DAWN Summit 2021

This event is outdated.

About the Summit

We are pleased to announce the Cosmic Dawn Center Summit 2021 which finally can be held in person again!


The purpose of the meeting is to get acquainted with what your colleagues are working on, foster new collaborations, and get an overview of the science of DAWN. There will of course also be time for discussions.

Time and venue

The meeting will take place on September 20 to 22 with events at NBB and at DTU.

Online attendees can find a Zoom link here.


  • Contributed talks are 10 minutes each, plus time for questions (so that a new speaker can start every 15 minutes).
  • Students' talks should be only 5 minutes, plus time for questions (so that a new speaker can start every 7½ minute).
  • Reviews are 20–25 minutes, plus questions (30 minutes in total)


Table of contents
    Add a header to begin generating the table of contents


     Science session:   First Galaxies
     Science session:   Interstellar Medium
     Science session:   Quenching
     Science session:   Galaxy evolution
     Science session:   Reionization
     Discussion session:   JWST
     Discussion session:   Midterm Evaluation
     Students' session:

    Monday 20.9

    Niels Bohr Building
    01.0H.146 (ground floor)
    09:30 – 10:00 
    Light breakfast ☕️ 🥐

    Welcome to the DAWN Summit 2021 (Sune Toft & Thomas Greve)

    Welcome to DAWN (Guarn)

    Welcome to the DAWN website (Peter Laursen)

    Science session: First Galaxies 
    Review: Victoria Strait

    While the un-obscured SFRD has been determined very accurately out to z~10-11 based on rest-frame UV observations, the contribution from dust obscured star-formation requires sensitive FIR-to-mm data and remains largely unexplored in the first 1 Gyr. I will present a first measurement of the IR luminosity function (LF) at z=7-8 based on the large ALMA Reionization Epoch Bright Emission Line Survey (REBELS). In particular, REBELS has constrained the dust continuum for 40 of the most massive z>6 galaxies, enabling us to derive their IR luminosities and obscured star-formation. We compare our measurements to theoretical models and discuss the contribution of un-obscured and dust-enshrouded star formation across the full cosmic history since z~8.

    11:30 – 11:45
    Morning break ☕️

    We report the discovery of the least luminous, the most distant red quasar, GNqz7-J1236, at redshift z=7.189 hosted by a dust-enshrouded starburst galaxy. Deep HST imaging and slitless spectroscopy show a strong Lyman break and point-source morphology. The rest-frame UV continuum is redder and 20 times fainter than the handful of luminous quasars known at these redshifts. The quasar is uniquely faint in X-rays, suggesting that a less massive black hole <~ 10^8 Msun is rapidly growing with super-Eddington accretion, while the host galaxy is the most luminous in dust emission at this epoch with SFR~1,600 Msun/yr. These properties confirm predictions from cosmological simulations suggest that GNqz7-J1236 is a direct progenitor of luminous quasars at later epochs.

    The 21cm signal from neutral hydrogen in the early universe will provide a new way to infer information about the properties of the universe's first stars and galaxies, and to investigate dark matter at small scales. However, forecasting and interpreting this signal is challenging due to the large number of - mostly unconstrained - parameters that affect it. I will describe the key features of the 21cm signal and its synergies with direct observations of galaxies to understand Cosmic Dawn, and present a new efficient tool to explore the astrophysical and cosmological parameter space.

    Using a large sample of sub-L galaxies, with similar UV magnitudes, MUV ≃ −19 at z ≃ 6, extracted from the FirstLight simulations, we show the diversity of galaxies at the end of the reionization epoch. We find a factor 40 variation in the specific star-formation rate (sSFR). This drives a 1 dex range in equivalent width of the [OIII]λ5007 line. [OIII]-bright ([OIII]/Hα> 1) emitters have higher ionization parameters and/or higher metallicities than Hα-bright ([OIII]/Hα< 1) galaxies. According to the surface brightness maps in both [OIII] and Hα, [OIII]-bright emitters are more compact than Hα-bright galaxies. Hα luminosity is higher than [OIII] if star formation is distributed over extended regions. [OIII] dominates if it is concentrated in compact clumps.

