key features of the velocity-space signatures of electron Landau damping as a function of varying plasma$\beta _i$to provide a critical framework for interpreting the results of field–particle correlation analysis ofin situspacecraft observations of plasma turbulence.

The early evolution of magnetar rotation – II. Rapidly rotating magnetars: implications for gamma-ray bursts and superluminous supernovae

ABSTRACT Rapidly rotating magnetars have been associated with gamma-ray bursts (GRBs) and superluminous supernovae (SLSNe). Using a suite of two-dimensional magnetohydrodynamic simulations at fixed neutrino luminosity and a couple of evolutionary models with evolving neutrino luminosity and magnetar spin period, we show that magnetars are viable central engines for powering GRBs and SLSNe. We also present analytical estimates of the energy outflow rate from the proto-neutron star (PNS) as a function of polar magnetic field strength B0, PNS angular velocity Ω⋆, PNS radius R⋆, and mass outflow rate $\dot{M}$. We show that rapidly rotating magnetars with spin periods P⋆ ≲ 4 ms and polar magnetic field strength B0 ≳ 1015 G can release 1050 to 5 × 1051 erg of energy during the first ∼2 s of the cooling phase. Based on this result, it is plausible that sustained energy injection by magnetars through the relativistic wind phase can power GRBs. We also show that magnetars with moderate field strengths of B0 ≲ 5 × 1014 G do not release a large fraction of their rotational kinetic energy during the cooling phase and, hence, are not likely to power GRBs. Although we cannot simulate to times greater than ∼3–5 s after a supernova, we can hypothesize that moderate field strength magnetars can brighten the supernova light curves by releasing their rotational kinetic energy via magnetic dipole radiation on time-scales of days to weeks, since these do not expend most of their rotational kinetic energy during the early cooling phase.

MF-Box: multifidelity and multiscale emulation for the matter power spectrum

ABSTRACT We introduce MF-Box, an extended version of MFEmulator, designed as a fast surrogate for power spectra, trained using N-body simulation suites from various box sizes and particle loads. To demonstrate MF-Box’s effectiveness, we design simulation suites that include low-fidelity (LF) suites (L1 and L2) at 256 and $100 \, \rm {Mpc\, ~}h^{-1}$, each with 1283 particles, and a high-fidelity (HF) suite with 5123 particles at $256 \, \rm {Mpc\, ~}h^{-1}$, representing a higher particle load compared to the LF suites. MF-Box acts as a probabilistic resolution correction function, learning most of the cosmological dependencies from L1 and L2 simulations and rectifying resolution differences with just three HF simulations using a Gaussian process. MF-Box successfully emulates power spectra from our HF testing set with a relative error of $\lt 3~{{\ \rm per\ cent}}$ up to $k \simeq 7 \, h\rm {Mpc}{^{-1}}$ at z ∈ [0, 3], while maintaining a cost similar to our previous multifidelity approach, which was accurate only up to z = 1. The addition of an extra LF node in a smaller box significantly improves emulation accuracy for MF-Box at $k \gt 2 \, h\rm {Mpc}{^{-1}}$, increasing it by a factor of 10. We conduct an error analysis of MF-Box based on computational budget, providing guidance for optimizing budget allocation per fidelity node. Our proposed MF-Box enables future surveys to efficiently combine simulation suites of varying quality, effectively expanding the range of emulation capabilities while ensuring cost efficiency.

The ratio of mesoscale convective system precipitation to total precipitation increases in future climate change scenarios

Mesoscale convective systems (MCSs) are a substantial source of precipitation in the eastern U.S. and may be sensitive to regional climatic change. We use a suite of convection-permitting climate simulations to examine possible changes in MCS precipitation. Specifically, annual and regional totals of MCS and non-MCS precipitation generated during a retrospective simulation are compared to end-of-21st-century simulations based on intermediate and extreme climate change scenarios. Both scenarios produce more MCS precipitation and less non-MCS precipitation, thus significantly increasing the proportion of precipitation associated with MCSs across the U.S.

