Search for: All records

Star formation has been observed to occur at globally low yet locally varying efficiencies. As such, accurate capture of star formation in numerical simulations requires mechanisms that can replicate both its smaller scale variations and larger scale properties. Magnetic fields are thought to play an essential role within the turbulent interstellar medium (ISM) and affect molecular cloud collapse. However, it remains to be fully explored how a magnetized model of star formation might influence galaxy evolution. We present a new model for a sub-grid star formation recipe that depends on the magnetic field. We run isolated disc galaxy simulations to assess its impact on the regulation of star formation using the code ramses. Building upon existing numerical methods, our model derives the star formation efficiency from local properties of the sub-grid magnetized ISM turbulence, assuming a constant Alfvén speed at sub-parsec scales. Compared to its non-magnetized counterpart, our star formation model suppresses the initial starburst by a factor of 2 while regulating star formation later on to a nearly constant rate of ∼1 M⊙ yr−1. Differences also arise in the local Schmidt law with a shallower power-law index for the magnetized star formation model. Our results encourage further examination into the notion that magnetic fields are likely to play a non-trivial role in our understanding of star and galaxy formation.

 

Abstract Recent work has pointed out the potential existence of a tight relation between the cosmological parameter Ω m , at fixed Ω b , and the properties of individual galaxies in state-of-the-art cosmological hydrodynamic simulations. In this paper, we investigate whether such a relation also holds for galaxies from simulations run with a different code that makes use of a distinct subgrid physics: Astrid. We also find that in this case, neural networks are able to infer the value of Ω m with a ∼10% precision from the properties of individual galaxies, while accounting for astrophysics uncertainties, as modeled in Cosmology and Astrophysics with MachinE Learning (CAMELS). This tight relationship is present at all considered redshifts, z ≤ 3, and the stellar mass, the stellar metallicity, and the maximum circular velocity are among the most important galaxy properties behind the relation. In order to use this method with real galaxies, one needs to quantify its robustness: the accuracy of the model when tested on galaxies generated by codes different from the one used for training. We quantify the robustness of the models by testing them on galaxies from four different codes: IllustrisTNG, SIMBA, Astrid, and Magneticum. We show that the models perform well on a large fraction of the galaxies, but fail dramatically on a small fraction of them. Removing these outliers significantly improves the accuracy of the models across simulation codes. 

The formation and evolution of galaxies have proved sensitive to the inclusion of stellar feedback, which is therefore crucial to any successful galaxy model. We present INFERNO, a new model for hydrodynamic simulations of galaxies, which incorporates resolved stellar objects with star-by-star calculations of when and where the injection of enriched material, momentum, and energy takes place. INFERNO treats early stellar kinematics to include phenomena such as walkaway and runaway stars. We employ this innovative model on simulations of a dwarf galaxy and demonstrate that our physically motivated stellar feedback model can drive vigorous galactic winds. This is quantified by mass and metal loading factors in the range of 10–100, and an energy loading factor close to unity. Outflows are established close to the disc, are highly multiphase, spanning almost 8 orders of magnitude in temperature, and with a clear dichotomy between mass ejected in cold, slow-moving (T ≲ 5 × 104 K, v < 100 km s−1) gas and energy ejected in hot, fast-moving (T > 106 K, v > 100 km s−1) gas. In contrast to massive disc galaxies, we find a surprisingly weak impact of the early stellar kinematics, with runaway stars having little to no effect on our results, despite exploding in diffuse gas outside the dense star-forming gas, as well as outside the galactic disc entirely. We demonstrate that this weak impact in dwarf galaxies stems from a combination of strong feedback and a porous interstellar medium, which obscure any unique signatures that runaway stars provide.

