arXiv subjects

Sadegh Khochfar

Publications and source records attributed to Sadegh Khochfar.

At least 37 records · Page 2Linked to original sources

arXiv2022

Linking the Internal Properties of Infant Globular Clusters to their Formation Environments

We investigate the formation of infant globular cluster (GC) candidates in high-resolution cosmological simulations from the First Billion Years (FiBY) project. By analysing the evolution of the systems in the energy and angular momentum plane, we identify the redshift at which the infant GCs first became gravitationally bound, and we find evidence of radial infall of their gaseous and stellar components. The collapse appears to be driven by internal self-gravity, however, the initial trigger is sourced from the external environment. The phase space behaviour of the infant GCs also allows us to identify some characteristic groupings of objects. Such a classification based on internal properties appears to be reflected in the formation environment: GC candidates that belong to the same class are found in host galaxies of similar morphology, with the majority of the infant GCs located in clumpy, irregular proto-galaxies. Finally, through the inspection of two GC candidates that contain only stars by z = 6, we find that supernova feedback is the main physical mechanism behind their dearth of gas and that the systems subsequently respond with an approximately adiabatic expansion. Such infant GC candidates already resemble the GCs we currently observe in the local Universe.

astro-ph.GA↗

arXiv2022

The fate of baryons in counterfactual universes

We present results from nine simulations that compare the standard $Λ$ Cold Dark Matter cosmology ($Λ$CDM) with counterfactual universes, for approximately $100\,{\rm Gyr}$ using the Enzo simulation code. We vary the value of $Λ$ and the fluctuation amplitude to explore the effect on the evolution of the halo mass function (HMF), the intergalactic medium (IGM) and the star formation history (SFH). The distinct peak in star formation rate density (SFRD) and its subsequent decline are both affected by the interplay between gravitational attraction and the accelerating effects of $Λ$. The IGM cools down more rapidly in models with a larger $Λ$ and also with a lower $σ_8$, reflecting the reduced SFRD associated with these changes -- although changing $σ_8$ is not degenerate with changing $Λ$, either regarding the thermal history of the IGM or the SFH. However, these induced changes to the IGM or ionizing background have little impact on the calculated SFRD. We provide fits for the evolution of the SFRD in these different universes, which we integrate over time to derive an asymptotic star formation efficiency. Together with Weinberg's uniform prior on $Λ$, the estimated probability of observers experiencing a value of $Λ$ no greater than the observed value is 13%, substantially larger than some alternative estimates. Within the Enzo model framework, then, observer selection within a multiverse is able to account statistically for the small value of the cosmological constant, although $Λ$ in our universe does appear to be at the low end of the predicted range.

astro-ph.CO↗

arXiv2022

Turbulent Cold Flows Gave Birth to the First Quasars

How quasars powered by supermassive black holes (SMBHs) formed less than a billion years after the Big Bang is still one of the outstanding problems in astrophysics 20 years after their discovery$^{1-4}$. Cosmological simulations suggest that rare cold flows converging on primordial haloes in low-shear environments could have created these quasars if they were 10$^4$ - 10$^5$ M$_{\odot}$ at birth but could not resolve their formation$^{5-8}$. Semianalytical studies of the progenitor halo of a primordial quasar found that it favours the formation of such seeds but could not verify if one actually appeared$^9$. Here we show that a halo at the rare convergence of strong, cold accretion flows creates massive BH seeds without the need for UV backgrounds, supersonic streaming motions, or even atomic cooling. Cold flows drive violent, supersonic turbulence in the halo that prevents star formation until it reaches a mass that triggers sudden, catastrophic baryon collapse that forms 31,000 and 40,000 M$_{\odot}$ stars. This simple, robust process ensures that haloes capable of forming quasars by z $>$ 6 produce massive seeds. The first quasars were thus a natural consequence of structure formation in cold dark matter cosmologies, not exotic, finely-tuned environments as previously thought$^{10-14}$.