Figure Rotation of IllustrisTNG Halos

We use the TNG50 and TNG50 dark matter (DM)-only simulations from the IllustrisTNG simulation suite to conduct an updated survey of halo figure rotation in the presence of baryons. We develop a novel methodology to detect coherent figure rotation about an arbitrary axis and for arbitrary durations, and we apply it to a catalog of 1577 DM halos from the DM-only run and 1396 DM halos from the DM+baryons (DM+B) run that are free of major mergers. Figure rotation was detected in 94% of DM-only halos and 82% of the DM+B halos. The pattern speeds of rotations lasting ≳1h −1 Gyr were log-normally distributed with medians of 0.25 h km s−1 kpc−1 for DM-only in agreement with past results, but 14% higher at 0.29 h km s−1 kpc−1 in the DM+B halos. We find that rotation axes are typically aligned with the halo minor or major axis in 57% of DM-only halos and in 62% of DM+B halos. The remaining rotation axes were not strongly aligned with any principal axis but typically lie in the plane containing the halo minor and major axes. Longer-lived rotations showed greater alignment with the halo minor axis in both simulations. Our results show that, in the presence of baryons, figure rotation is marginally less common, shorter-lived, faster, and better aligned with the minor axis than in DM-only halos. This updated understanding will be consequential for future efforts to constrain figure rotation in the Milky Way dark halo using the morphology and kinematics of tidal streams.

Hydrodynamical simulations of circumbinary accretion: balance between heating and cooling

ABSTRACT Hydrodynamical interaction in circumbinary discs (CBDs) plays a crucial role in various astrophysical systems, ranging from young stellar binaries to supermassive black hole binaries in galactic centres. Most previous simulations of binary-disc systems have adopted locally isothermal equation of state. In this study, we use the grid-based code Athena++ to conduct a suite of two-dimensional viscous hydrodynamical simulations of circumbinary accretion on a Cartesian grid, resolving the central cavity of the binary. The gas thermodynamics is treated by thermal relaxation towards an equilibrium temperature (based on the constant − β cooling ansatz, where β is the cooling time in units of the local Keplerian time). Focusing on equal mass, circular binaries in CBDs with (equilibrium) disc aspect ratio H/R = 0.1, we find that the cooling of the disc gas significantly influences the binary orbital evolution, accretion variability, and CBD morphology, and the effect depends sensitively on the disc viscosity prescriptions. When adopting a constant kinematic viscosity, a finite cooling time (β ≳ 0.1) leads to a binary inspiral as opposed to an outspiral and the CBD cavity becomes more symmetric. When adopting a dynamically varying α-viscosity, binary inspiral only occurs within a narrow range of cooling time (corresponding to β around 0.5).

First light and reionization epoch simulations (FLARES) IX: the physical mechanisms driving compact galaxy formation and evolution

ABSTRACT In the First Light And Reionization Epoch Simulations (Flares) suite of hydrodynamical simulations, we find the high-redshift (z > 5) intrinsic size–luminosity relation is, surprisingly, negatively sloped. However, after including the effects of dust attenuation, we find a positively sloped UV observed size–luminosity relation in good agreement with other simulated and observational studies. In this work, we extend this analysis to probe the underlying physical mechanisms driving the formation and evolution of the compact galaxies driving the negative size–mass/size–luminosity relation. We find the majority of compact galaxies (R1/2, ⋆ < 1 pkpc, which drive the negative slope of the size–mass relation, have transitioned from extended to compact sizes via efficient centralized cooling, resulting in high specific star formation rates in their cores. These compact stellar systems are enshrouded by non-star-forming gas distributions as much as 100 times larger than their stellar counterparts. By comparing with galaxies from the Eagle simulation suite, we find that these extended gas distributions ‘turn on’ and begin to form stars between z = 5 and 0 leading to increasing sizes, and thus the evolution of the size–mass relation from a negative to a positive slope. This explicitly demonstrates the process of inside-out galaxy formation in which compact bulges form earlier than the surrounding discs.

Aging in 10 Minutes: Do Age Simulation Suits Mimic Physical Decline in Old Age? Comparing Experimental Data with Established Reference Data

ABSTRACT Introduction Age simulation suits are increasingly used in health care education. However, empirical evidence that quantifies the simulated performance losses in established geriatric tests and compares those declines with reference data of older adults is scarce. Methods In a standardized lab setting, we compared performance of N = 61 participants (46 middle-aged, 15 young adults) with and without age simulation suit, for example in the Timed Up and Go Test (+dual task), Short Physical Performance Battery, grip strength, and 30-Second-Chair- Standing Test. Additionally, we compared the results with suit to established reference values of older adults in different age groups. Results Reduced performance was observed in both groups when wearing the suit, yet to different degrees dependent on the assessment and user age. For one, larger declines were observed in more challenging and complex tasks across age groups. In addition, comparisons with reference values revealed age-differential “instant aging” effects. Discussion A simulated “fourth age,” where frailty and impairments are accumulating, was not reached in the majority of assessments, especially not among younger participants. In conclusion, existing age simulation suits may have some educational and empathy potential, but so far, they fail in simulating the age period with most serious functional loss.