    12:30 – 13:30
    Lunch 🥪

    Science session: Interstellar Medium 
    Review: Francesco Valentino

    Using a sample of local IR bright galaxies observed with HST, we use the Hα and Paβ emission lines to calculate the extinction affecting the gas emission (Av,gas) for each object. With a new spatially resolved SED-Fitting tool, we infer the extinction affecting the stellar population (Av,stars). By comparing both Av estimates on a bin-to-bin basis, we find spatial bins that have a more extreme “extra attenuation” prescription than the usual consideration from Calzetti et al. 2000. This suggests that the attenuation effect caused by the stellar birth clouds in these galaxies could be higher than what we usually assume, and could establish the Balmer-to-Paschen decrement as a more robust indicator of extinction compared to commonly used techniques.

    Dust is made mostly in protostars, not AGBs, not supernovae, and not cold-grown in the ISM

    While gas mass reservoirs of SFGs are routinely measured out to the highest redshifts, this is a very challenging feat for high redshift QGs. We select samples in the GOODS-S field taking advantage of the existing ZFOURGE UV to mid-IR multi-wavelength catalogues and are complemented by the new GOODS-S ALMA 1.1 mm map. Selecting QGs based on their UVJ colours or their distance to the MS we measure the rest frame dust emission (or estimate the corresponding upper limits) in the the R-J tail of the populations through ALMA stacking and infer their fgas using a range of techniques. We find that galaxies residing in the envelope below the MS (1/5 < δMS < 1/2) have an average fgas that ranges from 67.5% down to 30.9% (depending on the adopted method), while U V J selected QGs have fgas < 18.8%.

    I will give a brief overview of ongoing and future projects on ISM, SF and quenching

    The first epoch of galaxy formation is governed by the infall of neutral, pristine gas. These neutral atomic hydrogen (HI) gas reservoirs will subsequently cool and condense into molecular gas and initiate star-formation. However, it has not yet been possible to estimate the HI gas content of high-redshift galaxies due to the weakness of the HI 21-cm line. In this talk I will present recent work on how to infer the HI gas content of distant galaxies using [CII] as a tracer, based on an empirical measurement of the [CII]-to-HI conversion factor in gamma-ray burst afterglows. This approach allows us to infer the HI gas mass of individual [CII]-detected galaxies and the cosmic HI gas mass density in galaxies from the end of the reionization epoch to the present.

    15:15 – 15:45
    Afternoon break ☕️

    In our work we present the analysis of the far-infrared properties of ∼4,700 star-forming galaxies at z <4.5, drawn from the deepest catalogues in the GOODS-N and COSMOS fields. We develop a novel panchromatic SED fitting algorithm, Stardust, that models the emission from stars, AGN, and infrared emission arising from dust heated by star formation without relying on energy balance assumptions. With our code we recover robust estimates of FIR properties of our objects. In particular, we focus on the dust-to-stellar mass ratio, which we find to be increasing by a factor of 10 from z= 0 to z= 2, while remaining flat at higher−z, mirroring the evolution of the gas fraction. Finally, we discover a rather puzzling population of gas giants, that seem to defy the current evolutionary trends.

    even if you think pahs and dust grains are boring I promise you'll not only think they're less boring after this talk, but also that I'm absolutely bananas for trying this project.

    Modeling Lyman α-emitting galaxies (and quasars) is essential to understanding their nature, but is complexificated by the scattering of Lyα photons in the interstellar and -galactic medium. Fitting observations to numerical models can help, but most such model are symmetrical and hence do not capture the intrinsic asymmetrical escape of Lyα photons. In this talk I'll present a method to model this effect, utilizing not only the spectral, but also the spatial information of the systems.

    Science session: Quenching 
    Review: Kate Whitaker

    The identification of numerous massive quiescent galaxies as early as z~4 has recently cast doubt on our understanding of how quenching works in the early universe. Whilst it was thought that these objects were the result of an instantaneous burst of star formation, there is growing evidence to suggest that there are multiple paths to quenching. We present the first results from a study of massive quiescent galaxies at 3<z<4 in the latest COSMOS catalog, which is unprecedented in its ultra-deep and homogenous optical/near-infrared photometry. We fit a sample of ~1200 galaxies with the bleeding edge SED fitting code dense basis and measure the quenching timescales and star formation histories, finding a diversity of quenching pathways already at z~3.5. 