astro-ph.GA↗

arXiv2022

Multi-Epoch Machine Learning 1: Unravelling Nature vs Nurture for Galaxy Formation

We present a novel machine learning method for predicting the baryonic properties of dark matter only subhalos from N-body simulations. Our model is built using the extremely randomized tree (ERT) algorithm and takes subhalo properties over a wide range of redshifts as its input features. We train our model using the IllustrisTNG simulations to predict blackhole mass, gas mass, magnitudes, star formation rate, stellar mass, and metallicity. We compare the results of our method with a baseline model from previous works, and against a model that only considers the mass history of the subhalo. We find that our new model significantly outperforms both of the other models. We then investigate the predictive power of each input by looking at feature importance scores from the ERT algorithm. We produce feature importance plots for each baryonic property, and find that they differ significantly. We identify low redshifts as being most important for predicting star formation rate and gas mass, with high redshifts being most important for predicting stellar mass and metallicity, and consider what this implies for nature vs nurture. We find that the physical properties of galaxies investigated in this study are all driven by nurture and not nature. The only property showing a somewhat stronger impact of nature is the present-day star formation rate of galaxies. Finally we verify that the feature importance plots are discovering physical patterns, and that the trends shown are not an artefact of the ERT algorithm.

astro-ph.GA↗

arXiv2022

The Cosmic Mach Number as an Environment Measure for the Underlying Dark Matter Density Field

Using cosmological dark matter only simulations of a $(1.6$ Gpc$/h)^3$ volume from the Legacy simulation project, we calculate Cosmic Mach Numbers (CMN) and perform a theoretical investigation of their relation with halo properties and features of the density field to gauge their use as an measure of the environment. CMNs calculated on individual spheres show correlations with both the overdensity in a region and the density gradient in the direction of the bulk flow around that region. To reduce the scatter around the median of these correlations, we introduce a new measure, the rank ordered Cosmic Mach number ($\hat{\mathcal{M}}_g$), which shows a tight correlations with the overdensity $δ=\frac{ρ-\barρ}{\barρ}$. Measures of the large scale density gradient as well as other average properties of the halo population in a region show tight correlations with $\hat{\mathcal{M}}_g$ as well. Our results in this first empirical study suggest that $\hat{\mathcal{M}}_g$ is an excellent proxy for the underlying density field and hence environment that can circumvent reliance on number density counts in a region. For scales between $10$ and $100 Mpc$/h, Mach numbers calculated using dark matter halos $(> 10^{12}$ M$_{\odot})$ that would typically host massive galaxies are consistent with theoretical predictions of the linear matter power spectrum at a level of $10\%$ due to non-linear effects of gravity. At redshifts $z\geq 3$, these deviations disappear. We also quantify errors due to missing large scale modes in simulations. Simulations of box size $\leq 1 $ Gpc/$h$ typically predict CMNs 10-30\% too small on scales of$\sim 100$ Mpc$/h$.

