Search

Machine learning based surrogate models offer researchers powerful tools for accelerating simulation-based workflows. However, as standard datasets in this space often cover small classes of physical behavior, it can be difficult to evaluate the efficacy of new approaches. To address this gap, we introduce the Well: a large-scale collection of datasets containing numerical simulations of a wide variety of spatiotemporal physical systems. The Well draws from domain experts and numerical software developers to provide 15TB of data across 16 datasets covering diverse domains such as biological systems, fluid dynamics, acoustic scattering, as well as magneto-hydrodynamic simulations of extra-galactic fluids or supernova explosions. These datasets can be used individually or as part of a broader benchmark suite. To facilitate usage of the Well, we provide a unified PyTorch interface for training and evaluating models. We demonstrate the function of this library by introducing example baselines that highlight the new challenges posed by the complex dynamics of the Well. The code and data is available at https://github.com/PolymathicAI/the_well., Comment: 38th Conference on Neural Information Processing Systems (NeurIPS 2024) Track on Datasets and Benchmarks

Dye attenuation, or photometric imaging, is an optical technique commonly used in fluid dynamics to measure tracer concentration fields and fluid thicknesses under the assumption that the motion of the dye is representative of the fluid motion and that its presence does not affect the behaviour of the system. However, in some systems, particularly living biological systems or those with strong chemical interactions and reactions, the addition of dye may non-trivially influence the system and may not follow the fluid containing it. To overcome this, we demonstrate how short-wave infrared imaging can be used to measure concentration and height profiles of water and other liquids without the introduction of dye for heights down to 0.2mm with spatial and temporal resolutions of the order of 50 microns per pixel and 120 fps respectively. We showcase the utility of this technique by demonstrating its ability to accurately track the temporal evolution of the total water content of two model systems, namely a water drop spreading on a glass slide and spreading within a hydrogel sheet, validating both against an analytical mass balance. Finally, we discuss how the spectral resolution of the present setup could be increased to the point that concentrations within a multi-component system containing more than one type of liquid could be quantified.

From microactuators to biological tissues, non-porous materials with the ability to strongly expand when in contact with a solvent are ubiquitous. Consequently, the swelling of polymer systems such as hydrogel has received recently much research attention. However, the related dehydration of these systems has received much less attention. Here, we present experiments investigating the rapid dehydration of a swollen hydrogel sheet whose surface exhibits a transient crumpling instability characterised by line segments of cusps patterning the surface of the gel into an array of bumps. We perform this dehydration through immersion in acetone, which is highly miscible in water, but poorly miscible in the hydrogel. We report the onset of a fascinating pattern formation where regions of the hydrogel sheet turn turbid. We find that the emerging pattern is independent of the overall extent of the hydrated swollen surface. The pattern wavelength only depends on the duration of hydration before immersion in acetone, growing temporally with a power law behaviour. We conclude through drawing comparisons between features of this dehydration induced pattern and the original crumpling instability in the water swollen hydrogel sheet.

We present $[OIII]/H_{\rm \beta}$ emision line flux ratio predictions for galaxies at $z \sim 7-9$ using the MAPPINGS V v5.2.0 photoionization modelling code combined with an analytic galaxy formation model. Properties such as pressure and ionization parameter that determine emission line properties are thought to evolve towards high redshift. In order to determine the range of expected interstellar conditions we extend previous modelling of the Star Formation Rate Density (SFRD) function to calculate the metallicity and ionization parameter, and incorporate the potential impact of turbulence on the density of the ISM. To validate our emission line predictions we calculate the [OIII] line luminosity and its dependence on UV luminosity, as well as the flux ratio $[OIII]/H_{\rm \beta}$ and its variation with the line luminosity, finding that both reproduce recent JWST observations from the FRESCO survey. We also use our model to predict the number counts of emission line galaxies across a range of redshift as well as the dependence of $[OIII]/H_{\rm \beta}$ on ionization parameter and metallicity. Finally, we show that the dependence of flux ratio on luminosity may provide a diagnostic of turbulent motion in galactic discs., Comment: 12 pages, 17 figures, Submitted to MNRAS

The redshift and size distributions of galaxy scale strong lenses depend on the evolution of early-type galaxies (ETGs) in the redshift range 0.2