Mglens: Modified gravity weak lensing simulations for emulation-based cosmological inference

ABSTRACT We present mglens, a large series of modified gravity lensing simulations tailored for cosmic shear data analyses and forecasts in which cosmological and modified gravity parameters are varied simultaneously. Based on the forge and bridgeN-body simulation suites presented in companion papers, we construct 100 × 5000 deg2 of mock Stage-IV lensing data from two 4D Latin hypercubes that sample cosmological and gravitational parameters in f(R) and nDGP gravity, respectively. These are then used to validate our inference analysis pipeline based on the lensing power spectrum, exploiting our implementation of these modified gravity models within the cosmosis cosmological inference package. Sampling this new likelihood, we find that cosmic shear can achieve 95 per cent CL constraints on the modified gravity parameters of log$_{10}[f_{R_0}] \lt $ −4.77 and log10[H0rc] > 0.09, after marginalizing over intrinsic alignments of galaxies and including scales up to ℓ = 5000. We also investigate the impact of photometric uncertainty, scale cuts, and covariance matrices. We finally explore the consequences of analysing mglens data with the wrong gravity model, and report catastrophic biases for a number of possible scenarios. The Stage-IV mglens simulations, the forge and bridge emulators and the cosmosis interface modules will be made publicly available upon journal acceptance.

Painting baryons on to N-body simulations of galaxy clusters with image-to-image deep learning

ABSTRACT Galaxy cluster mass functions are a function of cosmology, but mass is not a direct observable, and systematic errors abound in all its observable proxies. Mass-free inference can bypass this challenge, but it requires large suites of simulations spanning a range of cosmologies and models for directly observable quantities. In this work, we devise a U-net – an image-to-image machine learning algorithm – to ‘paint’ the illustristng model of baryons on to dark matter-only (DMO) simulations of galaxy clusters. Using 761 galaxy clusters with M200c ≳ 1014 M⊙ from the TNG300 simulation at z < 1, we train the algorithm to read in maps of projected dark matter mass and output maps of projected gas density, temperature, and X-ray flux. Despite being trained on individual images, the model reproduces the true scaling relation and scatter for the MDM–LX, as well as the distribution functions of the cluster X-ray luminosity and gas mass. For just one decade in cluster mass, the model reproduces three orders of magnitude in LX. The model is biased slightly high when using dark matter maps from the DMO simulation. The model performs well on inputs from TNG300-2, whose mass resolution is eight times coarser; further degrading the resolution biases the predicted luminosity function high. We conclude that U-net-based baryon painting is a promising technique to build large simulated cluster catalogues, which can be used to improve cluster cosmology by combining existing full-physics and large N-body simulations.

Emulating radiative transfer with artificial neural networks

ABSTRACT Forward-modeling observables from galaxy simulations enables direct comparisons between theory and observations. To generate synthetic spectral energy distributions (SEDs) that include dust absorption, re-emission, and scattering, Monte Carlo radiative transfer is often used in post-processing on a galaxy-by-galaxy basis. However, this is computationally expensive, especially if one wants to make predictions for suites of many cosmological simulations. To alleviate this computational burden, we have developed a radiative transfer emulator using an artificial neural network (ANN), ANNgelina, that can reliably predict SEDs of simulated galaxies using a small number of integrated properties of the simulated galaxies: star formation rate, stellar and dust masses, and mass-weighted metallicities of all star particles and of only star particles with age <10 Myr. Here, we present the methodology and quantify the accuracy of the predictions. We train the ANN on SEDs computed for galaxies from the IllustrisTNG project’s TNG50 cosmological magnetohydrodynamical simulation. ANNgelina is able to predict the SEDs of TNG50 galaxies in the ultraviolet (UV) to millimetre regime with a typical median absolute error of ∼7 per cent. The prediction error is the greatest in the UV, possibly due to the viewing-angle dependence being greatest in this wavelength regime. Our results demonstrate that our ANN-based emulator is a promising computationally inexpensive alternative for forward-modeling galaxy SEDs from cosmological simulations.