astro-ph.CO↗

arXiv2022

FOREVER22: galaxy formation in protocluster regions

We present results from a new cosmological hydrodynamics simulation campaign of protocluster (PC) regions, FOREVER22: FORmation and EVolution of galaxies in Extremely-overdense Regions motivated by SSA22. The simulations cover a wide range of cosmological scales using three different zoom set-ups in a parent volume of $(714.2~\rm cMpc)^{3}$: PCR (Proto-Cluster Region; $V= (28.6~{\rm cMpc})^{3} $, SPH particle mass, $m_{\rm{SPH}} = 4.1 \times 10^{6}~\rm M_{\odot}$ and final redshift, $z_{\rm end}=2.0$), BCG (Brightest proto-Cluster Galaxy; $V \sim (10~{\rm cMpc})^{3} $, $m_{\rm SPH} = 5.0\times10^{5}~\rm M_{\odot}$ and $z_{\rm end}=4.0$ ), and First ( $V \sim (3~{\rm cMpc})^{3} $, $m_{\rm SPH} = 7.9 \times 10^{3}~\rm M_{\odot}$ and $z_{\rm end}=9.5$) runs, that allow to focus on different aspects of galaxy formation. In the PCR runs, we follow 10 PCs, each harbouring 1 - 4 SMBHs with $M_{\rm BH} \ge 10^{9}~\rm M_{\odot}$. One of the PC cores shows a spatially close arrangement of seven starburst galaxies with ${\rm SFR} \gtrsim 100~\rm M_{\odot}~yr^{-1}$ each, that are dust-obscured and would appear as submillimeter galaxies with flux $\gtrsim 1~$ mJy at $1.1~ \rm mm$ in observations. The BCG runs show that the total SFRs of haloes hosting BCGs are affected by AGN feedback, but exceed $1000~\rm M_{\odot}~yr^{-1}$ at $z \lesssim 6$. The First runs resolve mini-haloes hosting population (Pop) III stars and we show that, in PC regions, the dominant stellar population changes from Pop III to Pop II at $z \gtrsim 20$, and the first galaxies with ${\rm SFR} \gtrsim 18~\rm M_{\odot}~yr^{-1}$ form at $z \sim 10$. These can be prime targets for future observations with the James Webb Space Telescope. Our simulations successfully reproduce the global star formation activities in observed PCs and suggest that PCs can kickstart cosmic reionization.

astro-ph.GA↗

arXiv2021

The Birth Mass Function of Pop III Stars

Population III (Pop III) stars ended the cosmic Dark Ages and began early cosmological reionization and chemical enrichment. However, in spite of their importance to the evolution of the early Universe, their properties remain uncertain because of limitations to previous numerical simulations and the lack of any observational constraints. Here we investigate Pop III star formation in five primordial halos with 3D radiation-hydrodynamical cosmological simulations. We find that multiple stars form in each minihalo and that their numbers increase over time, with up to 23 stars forming in one of the halos. Radiative feedback from the stars generates strong outflows, deforms the surrounding protostellar disk, and delays star formation for a few thousand years. Star formation rates vary with halo and depend on mass accretion onto the disk, halo spin number, and the fraction of massive stars in the halo. Stellar masses in our models range from 0.1-37 $\rm M_{\odot}$, and of the 55 stars that form in our models twelve are $\rm > 10~ M_{\odot}$ and most of the others are 1-10 $\rm M_{\odot}$. Our simulations thus suggest that Pop III stars have characteristic masses of 1-10 $\rm M_{\odot}$ and a top-heavy IMF with dN/dM $\propto M_*^{-1.18}$. Up to 70\% of the stars are ejected from their disks by three-body interactions which, along with ionizing UV feedback, limits their final masses.

astro-ph.GA↗

arXiv2021

Radiation Hydrodynamical Simulations of the Birth of Intermediate-Mass Black Holes in the First Galaxies

The leading contenders for the seeds of $z > 6$ quasars are direct-collapse black holes (DCBHs) forming in atomically-cooled halos at $z \sim$ 20. However, the Lyman-Werner (LW) UV background required to form DCBHs of 10$^5$ \Ms\ are extreme, about 10$^4$ J$_{21}$, and may have been rare in the early universe. Here we investigate the formation of intermediate-mass black holes (IMBHs) under moderate LW backgrounds of 100 and 500 J$_{21}$ that were much more common at early times. These backgrounds allow halos to grow to a few 10$^6$ - 10$^7$ \Ms\ and virial temperatures of nearly 10$^4$ K before collapsing but do not completely sterilize them of H$_2$. Gas collapse then proceeds via Ly$α$ and rapid H$_2$ cooling at rates that are 10 - 50 times those in normal Pop III star-forming haloes but less than those in purely atomically-cooled haloes. Pop III stars accreting at such rates become blue and hot, and we find that their ionizing UV radiation limits their final masses to 1800 - 2800 \Ms\, at which they later collapse to IMBHs. Moderate LW backgrounds thus produced IMBHs in far greater numbers than DCBHs in the early universe.

