DAWN Summit 2022

Registration is now closed. If you wish to attend, you need to write directly to the organizer and apologize for being a sløv banan.

Not to be confused with the temporally adjacent DAWN Conference.

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

    About the Summit

    We are pleased to announce the annual summit at the Cosmic Dawn Center.


    Each year, new faces appear at DAWN, and you forget what your old colleagues are spending their time on. The purpose of the DAWN Summit is to get to know each other, scientifically and socially, get an overview of the science of DAWN, and hopefully spawn new collaborations.

    Time and venue

    The meeting will take place on 20–21 June.

    Monday 20 June, 9:00–16:30, at DTU

    The first day will be at DTU, which is north of Copenhagen.

    • Cycling from the Niels Bohr Building (NBB) is 11 km, and a whole bunch of DAWNers will probably go together.
    • Alternatively, Bus 150S (and E15) leaves every 5 minutes from Nørreport in the center of Copenhagen, past the NBB, and to DTU in 25 minutes. Get off at "Rævehøjvej, DTU" and walk 750 m to Anker Engelundsvej 101, entrance A at the parking lot, then find Meeting Room M1.

    See also here how to get from the airport to Nørreport.

    Tuesday 21 june, 9:00–16:00, at UCPH

    The second day will be at the Lundbeckfond Auditorium, 367 m from DAWN/NBB.


    This year, we would like talks to have a special focus on JWST, if appropriate (i.e. if you're not really working on JWST stuff, don't worry, we still want to hear about your work).

    Unless recordings or call-ins are specifically requested, the meeting will be in-person only. This is to avoid the significant overhead when dealing with bad connections, computer setups, etc. between the talks.

    Contributed talks

    should be roughly 10+3 minutes, so that a new speaker may start every 15 minutes.

    Review talks

    will be 20–25 minutes, plus time for questions and discussions; 30 minutes in total including overhead. They will focus on advances in the field particularly within the last year, as well as expected advances when JWST starts observing.

    List of participants

    Listed in order of registration time, with Desika registering only 4'14" after receiving the invitation:

    1. 🥇 (+4'14") Desika Narayanan (U. of Florida)
    2. 🥈 (+6'07") Lijie Liu (DTU)
    3. 🥉 (+7'07") John Weaver (NBI)
    4. Francesco Valentino (NBI)
    5. Pascal Oesch (NBI & Geneva)
    6. Kasper Heintz (NBI)
    7. David Vizgan (DTU)
    8. Rebeca Reyes Carrion (U. of Puerto Rico Mayaguez)
    9. Helena Baungaard-Sørensen (NBI)
    10. Malte Brinch (DTU)
    11. David Blanquez (DTU)
    12. Shuowen Jin (DTU)
    13. Birgitta Nordström (NBI)
    14. Joonas Viuho (NBI)
    15. Peter Laursen (NBI)
    16. Trity Pourbahrami (Gordon & Betty Moore Foundation)
    17. Claudia Lagos (U. of Western Australia)
    18. Tom Reynolds (NBI)
    19. Steven Gillman (DTU)
    20. Ting-Yi Lu (NBI)
    21. Lauren Elicker (U. of Cincinnati)
    22. Thomas Herard (Sorbonne U. + Paris Observatory)
    23. Anne Hutter (Kapteyn Astronomical Institute + NBI)
    24. Adam Carnall (Edinburgh U.)
    25. Gonzalo Prieto Lyon (NBI)
    26. Guozhen Ma (NBI)
    27. Ezra Huscher (DAWN-IRES)
    28. Simone Vejlgaard Nielsen (NBI)
    29. sune Toft (NBI)
    30. Luis Colina (CAB (CSIC))
    31. Albert Sneppen (NBI)
    32. Eric Rumsfeld (NBI)
    33. Clara Giménez Arteaga (NBI)
    34. Ioannis Mageiras (NBI)
    35. Nikolaj Bjerregaard Sillassen (DTU)
    36. Lise Christensen (DAWN)
    37. Charles Steinhardt (NBI)
    38. Natalie Allen (NBI)
    39. Thomas Greve (DTU)
    40. Bo Milvang-Jensen (NBI)
    41. Casey Carlile (U. of Kansas)
    42. Sarah Bodansky (U. of Massachusetts Amherst)
    43. Lukas Zalesky (Institute for Astronomy, U. of Hawaii)
    44. Hanga Andras-Letanovszky (NBI/DAWN-IRES)
    45. Mahsa Kohandel (Scuola Normale Superiore)
    46. Sian Phillips (ARI (Liverpool JMU))
    47. Daniele Malesani (Radboud U.)
    48. Iris Jermann (U. of Geneva)
    49. Zoe Kearney (McGill U.)
    50. Kartik Sheth (NASA HQ)
    51. Minju Lee (DTU)
    52. Charlotte Mason (NBI)
    53. Jiaming Yao (NBI)
    54. Kate Whitaker (UMass Amherst)
    55. Iary Davidzon (NBI)
    56. Francesca Rizzo (NBI)
    57. Seiji Fujimoto (NBI)
    58. Victoria Strait (NBI)
    59. Aswin P Vijayan (DTU)
    60. gabriel brammer (NBI)
    61. Jackson Mann (UC Riverside)
    62. Meghana Killi (NBI)
    63. Francesco Valentino (NBI)
    64. Vasily Kokorev (NBI)
    65. Vadim Rusakov (NBI)
    66. Conor McPartland (NBI)
    67. Bitten Gullberg (DTU)


    Times are approximate. Click talks to view abstracts.

    You can download a .pdf version of the program by clicking this button:

    Monday 20 June

    Monday's program will end around 16:30.