We investigate the detectability of Lyman-$\alpha$ (Ly$\alpha$) emission from galaxies at the onset of cosmic reionization, aiming to understand the conditions necessary for detecting high-redshift sources like JADES-GS-z13-1-LA at $z=13$. By integrating galaxy formation models with detailed intergalactic medium (IGM) reionization simulations, we construct high-redshift galaxy catalogs to model intrinsic Ly$\alpha$ profiles and assess their transmission through the IGM. For a galaxy with $M_{\rm UV}\sim -18.5$ like JADES-GS-z13-1-LA, our fiducial model predicts a Ly$\alpha$ transmission of ${\sim}13$% and there is a probability of observing Ly$\alpha$ emission with an equivalent width >40A of up to 10%. We also explore how variations in the UV ionizing escape fraction, dependent on host halo mass, impact Ly$\alpha$ detectability. Our findings reveal that reionization morphology significantly influences detection chances -- models where reionization is driven by low-mass galaxies can boost the detection probability to as much as 12%, while those driven by massive galaxies tend to reduce ionized regions around faint emitters, limiting their detectability. This study underscores the importance of reionization morphology in interpreting high-redshift Ly$\alpha$ observations., Comment: 7 pages, 3 figures, a theory follow up on 2408.16608, comments welcome

We report evidence of galaxy assembly bias - the correlation between galaxy properties and biased secondary halo properties at fixed halo mass (M$_H$) - in the stellar-to-halo mass relation (SHMR) for red central galaxies from the Sloan Digital Sky Survey. In the M$_H = 10^{11.5}-10^{13.5} h^{-1}$ M$_{\odot}$ range, central galaxy stellar mass (M$_*$) is correlated with the number density of galaxies within $10 h^{-1}$ Mpc ($\delta_{10}$), a common proxy for halo formation time. This galaxy assembly bias signal is also present when M$_H$, M$_*$, and $\delta_{10}$ are substituted with group luminosity, galaxy luminosity, and metrics of the large-scale density field. To associate differences in $\delta_{10}$ with variations in halo formation time, we fitted a model that accounts for (1) errors in the M$_H$ measured by the Tinker 2021, 2022 group catalog and (2) the level of correlation between halo formation time and M$_*$ at fixed M$_H$. Fitting of this model yields that (1) errors in M$_H$ are 0.15 dex and (2) halo formation time and M$_*$ are strongly correlated (Spearman's rank correlation coefficient ~0.85). At fixed M$_H$, variations of ~0.4 dex in M$_*$ are associated with ~1-3 Gyr variations in halo formation time and in galaxy formation time (from stellar population fitting; Oyarz\'un et al. 2022). These results are indicative that halo properties other than M$_H$ can impact central galaxy assembly., Comment: Accepted for publication in ApJ. 11 pages and 6 figures

Many lung diseases require detailed visualisation for accurate diagnosis and treatment. High-resolution computed tomography (CT) is the gold-standard technique for non-invasive lung disease detection, but it presents a risk to the patient through the relatively high ionising radiation dose required. Utilising the X-ray phase information may allow improvements in image resolution at equal or lower radiation levels than current clinical imaging. Propagation-based phase-contrast imaging requires minimal adaption of existing medical systems, and is well suited to lung imaging due to the strong phase gradients introduced by the lung-air material interfaces. Herein, propagation-based phase contrast CT is demonstrated for large animals, namely lambs, as a model for paediatric patients, using monochromatic radiation and a photon-counting detector at the Imaging and Medical Beamline of the Australian Synchrotron. Image quality, normalised against radiation dose, was optimised as a function of the beam energy and propagation distance, with the optimal conditions used to test the available image quality at very low radiation dose. The resulting CT images demonstrate superior resolution to existing high-resolution CT systems, pushing dose to the quantum limit to comply with current Australian guidelines for infant chest CT exposure of $<2.5\:\text{mSv}$ effective dose. Constituent raw projections are shown to have significant proportions of pixels with zero photon counts that would create severe information loss in conventional CT. Phase retrieval enabled clear visualisation of minor lung airways at doses up to 1,225$\pm$31\% times lower than conventional CT reconstruction, at a voxel size of just 75$\mathrm{\mu}$m.