astro-ph.GA↗

arXiv2021

Formation of the first galaxies in the aftermath of the first supernovae

We perform high-resolution cosmological hydrodynamic simulations to study the formation of the first galaxies that reach the masses of $10^{8-9}~h^{-1}~M_\odot$ at $z=9$. The resolution of the simulations is high enough to resolve minihaloes and allow us to successfully pursue the formation of multiple Population (Pop) III stars, their supernova (SN) explosions, resultant metal-enrichment of the inter-galactic medium (IGM) in the course of the build-up of the system. Metals are ejected into the IGM by multiple Pop III SNe, but some of the metal-enriched gas falls back onto the halo after $\gtrsim 100~\rm Myr$. The star formation history of the first galaxy depends sensitively on the initial mass function (IMF) of Pop III stars. The dominant stellar population transits from Pop III to Pop II at $z\sim 12-15$ in the case of power-law Pop III IMF, ${\rm d}n/{\rm d}M \propto M^{-2.35}$ with the mass range $10-500~M_\odot$. At $z\lesssim 12$, stars are stably formed in the first galaxies with a star formation rate of $\sim 10^{-3}$-$10^{-1}~M_\odot/{\rm yr}$. In contrast, for the case with a flat IMF, gas-deprived first galaxies form due to frequent Pop III pair-instability SNe, resulting in the suppression of subsequent Pop II star formation. In addition, we calculate UV continuum, Ly$α$- and H$α$-line fluxes from the first galaxies. We show that the James Webb Space Telescope will be able to detect both UV continuum, Ly$α$ and H$α$ line emission from first galaxies with halo mass $\gtrsim 10^{9}~M_\odot$ at $z \gtrsim 10$.

astro-ph.GA↗

arXiv2021

Evolving beyond z=0: insights about the future of stars and the intergalactic medium

We present results from seven cosmological simulations that have been extended beyond the present era as far as redshift $z=-0.995$ or $t\approx96\,{\rm Gyr}$, using the Enzo simulation code. We adopt the calibrated star formation and feedback prescriptions from our previous work on reproducing the Milky Way with Enzo with modifications to the simulation code, chemistry and cooling library. We then consider the future behaviour of the halo mass function (HMF), the equation of state (EOS) of the IGM, and the cosmic star formation history (SFH). Consistent with previous work, we find a freeze-out in the HMF at $z\approx-0.6$. The evolution of the EOS of the IGM presents an interesting case study of the cosmological coincidence problem, where there is a sharp decline in the IGM temperature immediately after $z=0$. For the SFH, the simulations produce a peak and a subsequent decline into the future. However, we do find a turnaround in the SFH after $z\approx-0.98$ in some simulations, probably due to the limitations of the criteria used for star formation. By integrating the SFH in time up to $z=-0.92$, the simulation with the best spatial resolution predicts an asymptotic total stellar mass that is very close to that obtained from extrapolating the fit of the observed SFR. Lastly, we investigate the future evolution of the partition of baryons within a Milky Way-sized galaxy, using both a zoom and a box simulation. Despite vastly different resolutions, these simulations predict individual haloes containing an equal fraction of baryons in stars and gas at the time of freeze-out ($t\approx30\,{\rm Gyr}$).

astro-ph.GA↗

arXiv2021

Mapping the inhomogeneous Universe with Standard Sirens: Degeneracy between inhomogeneity and modified gravity theories