    In this talk, I will highlight a few recent results from the ongoing REQUIEM Survey (REsolving QUIEscent Magnified galaxies). With Hubble and ALMA data, we explore the gradients in the stellar populations and the cold dust and gas content of a premier sample of strong gravitationally lensed quiescent galaxies at cosmic noon. We are also using results from the SIMBA simulation to place the REQUIEM results in context and explore the prospects for using the cold dust continuum to trace molecular gas reservoirs. 

    In this talk, I will present a few results on the formation and death of distant massive galaxies obtained during the first 1.5 years of the "Galaxies: Rise And Death" (GRAD) project. I will touch upon an ALMA study of the cold molecular gas in main-sequence star-forming galaxies, showing what the ISM properties can teach us about their "normality" and the role played by active galactic nuclei during the growth of these galaxies. I will also present recent work aimed at identifying the first quenched galaxies in the Universe and how well state-of-the-art cosmological simulations can reproduce the early appearance and death of these systems in the history of the Universe.

    The broad line quasar SDSS 0956+5128 was shown to demonstrate three distinct emission line velocities. Balmer lines Ha, Hb and MgII exhibited different extreme velocity offsets with respect the systemic redshift, 1200 and 4100 km/s, respectively. The theory of a post-merger recoiling black hole shaping the central region held consistent against the existing data, although lacked further evidence. By analysing new observations of Lya and CIV lines and resolved images of the host galaxy, we rule out the possibility of the recoiling black hole. Instead, the cumulative evidence appears consistent with an energetic event producing an outflow of matter in the central broad line region.

    Tuesday 21.9

    DTU Space
    Room S08 in the main 101 building (indicated red on this map)
    09:30 – 10:00
    Light breakfast ☕️🥐
    10:00 – 10:20 
    Conor McPartland: Recent updates from the H20 survey

    Science session: Galaxy Evolution 
    Review: Bitten Gullberg

    In this talk I will try to present a brief and hopefully a coherent picture of some of the works I have done during my PhD using simulations of galaxy formation and evolution to study the evolution of dust in galaxies as well as its effect on the galaxy spectral energy distributions (SEDs). 


    11:20 – 11:30
    Morning break ☕️

    The dynamical properties of high redshift galaxies reveal important clues on their formation and assembly, providing also strong constraints on galaxy formation models. In this talk, I will present some preliminary results on the morpho-kinematic analysis of star-forming galaxies at the epoch of reionization. Using mock data of galaxies from a zoom-in cosmological simulation, I am defining the data quality (angular resolutions and signal-to-noise ratio) of future ALMA (and JWST) observations needed for a robust characterization of the assembly mode of primeval galaxies. Finally, I will present a project aiming at providing a systemic kinematic analysis of all z ~ 0.1 - 4 star-forming galaxies with high-resolution ALMA archive data.

    John Weaver: COSMOS2020: A consistent view of galaxy mass assembly since z~7

    IR-luminous galaxies are very common in. the high-z universe. I´ll be presenting some recent results based on ALMA data of a large sample of low-z U/LIRGs. In particular I´ll be discussing the power source, the compactness of the dust and. molecular gas, their distribution relative to the ionized ISM and stellar host. Comparison with both low- and high-z samples will be discussed.

    Being the largest, most massive and, arguably, the most beautiful gravitationally bound structures in the Universe, galaxy clusters demand our fascination and scientific scrutiny. I will talk about the latest findings from the SPT-SMG and RAGERS surveys - two surveys that traces the build-up of dusty starburst galaxies in distant clusters at Cosmic Dawn (z~4-6) and Cosmic Noon (z~1-3), respectively.

    12:30 – 13:30
    Lunch 🥪

    We have conducted a systematic search for protocluster candidates in the redshift range 6 < z < 10 in the COSMOS field using the recently released COSMOS2020 source catalogue. We select galaxies using a number of selection criteria to obtain a sample of galaxies that have a high probability of being inside a given redshift bin. We then apply overdensity analysis to the bins using two estimators, a Weighted Adaptive Kernel Estimator and a Weighted Voronoi Tessellation Estimator. We have discovered a number of highly significant (>4 sigma) galaxy overdensities across the redshift range. We compare the evolution in the total star-formation rate and stellar mass content of the protocluster candidates across the redshift range 6 < z < 10.