The hosts of two low-luminosity high-z quasars, J2255+0251 and J2236+0032, were recently detected using JWST's NIRCam instrument. These represent the first high-z quasar host galaxy stellar detections and open a new window into studying high-z quasars. We examine the implications of the measured properties of J2255+0251 and J2236+0032 within the context of the hydrodynamic simulation BlueTides at z = 6.5. We find that these observed quasars fall on the BlueTides stellar to black hole mass relation and have similar luminosities to the brightest simulated quasars. We predict their star formation rates, estimating approximately $10^{2-3}$ $M_{\odot}/ \rm yr$ for both quasar hosts. J2255+0251 and J2236+0032's host galaxy radii also fall within estimates of the radii of the simulated host galaxies of similar luminosity quasars. We generate mock JWST NIRCam images of analogs to the observed quasars within BlueTides and perform a point source removal to illustrate both a qualitative and quantitative comparison of the measured and simulated radii and magnitudes. The quasar subtraction works well for similar luminosity quasars, and the recovered host images are consistent with what was observed for J2255+0251 and J2236+0032, further supporting the success of those observations. We also use our mock imaging pipeline to make predictions for the detection of J2255+0251 and J2236+0032's hosts in upcoming JWST observations. We anticipate that the simulation analogs of future high-z quasar host discoveries will allow us to make accurate predictions of their properties beyond the capabilities of JWST., Comment: Accepted to MNRAS. 15 pages, 11 figures

We implemented Population III (Pop. III) star formation in mini-halos within the MERAXES semi-analytic galaxy formation and reionisation model, run on top of a N-body simulation with $L = 10 h^{-1}$ cMpc with 2048$^3$ particles resolving all dark matter halos down to the mini-halos ($\sim 10^5 M_\odot$). Our modelling includes the chemical evolution of the IGM, with metals released through supernova-driven bubbles that expand according to the Sedov-Taylor model. We found that SN-driven metal bubbles are generally small, with radii typically of 150 ckpc at z = 6. Hence, the majority of the first galaxies are likely enriched by their own star formation. However, as reionization progresses, the feedback effects from the UV background become more pronounced, leading to a halt in star formation in low-mass galaxies, after which external chemical enrichment becomes more relevant. We explore the sensitivity of the star formation rate density and stellar mass functions on the unknown values of free parameters. We also discuss the observability of Pop. III dominated systems with JWST, finding that the inclusion of Pop. III galaxies can have a significant effect on the total UV luminosity function at z = 12 - 16. Our results support the idea that the excess of bright galaxies detected with JWST might be explained by the presence of bright top-heavy Pop. III dominated galaxies without requiring an increased star formation efficiency., Comment: 19 Pages, 22 Figures, 2 tables, 2 Appendices - Accepted for publication in MNRAS (14th of Feb. 2024) - Meraxes is now publicly available!

Accuracy in the topology and statistics of a simulated Epoch of Reionization (EoR) are vital to draw connections between observations and physical processes. While full radiative transfer models produce the most accurate reionization models, they are highly computationally expensive, and are infeasible for the largest cosmological simulations. Instead, large simulations often include EoR models that are pre-computed via the initial density field, or post-processed where feedback effects are ignored. We introduce Astrid-ES, a resimulation of the Astrid epoch of reionisation $20 > z > 5.5$ which includes an on-the-fly excursion-set reionization algorithm. Astrid-ES produces more accurate reionization histories without significantly impacting the computational time. This model directly utilises the star particles produced in the simulation to calculate the EoR history and includes a UV background which heats the gas particles after their reionization. We contrast the reionization topology and statistics in Astrid-ES with the previously employed parametric reionisation model, finding that in Astrid-ES, ionised regions are more correlated with galaxies, and the 21cm power-spectrum shows an increase in large scale power. We calculate the relation between the size of HII regions and the UV luminosity of the brightest galaxy within them. Prior to the overlap phase, we find a power-law fit of $\mathrm{log} (R) = -0.314 M_\mathrm{UV} - 2.550 \mathrm{log}(1+z) + 7.408$ with a standard deviation $\sigma_R < 0.15 \mathrm{dex}$ across all mass bins. We also examine the properties of halos throughout reionization, finding that while the properties of halos in the simulation are correlated with the redshift of reionisation, they are not greatly affected by reionisation itself., Comment: 12 pages, 11 figures. Accepted for publication in Monthly Notices of the Royal Astronomical Society