The detection of gravitational waves (GWs) and an accompanying electromagnetic (E/M) counterpart have been suggested as a future probe for cosmology and theories of gravity. In this paper, we present calculations of the luminosity distance of sources taking into account inhomogeneities in the matter distribution that are predicted in numerical simulations of structure formation. In addition, we show that inhomogeneities resulting from clustering of matter can mimic certain classes of modified gravity theories, or other effects that dampen GW amplitudes, and deviations larger than $δν\sim \mathcal{O}(0.1)\ (99\%\ \rm{C.L.})$ to the extra friction term $ν$, from zero, would be necessary to distinguish them. For these, we assume mock GWs sources, with known redshift, based on binary population synthesis models, between redshifts $z=0$ and $z=5$. We show that future GW detectors, like Einstein Telescope or Cosmic Explorer, will be needed for strong constraints on the inhomogeneity parameters and breaking the degeneracy between modified gravity effects and matter anisotropies by measuring $ν$ at $5 \%$ and $1 \%$ level with $100$ and $350$ events respectively.

astro-ph.CO↗

arXiv2020

Starbursting [O III] emitters and quiescent [C II] emitters in the reionization era

Recent observations have successfully detected [O III] $88.3\,{\rm μm}$ and [C II] $157.6\,{\rm μm}$ lines from galaxies in the early Universe with the Atacama Large Millimeter Array (ALMA). Combining cosmological hydrodynamic simulations and radiative transfer calculations, we present relations between the metal line emission and galaxy evolution at $z=6-15$. We find that galaxies during their starburst phases have high [O III] luminosity of $\sim 10^{42}~\rm erg~s^{-1}$. Once supernova feedback quenches star formation, [O III] luminosities rapidly decrease and continue to be zero for $\sim 100\,{\rm Myr}$. The slope of the relation between $\log{(\rm SFR/M_{\odot}~yr^{-1})}$ and $\log{(L_{\rm [O_{III}]}/{\rm L_{\odot}})}$ at $z=6-9$ is 1.03, and 1.43 for $\log{(L_{\rm [C_{II}]}/{\rm L_{\odot}})}$. As gas metallicity increases from sub-solar to solar metallicity by metal enrichment from star formation and feedback, the line luminosity ratio $L_{\rm [O_{III}]} / L_{\rm [C_{II}]}$ decreases from $\sim 10$ to $\sim 1$ because the O/C abundance ratio decreases due to carbon-rich winds from AGB stars and the mass ratio of {\sc H\,ii} to {\sc H\,i} regions decreases due to rapid recombination. Therefore, we suggest that the combination of [O III] and [C II] lines is a good probe to investigate the relative distribution of ionized and neutral gas in high-$z$ galaxies. In addition, we show that deep [C II] observations with a sensitivity of $\sim 10^{-2}~{\rm mJy~arcsec^{-2}}$ can probe the extended neutral gas disks of high-$z$ galaxies.

astro-ph.GA↗

arXiv2020

Inception of a first quasar at cosmic dawn

Earliest quasars at the cosmic dawn are powered by mass accretion onto supermassive black holes of a billion solar masses. Massive black hole seeds forming through the direct collapse mechanism are considered the most promising candidates but how do they grow and co-evolve with their host galaxies at early cosmic times remains unknown. We here present results from a cosmological radiation hydrodynamical simulation including self-consistent modeling of both Pop III and Pop II star formation, their radiative and supernova feedback in the host galaxy along with X-ray feedback from an accreting massive black hole (MBH) of $\rm 10^5 ~M_{\odot}$ in a halo of $\rm 2 \times 10^9~M_{\odot}$ from $z=26$ down to $z=16$. Our results show that energy deposition from X-rays in the proximity of MBH suppresses Pop III star formation for about 12 Myr while at the same time these X-rays catalyze $\rm H_2$ formation which leads to the formation of a Pop III star cluster of 500 $\rm M_{\odot}$ in the close vicinity of the MBH. We find that mode of star formation for Pop III is episodic and bursty due to the clumpy accretion while for Pop II it is continuous. The stellar mass of the host galaxy at $z \sim 16$ is $\rm 2 \times 10^7~M_{\odot}$ with a star formation rate (SFR) of $\rm \sim 0.1-1~M_{\odot}/yr$. In total, the MBH accretes $\rm 1.5 \times 10^6~M_{\odot}$ during 120 Myr with the mean accretion rate of $\rm \sim 0.01~M_{\odot}/yr$ corresponding to an average Eddington fraction of 50\%.