    I will present a summary of my Master's thesis where we investigated the reason for compactness of Lyα emitters (LAEs) at high z. We studied the morphology and SEDs of a sample of 17 bright, yet extremely compact (∼ 1 kpc size) LAEs at z ∼ 3.5−7 selected from a parent sample of 1060 Lyman-break galaxies (LBGs) observed in HST data in 4 of the 5 CANDELS fields. Based on our results and findings from literature, we tentatively concluded that the compact nature is partly a true physical feature of high-z LAEs and partly due to observational limitations.

    I will introduce my recent studies on galaxy evolution in different environments at cosmic noon. The talk will cover case studies targeting a protocluster at z=2.5, a study in progress using ALMA archive to explore the gas content and improving the (cold) gas scaling relations, and expected future outcomes in the coming years as part of this long-term project. If time allows, I will also briefly introduce my side work for probing ISM properties of massive galaxy at high redshift (z>4).

    14:30 – 15:00
    Afternoon break ☕️

    I will present the project which will be my aim focus during my Carlsberg Reintegration fellowship at DAWN/DTU. This project (PRISMLIGHT) builds on the ALMA SPECtral line Survey (ASPECS), which unique by identifying CO-bright galaxies, irrespective of their SFRs and stellar masses. ASPECS therefore provides an unbiased view of the physical conditions of the ISM in galaxies at z =1-3, spanning a large range in SFR (~3-300Msun/yr) and stellar masses (~3x10^9-3x10^11Msun). Using near-infrared KMOS data, I will aim to measurement the dynamical masses and metallicities of a subsample of the ASPECS galaxies. Such measurements are instrumental to better understand galaxy evolution, as they reveal how stars are formed in both rapidly evolving and more quiescent environments.

    Using NOEMA and ALMA line scans, we spectroscopically confirmed redshifts of 10 dusty galaxies at z=3.6--5.8 via solidly detecting multiple CO and/or CI transitions. Detailed SED analysis shows that they have extremely cold dust temperatures and abnormally steep Rayleigh-Jeans slopes, indicating CMB impacts on mm continuums. High resolution ALMA imaging revealed their compact morphologies. However, their extremely cold dust temperatures appear to conflict with the high IR surface density, the massive and compact dust content, strongly suggesting that their dust contents are optically thick in far-infrared. I will show multiple evidence of optically thick dust in this sample. This work paves the way for diagnostics of optically thick dust in the early universe.

    The demographic of local main-sequence galaxies is well characterised by the Hubble sequence and several tight scaling relations, that interconnect the galaxies' fundamental properties. In this talk, I will discuss my ongoing work to constrain the properties of the ISM at z=1-2 that define the secular evolution of main-sequence galaxies and lead to the emergence of these scaling relations.

    15:45 – 15:50
    Short break to breath and stop thinking about galaxy evolution and start thinking about reionization

    Science session: Reionization 
    Review: Charlotte Mason

    Constraints on physical properties of galaxies in the redshift range z~6-10 are key for a full understanding of the process of reionization and early galaxy evolution, including the onset of star formation. I will present results on the highlights from my study of ~200 z ~ 6-10 galaxy candidates from the Reionization Lensing Cluster Survey (RELICS) survey using photometric redshift fitting. I will also discuss a remarkable extreme Lyman alpha emitting source at z~7 and the implications of exciting future data. Each of the galaxies I will discuss are perfect laboratories for JWST, ALMA, and ELTs to study in great detail the physical properties of galaxies in the early universe.

    Scattered light produced by diffraction gratings, the main dispersive elements used in astronomical spectrographs, is a long-known source of uncertainty in astronomical observation. Until this far, the scattered light have not been the most serious problem affecting the observation. However, recently the detector technologies used in astronomy have become increasingly better and single photon detection have been demonstrated in laboratory environments for various different detector types used in optical and NIR wavelengths. Consequently, understanding both the spectrographs themselves and night sky emission have become more crucial. In this presentation I will give you an overview of my PhD project and present the progress so far.