Using a semi-analytic galaxy-formation model, we study analogues of 8 recently discovered JWST galaxies at $z>{\sim}12$. We select analogues from a cosmological simulation with a $(311{\rm cMpc})^3$ volume and an effective particle number of $10^{12}$ enabling resolution of every atomic-cooling galaxy at $z{\le}20$. We vary model parameters to reproduce the observed UV luminosity function at $5{<}z{<}13$, aiming for a statistically representative high-redshift galaxy mock catalogue. Using the forward-modelled JWST photometry, we identify analogues from this catalogue and study their properties as well as possible evolutionary paths and local environments. We find faint JWST galaxies ($M_{\rm UV}>{\sim}-19.5$) to remain consistent with standard galaxy-formation model and that our fiducial catalogue includes large samples of their analogues. The properties of these analogues broadly agree with conventional SED fitting results, except for having systematically lower redshifts due to the evolving UV luminosity function, and for having higher specific star formation rates as a result of burstier histories in our model. On the other hand, only a handful of bright galaxy analogues can be identified for the observed $z{\sim}12$ galaxies. Moreover, in order to reproduce the $z>{\sim}16$ JWST galaxy candidates, boosted star-forming efficiencies and reduced feedback regulation are necessary relative to models of lower-redshift populations. This suggests star formation in the first galaxies could differ significantly from their lower-redshift counterparts. We also find that these candidates are subject to low-redshift contamination, which is present in our fiducial results as both the dusty or quiescent galaxies at $z{\sim}5$., Comment: updated with a couple of references and some typos fixed. 18 pages, 11 figures, submitted to MNRAS, comments welcomed

The hyperfine 21-cm transition of neutral hydrogen from the early Universe ($z>5$) is a sensitive probe of the formation and evolution of the first luminous sources. Using the Fisher matrix formalism we explore the complex and degenerate high-dimensional parameter space associated with the high-$z$ sources of this era and forecast quantitative constraints from a future 21-cm power spectrum (21-cm PS) detection. This is achieved using MERAXES, a coupled semi-analytic galaxy formation model and reionisation simulation, applied to an $N$-body halo merger tree with a statistically complete population of all atomically cooled galaxies out to $z\sim20$. Our mock observation assumes a 21-cm detection spanning $z \in [5, 24]$ from a 1000 h mock observation with the forthcoming Square Kilometre Array and is calibrated with respect to ultraviolet luminosity functions (UV LFs) at $z\in[5, 10]$, the optical depth of CMB photons to Thompson scattering from Planck, and various constraints on the IGM neutral fraction at $z > 5$. In this work, we focus on the X-ray luminosity, ionising UV photon escape fraction, star formation and supernova feedback of the first galaxies. We demonstrate that it is possible to recover 5 of the 8 parameters describing these properties with better than $50$ per cent precision using just the 21-cm PS. By combining with UV LFs, we are able to improve our forecast, with 5 of the 8 parameters constrained to better than 10 per cent (and all below 50 per cent)., Comment: 14 pages, 5 figures (+3 appendix), Revised and MNRAS accepted version

The fraction of ionizing photons that escape their host galaxies to ionize hydrogen in the inter-galactic medium (IGM) is a critical parameter in analyses of the reionization era. In this paper we use the Meraxes semi-analytic galaxy formation model to infer the mean ionizing photon escape fraction and its dependence on galaxy properties through joint modelling of the observed high redshift galaxy population and existing constraints on the reionization history. Using a Bayesian framework, and under the assumption that escape fraction is primarily related to halo mass, we find that the joint constraints of the UV luminosity function, CMB optical depth, and the Ly$\alpha$ forest require an escape fraction of $(18\pm5)\%$ for galaxies within haloes of $M\lesssim10^{9}$M$_\odot$ and $(5\pm2)\%$ for more massive haloes. In terms of galaxy properties, this transition in escape fraction occurs at stellar masses of $M_\star\sim10^7$M$_\odot$, nearly independent of redshift. As a function of redshift, reionization is dominated by the smaller $M_\star\lesssim10^7$M$_\odot$ galaxies with high escape fractions at $z\gtrsim6$ and by the larger $M_\star\gtrsim10^7$M$_\odot$ galaxies with lower escape fractions at $z\lesssim6$. Galaxies with star formation rates of $10^{-2.5}$M$_\odot$yr$^{-1}$ to $10^{-1.5}$M$_\odot$yr$^{-1}$ provide the dominant source of ionizing photons throughout reionization. Our results are consistent with recent direct measurements of a $\sim5\%$ escape fraction from massive galaxies at the end of reionization and support the picture of low mass galaxies being the dominant sources of ionizing photons during reionization., Comment: Updated to match accepted version for publication in MNRAS. Includes two new figures showing model predictions and related discussion