astro-ph.GA↗

arXiv2020

The First Billion Years Project: Finding Infant Globular Clusters at z=6

We explored a suite of high-resolution cosmological simulations from the First Billion Years Project (FiBY) at $z \geq 6$. All substructures within the simulations have been identified with the SUBFIND algorithm. From our analysis, two distinct groups of objects emerge. We hypothesise that the substructures in the first group, which appear to have a high baryon fraction ($f_{\rm b} \geq 0.95$), are possible infant GC candidates. Objects belonging to the second group have a high stellar fraction ($f_{\rm star} \geq 0.95$) and show a potential resemblance to infant ultra-faint dwarf galaxies. The high baryon fraction objects identified in this study are characterised by a stellar content similar to the one observed in present-day GCs, but they still contain a high gas fraction ($f_{\rm gas} \sim 0.95$) and a relatively low amount of dark matter. They are compact, dense systems. Their sizes are consistent with recent estimates based on the first observations of possible proto-GCs at high redshifts. These types of infant GC candidates appear to be more massive and more abundant in massive host galaxies, indicating that the assembly of galaxies via mergers may play an important role in building several GC-host scaling relations. Specifically, we express the relation between the mass of the most massive infant GC and its host stellar mass as $\log(M_{\rm cl}) = (0.31\pm0.15)\log(M_{\rm *,gal} + (4.17\pm1.06)$. We also report a new relation between the most massive infant GC and the parent specific star formation rate of the form $\log(M_{\rm cl}) = (0.85\pm0.30)\log(sSFR) + α$ that describes the data at both low and high redshift. Finally, we assess the present-day GC mass (GC number) -- halo mass relation offers a satisfactory description of the behaviour of our infant GC candidates at high redshift, suggesting that such a relation may be set at formation.

astro-ph.GA↗

arXiv2020

The Birth of Binary Direct-Collapse Black Holes

Supermassive primordial stars forming during catastrophic baryon collapse in atomically-cooling halos at $z \sim$ 15 - 20 may be the origin of the first quasars in the universe. However, no simulation to date has followed the evolution of these halos at resolutions that are high enough or for times that are long enough to determine if collapse actually produces SMSs. Here we report new cosmological simulations of baryon collapse in atomically-cooled halos for times that are long enough for SMSs to form and die as direct-collapse black holes (DCBHs). We find that the high infall rates required to build up such stars do persist until the end of their lives and could fuel the rapid growth of their BHs thereafter. Our simulations also demonstrate that binary and even small multiples of SMSs can form in low-spin and high-spin halos, respectively. This discovery raises the exciting prospect of detecting gravitational waves from DCBH mergers with LISA and tidal disruption events in the near infrared with the {\em James Webb Space Telescope} and ground-based telescopes in the coming decade.

astro-ph.GA↗

arXiv2020

Calibration of a star formation and feedback model for cosmological simulations with Enzo

We present results from seventy-one zoom simulations of a Milky Way-sized (MW) halo, exploring the parameter space for a widely-used star formation and feedback model in the {\tt Enzo} simulation code. We propose a novel way to match observations, using functional fits to the observed baryon makeup over a wide range of halo masses. The model MW galaxy is calibrated using three parameters: the star formation efficiency $\left(f_*\right)$, the efficiency of thermal energy from stellar feedback $\left(ε\right)$ and the region into which feedback is injected $\left(r\ {\rm and}\ s\right)$. We find that changing the amount of feedback energy affects the baryon content most significantly. We then identify two sets of feedback parameter values that are both able to reproduce the baryonic properties for haloes between $10^{10}\,\mathrm{M_\odot}$ and $10^{12}\,\mathrm{M_\odot}$. We can potentially improve the agreement by incorporating more parameters or physics. If we choose to focus on one property at a time, we can obtain a more realistic halo baryon makeup. We show that the employed feedback prescription is insensitive to dark matter mass resolution between $10^5\,{\rm M_\odot}$ and $10^7\,{\rm M_\odot}$. Contrasting both star formation criteria and the corresponding combination of optimal feedback parameters, we also highlight that feedback is self-consistent: to match the same baryonic properties, with a relatively higher gas to stars conversion efficiency, the feedback strength required is lower, and vice versa. Lastly, we demonstrate that chaotic variance in the code can cause deviations of approximately 10\% and 25\% in the stellar and baryon mass in simulations evolved from identical initial conditions.