 

Abstract We present a suite of high-resolution simulations of an isolated dwarf galaxy using four different hydrodynamical codes: Gizmo , Arepo , Gadget , and Ramses . All codes adopt the same physical model, which includes radiative cooling, photoelectric heating, star formation, and supernova (SN) feedback. Individual SN explosions are directly resolved without resorting to subgrid models, eliminating one of the major uncertainties in cosmological simulations. We find reasonable agreement on the time-averaged star formation rates as well as the joint density–temperature distributions between all codes. However, the Lagrangian codes show significantly burstier star formation, larger SN-driven bubbles, and stronger galactic outflows compared to the Eulerian code. This is caused by the behavior in the dense, collapsing gas clouds when the Jeans length becomes unresolved: Gas in Lagrangian codes collapses to much higher densities than that in Eulerian codes, as the latter is stabilized by the minimal cell size. Therefore, more of the gas cloud is converted to stars and SNe are much more clustered in the Lagrangian models, amplifying their dynamical impact. The differences between Lagrangian and Eulerian codes can be reduced by adopting a higher star formation efficiency in Eulerian codes, which significantly enhances SN clustering in the latter. Adopting a zero SN delay time reduces burstiness in all codes, resulting in vanishing outflows as SN clustering is suppressed. 

When the first galaxies formed and starlight escaped into the intergalactic medium to reionize it, galaxy formation and reionization were both highly inhomogeneous in time and space, and fully coupled by mutual feedback. To show how this imprinted the UV luminosity function (UVLF) of reionization-era galaxies, we use our large-scale, radiation-hydrodynamics simulation CoDa II to derive the time- and space-varying halo mass function and UVLF, from z ≃ 6–15. That UVLF correlates strongly with local reionization redshift: earlier-reionizing regions have UVLFs that are higher, more extended to brighter magnitudes, and flatter at the faint end than later-reionizing regions observed at the same z. In general, as a region reionizes, the faint-end slope of its local UVLF flattens, and, by z = 6 (when reionization ended), the global UVLF, too, exhibits a flattened faint-end slope, ‘rolling-over’ at MUV ≳ −17. CoDa II’s UVLF is broadly consistent with cluster-lensed galaxy observations of the Hubble Frontier Fields at z = 6–8, including the faint end, except for the faintest data point at z = 6, based on one galaxy at MUV = −12.5. According to CoDa II, the probability of observing the latter is $\sim 5~{{\ \rm per\ cent}}$. However, the effective volume searched at this magnitude is very small, and is thus subject to significant cosmic variance. We find that previous methods adopted to calculate the uncertainty due to cosmic variance underestimated it on such small scales by a factor of 2–4, primarily by underestimating the variance in halo abundance when the sample volume is small.

 

Abstract We train graph neural networks to perform field-level likelihood-free inference using galaxy catalogs from state-of-the-art hydrodynamic simulations of the CAMELS project. Our models are rotational, translational, and permutation invariant and do not impose any cut on scale. From galaxy catalogs that only contain 3D positions and radial velocities of ∼1000 galaxies in tiny ( 25 h − 1 Mpc ) 3 volumes our models can infer the value of Ω m with approximately 12% precision. More importantly, by testing the models on galaxy catalogs from thousands of hydrodynamic simulations, each having a different efficiency of supernova and active galactic nucleus feedback, run with five different codes and subgrid models—IllustrisTNG, SIMBA, Astrid, Magneticum, SWIFT-EAGLE—we find that our models are robust to changes in astrophysics, subgrid physics, and subhalo/galaxy finder. Furthermore, we test our models on 1024 simulations that cover a vast region in parameter space—variations in five cosmological and 23 astrophysical parameters—finding that the model extrapolates really well. Our results indicate that the key to building a robust model is the use of both galaxy positions and velocities, suggesting that the network has likely learned an underlying physical relation that does not depend on galaxy formation and is valid on scales larger than ∼10 h −1 kpc. 

We analyze the circumgalactic medium (CGM) for eight commonly-used cosmological codes in the AGORA collaboration. The codes are calibrated to use identical initial conditions, cosmology, heating and cooling, and star formation thresholds, but each evolves with its own unique code architecture and stellar feedback implementation. Here, we analyze the results of these simulations in terms of the structure, composition, and phase dynamics of the CGM. We show properties such as metal distribution, ionization levels, and kinematics are effective tracers of the effects of the different code feedback and implementation methods, and as such they can be highly divergent between simulations. This is merely a fiducial set of models, against which we will in the future compare multiple feedback recipes for each code. Nevertheless, we find that the large parameter space these simulations establish can help disentangle the different variables that affect observable quantities in the CGM, e.g., showing that abundances for ions with higher ionization energy are more strongly determined by the simulation’s metallicity, while abundances for ions with lower ionization energy are more strongly determined by the gas density and temperature.