The stratified inclined duct (SID) sustains an exchange flow in a long, gently sloping duct as a model for continuously-forced density-stratified flows such as those found in estuaries. Experiments have shown that the emergence of interfacial waves and their transition to turbulence as the tilt angle is increased appears linked to a threshold in the exchange flow rate given by inviscid two-layer hydraulics. We uncover these hydraulic mechanisms with (i) recent direct numerical simulations (DNS) providing full flow data in the key flow regimes (Zhu & Atoufi et al., arXiv:2301.09773, 2023), (ii) averaging these DNS into two layers, (iii) an inviscid two-layer shallow water and instability theory to diagnose interfacial wave behaviour and provide physical insight. The laminar flow is subcritical and stable throughout the duct and hydraulically controlled at the ends of the duct. As the tilt is increased, the flow becomes everywhere supercritical and unstable to long waves. An internal undular jump featuring stationary waves first appears near the centre of the duct, then leads to larger-amplitude travelling waves, and to stronger jumps, wave breaking and intermittent turbulence at the largest tilt angle. Long waves described by the (nonlinear) shallow water equation are locally interpreted as linear waves on a two-layer parallel base flow described by the Taylor-Goldstein equation. This link helps us interpret long-wave instability and contrast it to short-wave (e.g. Kelvin-Helmholtz) instability. Our results suggest a transition to turbulence in SID through long-wave instability relying on vertical confinement by the top and bottom walls.

The stratified inclined duct (SID) experiment consists of a zero-net-volume exchange flow in a long tilted rectangular duct, which allows the study of realistic stratified shear flows with sustained internal forcing. We present the first three-dimensional direct numerical simulations (DNS) of SID to explore the transitions between increasingly turbulent flow regimes first described by Meyer \& Linden (\textit{J. Fluid Mech.} \textbf{753}, 242-253, 2014). We develop a numerical set-up that faithfully reproduces the experiments and sustains the flow for arbitrarily long times at minimal computational cost. We recover the four qualitative flow regimes found experimentally in the same regions of parameter space: laminar flow, waves, intermittent turbulence, and fully-developed turbulence. We find good qualitative and quantitative agreement between DNS and experiments and highlight the added value of DNS to complement experimental diagnostics and increase our understanding of the transition to turbulence, both temporally (laminar/turbulent cycles) and parametrically (as the tilt angle of the duct and the Reynolds number are increased). These results demonstrate that numerical studies of SID -- and deeper integration between simulations and experiments -- have the potential to lead to a better understanding of stratified turbulence in environmental flows.

Multimodal characterization is commonly required for understanding materials. User facilities possess the infrastructure to perform these measurements, albeit in serial over days to months. In this paper, we describe a unified multimodal measurement of a single sample library at distant instruments, driven by a concert of distributed agents that use analysis from each modality to inform the direction of the other in real time. Powered by the Bluesky project at the National Synchrotron Light Source II, this experiment is a world's first for beamline science, and provides a blueprint for future approaches to multimodal and multifidelity experiments at user facilities., Comment: 36th Conference on Neural Information Processing Systems (NeurIPS 2022). AI4Mat Workshop