astro-ph.GA↗

arXiv2019

UV regulated star formation in high-redshift galaxies

The first galaxies forming a few hundred million years after the Big Bang are the key drivers of cosmic evolution and ideal laboratories to study theories of galaxy formation. We here study the role of UV radiation in suppressing star formation in primordial galaxies by destroying molecular hydrogen, the main coolant in primordial gas and provide estimates of cold dense gas at the onset of star formation. To accomplish this goal, we perform three dimensional cosmological simulations of minihalos in different environments forming at $ z\sim 25$ by varying strength of background UV flux below the Lyman limit between 0.01-1000 in units of $\rm J_{21}=10^{-21}~erg/cm^2/s/Hz/sr$. Particularly, we include photo-detachment of $\rm H^-$, the self-shielding of $\rm H_2$ which both were neglected in previous studies and use updated reaction rates. Our results show that depending on the background level $\rm H_2$ formation is suppressed, delaying gravitational collapse until halos reach the atomic cooling limit. We find that the formation of cold dense molecular gas and subsequently star formation gets delayed by 100 to 230 Myr depending on the level of the background radiation and the growth history of the dark matter halos. The fraction of dense self-shielded gas is a strong function of the background flux and exponentially declines with the strength of incident UV flux above $\rm J_{21} \geq 1$. We find that taking into account $\rm H_2$ self-shielding is crucial for accurately estimating the amount of cold dense gas available for star formation.

astro-ph.GA↗

arXiv2019

Galaxy evolution and radiative properties in the early Universe: multi-wavelength analysis in cosmological simulations

Recent observations have successfully detected UV or infrared flux from galaxies at the epoch of reionization. However, the origin of their radiative properties has not been fully understood yet. Combining cosmological hydrodynamic simulations and radiative transfer calculations, we present theoretical predictions of multi-wavelength radiative properties of the first galaxies at z=6-15. We find that most of the gas and dust are ejected from star-forming regions due to supernova (SN) feedback, which allows UV photons to escape. We show that the peak of SED rapidly shifts between UV and infrared wavelengths on a timescale of 100 Myr due to intermittent star formation and feedback. When dusty gas covers the star-forming regions, the galaxies become bright in the observed-frame sub-millimeter wavelengths. In addition, we find that the escape fraction of ionizing photons also changes between 1-40% at z>10. The mass fraction of HII region changes with the star formation history, resulting in the fluctuations of metal lines and Lyman-alpha line luminosities. In the starbursting phase of galaxies with the halo mass $\sim 10^{11}\,{\rm M_{\odot}}$ ($10^{12}\,{\rm M_{\odot}}$), the simulated galaxy has $L_{\rm [O\,III]} \sim 10^{42}\, (10^{43})\, {\rm erg\,s^{-1}}$, which is consistent with the observed star-forming galaxies at z>7. Our simulations suggest that deep [C II] observation with ALMA can trace the distribution of neutral gas extending over $\sim 20$ physical kpc. We also find that the luminosity ratio $L_{\rm [O\,III]}/L_{\rm [C\,II]}$ decreases with bolometric luminosity due to metal enrichment. Our simulations show that the combination of multi-wavelength observations by ALMA and JWST will be able to reveal the multi-phase ISM structure and the transition from starbursting to outflowing phases of high-z galaxies.

astro-ph.GA↗