 

Abstract We train graph neural networks on halo catalogs from Gadget N -body simulations to perform field-level likelihood-free inference of cosmological parameters. The catalogs contain ≲5000 halos with masses ≳10 10 h −1 M ⊙ in a periodic volume of ( 25 h − 1 Mpc ) 3 ; every halo in the catalog is characterized by several properties such as position, mass, velocity, concentration, and maximum circular velocity. Our models, built to be permutationally, translationally, and rotationally invariant, do not impose a minimum scale on which to extract information and are able to infer the values of Ω m and σ 8 with a mean relative error of ∼6%, when using positions plus velocities and positions plus masses, respectively. More importantly, we find that our models are very robust: they can infer the value of Ω m and σ 8 when tested using halo catalogs from thousands of N -body simulations run with five different N -body codes: Abacus, CUBEP 3 M, Enzo, PKDGrav3, and Ramses. Surprisingly, the model trained to infer Ω m also works when tested on thousands of state-of-the-art CAMELS hydrodynamic simulations run with four different codes and subgrid physics implementations. Using halo properties such as concentration and maximum circular velocity allow our models to extract more information, at the expense of breaking the robustness of the models. This may happen because the different N -body codes are not converged on the relevant scales corresponding to these parameters. 

Abstract Galaxies can be characterized by many internal properties such as stellar mass, gas metallicity, and star formation rate. We quantify the amount of cosmological and astrophysical information that the internal properties of individual galaxies and their host dark matter halos contain. We train neural networks using hundreds of thousands of galaxies from 2000 state-of-the-art hydrodynamic simulations with different cosmologies and astrophysical models of the CAMELS project to perform likelihood-free inference on the value of the cosmological and astrophysical parameters. We find that knowing the internal properties of a single galaxy allows our models to infer the value of Ω m , at fixed Ω b , with a ∼10% precision, while no constraint can be placed on σ 8 . Our results hold for any type of galaxy, central or satellite, massive or dwarf, at all considered redshifts, z ≤ 3, and they incorporate uncertainties in astrophysics as modeled in CAMELS. However, our models are not robust to changes in subgrid physics due to the large intrinsic differences the two considered models imprint on galaxy properties. We find that the stellar mass, stellar metallicity, and maximum circular velocity are among the most important galaxy properties to determine the value of Ω m . We believe that our results can be explained by considering that changes in the value of Ω m , or potentially Ω b /Ω m , affect the dark matter content of galaxies, which leaves a signature in galaxy properties distinct from the one induced by galactic processes. Our results suggest that the low-dimensional manifold hosting galaxy properties provides a tight direct link between cosmology and astrophysics. 

ABSTRACT We present a novel set of stellar feedback models, implemented in the moving-mesh code arepo, designed for galaxy formation simulations with near-parsec (or better) resolution. These include explicit sampling of stars from the IMF, allowing feedback to be linked to individual massive stars, an improved method for the modelling of H ii regions, photoelectric (PE) heating from a spatially varying FUV field and supernova feedback. We perform a suite of 32 simulations of isolated $M_\mathrm{vir} = 10^{10}\, \mathrm{M_\odot }$ galaxies with a baryonic mass resolution of $20\, \mathrm{M_\odot }$ in order to study the non-linear coupling of the different feedback channels. We find that photoionization (PI) and supernova feedback are both independently capable of regulating star formation to the same level, while PE heating is inefficient. PI produces a considerably smoother star formation history than supernovae. When all feedback channels are combined, the additional suppression of star formation rates is minor. However, outflow rates are substantially reduced relative to the supernova only simulations. We show that this is directly caused by a suppression of supernova clustering by the PI feedback, disrupting star-forming clouds prior to the first supernovae. We demonstrate that our results are robust to variations of our star formation prescription, feedback models and the baryon fraction of the galaxy. Our results also imply that the burstiness of star formation and the mass loading of outflows may be overestimated if the adopted star particle mass is considerably larger than the mass of individual stars because this imposes a minimum cluster size.