The present work focuses on a specific bouncing behaviour as a particle settling through a three-layer stratified fluid in the absence of neutral buoyant position, which was firstly discovered by Abaid, N., Adalsteinsson D., Agyapong A. & McLaughlin, R.M. (2004) in salinity-induced stratification. Both experiments and numerical simulations are carried out. In our experiments, illuminated by a laser sheet on the central plane of the particle, its bouncing behaviour is well captured. We find that the bouncing process starts after the wake detaches from the particle. The PIV results show that an upward jet is generated at the central axis behind the particle after the wake breaks. By conducting a force decomposition procedure, we quantify the enhanced drag caused by the buoyancy of the wake ($F_{sb}$) and the flow structure ($F_{sj}$). It is noted that $F_{sb}$ contributes primarily to the enhanced drag at the early stage, which becomes less dominant after the detachment of the wake. In contrast, $F_{sj}$ plays a pivotal role in reversing the particle's motion. We conjecture that the jet flow is a necessary condition for the occurrence of bouncing motion. Then, we examine the minimal velocities (negative values when bounce occurs) of the particle by varying the lower Reynolds number $Re_l$, the Froude number $Fr$ and the upper Reynolds number $Re_u$ within the ranges $1 \leq Re_l\leq 125$, $115 \leq Re_u\leq 356$ and $2 \leq Fr\leq 7$. We find that the bouncing behaviour is primarily determined by $Re_l$. In our experiments, the bouncing motion is found to occur below a critical lower Reynolds number around $Re^ \ast _{l}=30$. In the numerical simulations, the highest value for this critical number is $Re^ \ast _{l}=46.2$, limited in the currently studied parametric ranges.

The bright end of the rest-frame UV luminosity function (UVLF) of high-redshift galaxies is modified by gravitational lensing magnification bias. Motivated by recent discoveries of very high-z galaxies with JWST, we study the dependence of magnification bias on the finite size of sources at $6

Modeling the 21cm global signal from the Cosmic Dawn is challenging due to the many poorly constrained physical processes that come into play. We address this problem using the semi-analytical code "Cosmic Archaeology Tool" (CAT). CAT follows the evolution of dark matter halos tracking their merger history and provides an ab initio description of their baryonic evolution, starting from the formation of the first (Pop III) stars and black holes (BHs) in mini-halos at z > 20. The model is anchored to observations of galaxies and AGN at z < 6 and predicts a reionization history consistent with constraints. In this work we compute the evolution of the mean global 21cm signal between $4\leq z \leq 40$ based on the rate of formation and emission properties of stars and accreting black holes. We obtain an absorption profile with a maximum depth $\delta {\rm T_b} = -95$ mK at $z \sim 26.5$ (54 MHz). This feature is quickly suppressed turning into an emission signal at $z = 20$ due to the contribution of accreting BHs that efficiently heat the IGM at $z < 27$. The high-$z$ absorption feature is caused by the early coupling between the spin and kinetic temperature of the IGM induced by Pop III star formation episodes in mini-halos. Once we account for an additional radio background from early BHs, we are able to reproduce the timing and the depth of the EDGES signal only if we consider a smaller X-ray background from accreting BHs, but not the shape., Comment: 16 pages, 10 figures, Accepted for publication in MNRAS

We predict the 21-cm global signal and power spectra during the Epoch of Reionisation using the MERAXES semi-analytic galaxy formation and reionisation model, updated to include X-ray heating and thermal evolution of the intergalactic medium. Studying the formation and evolution of galaxies together with the reionisation of cosmic hydrogen using semi-analytic models (such as MERAXES) requires N-body simulations within large volumes and high mass resolutions. For this, we use a simulation of side-length $210~h^{-1}$ Mpc with $4320^3$ particles resolving dark matter haloes to masses of $5\times10^8~h^{-1}~M_\odot$. To reach the mass resolution of atomically cooled galaxies, thought to be the dominant population contributing to reionisation, at $z=20$ of $\sim 2\times10^7~h^{-1}~M_\odot$, we augment this simulation using the DARKFOREST Monte-Carlo merger tree algorithm (achieving an effective particle count of $\sim10^{12}$). Using this augmented simulation we explore the impact of mass resolution on the predicted reionisation history as well as the impact of X-ray heating on the 21-cm global signal and the 21-cm power spectra. We also explore the cosmic variance of 21-cm statistics within $70^{3}$ $h^{-3}$ Mpc$^3$ sub-volumes. We find that the midpoint of reionisation varies by $\Delta z\sim0.8$ and that the cosmic variance on the power spectrum is underestimated by a factor of $2-4$ at $k\sim 0.1-0.4$ Mpc$^{-1}$ due to the non-Gaussian nature of the 21-cm signal. To our knowledge, this work represents the first model of both reionisation and galaxy formation which resolves low-mass atomically cooled galaxies while simultaneously sampling sufficiently large scales necessary for exploring the effects of X-rays in the early Universe., Comment: 17 pages, 13 figures, revised and MNRAS accepted version