    Over the past twenty years, deep and sophisticated surveys have uncovered thousands of faint young galaxies that were already in place when the Universe was less than a billion years old. These efforts are characterizing precisely the overall abundance and spatial distribution of the first galaxy population, measurements which are critical for models of early structure formation. Did these galaxies drive cosmological reionization, the process that transformed the Universe into a plasma? Unfortunately, while we now know that those galaxies were there, we know little about know what they contributed, if anything, to reionization. In order to make progress, it is necessary to understand how they enriched and irradiated their local environments. I use numerical simulations as a reference point for studying how observations of galaxies and the circumgalactic and intergalactic media address this problem. My goals are to understand how measurements of galaxies and their environments can be used both to chart the progress of reionization, and detect directly the contribution that galaxies made to it.

    Wednesday 22.9

    Venue: Charlottenlund Fort
    09:30 – 10:00
    Light breakfast☕️🥐

    Some of the most metal poor (first) stars are more abundant i r-process elements than others. Why? 

    The future of the journal A&A.

    The European Astronomical Society is growing.

    10:10 – 12:30 
    Discussion session: JWST

    Peter Jakobsen: Status & Future Timeline of the JWST
    Pascal, Victoria, and Gabe: JWST Working Groups
    11:15 – 11:30
    Morning break ☕️
    Instruments & methods:
    → Steve Gillman: MIRI
    → Nina: NIRspec
    → Gabe: NIRISS & NIRcam
    Successful DAWN Proposals in Cycle 1:
    → Seiji
    → John
    → Gabe
    → Pascal
    12:30 – 13:30
    Lunch 🥪

    13:30 – 14:15 
    Students' session

    Both global, intermediate and local scales of Climate Change have been studied extensively, but a unified diagnostic framework for examining all spatial scales concurrently has remained elusive. Here we present a new tool-set using spherical harmonics to examine climate change from 1850 to 2021 on spatial scales ranging from 20.000 to 200 km. First, this approach confirms previous observations with respect to spatial variations for the lowest modes and largest scales, while local fluctuations are shown to be consistent with 2-dimensional turbulence. The observed turbulent cascade suggests that climate change feeds increasing volatility on all spatial scales from 2.000 to 200 km. Surprisingly, this increase is not spherically symmetric, but driven by growing volatility along longitudes.

    The far-infrared SEDs of quiescent galaxies are usually faint, but stacking SEDs can help us increase signal-to-noise ratio and deduce their average properties. Before doing these, now I am trying to use UVJ diagram and apply different cuts to select out quiescent galaxies from star-bursting galaxies.

    Using Deep Learning to build fast and accurate metamodels of existing simulation software reduces the computational cost of running expensive, slow simulations. These emulators can be used where the sheer number of evaluations of the original simulation software would make the task prohibitly expensive. For the case of CLOUDY, the DeepCLOUDY emulator will allow us to look at larger cosmological volumes.


    We investigate the utility of the fine structure line of ionized carbon ([CII]) at 158 microns as a tracer of molecular and atomic hydrogen in the early universe. We present a conversion factor for [CII] luminosity and molecular gas mass and investigate secondary contributions due to star formation rate and metallicity. We conclude with a discussion of tracing atomic hydrogen with [CII], which will be the focus of my Fulbright scholarship. 

    Because of the different abundances and origins of nucleosynthesis that isotopes have, isotopes and isotope ratios start to become powerful tools to study star formation and galaxy evolution. It recently has been proposed that 13C/18O can be used to indicate a top-heavy initial mass function in distant starburst galaxies. However, the same effect can also be cased by a very young starburst. Chemical fractionation can also play a role. Both effects have not been considered by detailed analysis. To investigate these results, we have investigated the correlation between the clusters ages and the isotope ratios in the central molecular zone of NGC253. We have also used a time dependent gas-grain chemical model UCLCHEM to evaluate the effects of chemical fractionation in starburst environment. 

    14:15 – 14:45
    Afternoon break ☕️

    14:45 – 16:00 
    Discussion session: Midterm evaluation