    Venue: Meeting room M1, Anker Engelunds Vej 101 (DTU)


    Light breakfast


    Sune Toft & Thomas Greve


    Special session #1: JWST updates

    Review by Gabe Brammer

    An overview of the JWST MIRI GTO Imaging program inc. strategy, simulations, depth, science analysis and future plans.



    Coffee break


    Galaxy evolution

    Review by Minju Lee


    Based on 1 million priors from the COSMOS2020, radio and ALMA archival catalogs, we have deblended the deepest and the largest FIR to radio data sets in COSMOS using our Super-deblending thechnique. Besides the deblended MIPS and Herschel photometry, the new super-deblended catalog is significantly improved in FIR and radio by incorporating the latest SCUBA2, MeerKAT, VLA and GMRT images. The new catalog includes 15k FIR/submm detections (S/N>5) and 30k radio detections (S/N>3), which is now ready for public sharing. I will present the new catalog, report our recent work on a sample of z=3.5--6 galaxies with optically thick dust in FIR, and finding of a high-z protocluster in COSMOS.

    I will present a public code GalCluster, which is specialized for (proto)cluster search using overdensity mapping. This code has been verified in the COSMOS field by re-discovering two known protoclusters at z>2, and it is ready for public use. By applying GalCluster in the COSMOS2020 catalog, we discovered a high overdensity of galaxies HPC1001 at z~3.7, which has the densest structure even seen at z>3 with massive dark matter halo mass. I will report the finding of HPC1001 and characterize physical quantities in this structure.



    Sandwiches will be provided.


    The Cosmic Dawn Survey is a 50 square degree multi-wavelength survey of the Euclid Deep and Calibration fields. A key component of this endeavor is the The Hawaii Twenty Square Degree Survey (H20) which targets the Euclid Deep Field North and Euclid Deep Field Fornax. These two 10 deg2 fields benefit from the deepest coverage (~25 mag AB) from Spitzer over such large areas. Our survey provides complementary imaging from CFHT MegaCam and Subaru Hyper Suprime-Cam needed to determine photometric redshifts and therefore utilize the Spitzer data to their fullest extent, as well as Keck DEIMOS spectroscopy for follow-up. Combining this with the near-IR imaging from Euclid will enable studies of galaxy demographics to z ~ 10 and beyond.

    Galaxies at high redshift have complex morphologies, with uneven distributions of stars, gas, and dust. To characterize the complexity, we study the spatial offsets among tracers of various galaxy components - the UV continuum emission tracing the stars, the singly-ionized carbon [CII] tracing multiple phases of enriched gas in the ISM, and the FIR continuum tracing the dust. For this, we use a statistically significant sample from the ALPINE survey consisting of 75 main-sequence star-forming galaxies at z ~ 4-6. We find significant spatial offsets among the various phases, suggesting that the general assumption of co-spatial emission in galaxy simulations needs to be revisited.

    Most galaxies follow the star forming main sequence encoded by the star formation rates and stellar masses. The decoupling of these properties signifies quenching of star formation. By investigating the mass scale of the youngest stars in galaxies at various redshifts, there appear hints of a new galaxy population that deviates from the main sequence. High star formation rates, massive young stars and small overall stellar masses in these galaxies are consistent with the yet elusive early phase of galaxy formation. These hints allow hypothesising formation mechanisms and making predictions about their structural appearance. Here, the morphological predictions are tested using resolved HST observations of galaxies at low redhsift.



    Coffee break



    Review by Adam Carnall

    I will discuss current and future efforts to constrain the star-formation histories (SFHs) and stellar metallicities of the most massive galaxies, from the end of the peak epoch of cosmic star formation at z~1, all the way back to the first billion years of cosmic history. I will begin by discussing results from the VANDELS survey, which provides high-SNR rest-frame UV-optical continuum spectroscopy for a mass-complete sample of massive quiescent galaxies at z>~1 for the first time, allowing detailed studies of their SFHs and metallicities to understand when and how these galaxies quenched. I will also discuss massive, quiescent galaxies at the highest redshifts, focusing on the upcoming revolution we can expect as a result of JWST data.

    While standard SED-fitting codes assume simple analytical shapes, more complex star formation histories (SFH) may be needed to explain the previous evolution of what we observe as massive quiescent galaxies at high redshift. Nonparametric SED-fitting tools offer a solution to this. We explore these issues by diving into Prospector's nonparametric SED-fitting code in anticipation of VLT X-Shooter spectroscopic observations of a sample of 7 massive quiescent galaxies in the COSMOS-field at z~2.5-3. In order to draw conclusions on the reliability of this method, posterior SEDs, as well as various properties, and constrains on SFHs will be analysed for reconstructing continuous and bursty SFHs.


    Special session #2: Outreach and science communication

    Do you want your science boosted to reach more colleagues, funding agents, and the general public? Learn here how.

    I will give an update on how I have been contributing to the areas of science communication, science philanthropy, and inclusive science since leaving my position at Caltech in 2019 to join the Gordon and Betty Moore Foundation.

    Tuesday 21 June

    Tuesday's program will end around 16:00.

    Venue: Lundbeckfond-auditoriet, Ole Maaløes Vej 5 (UCPH)


    Light breakfast


    Putting in all together (theory)

    Review by Desika Narayanan

    Observations from have robustly shown that the velocity dispersion of gas in galaxies increases significantly from z=0 to z=2-3, which has been interpreted as galaxies at high z being more turbulent. Although hydrodynamical simulations have shown to agree with this trend, there has been a lack of in-depth studies of the physical origin of this. I will present an exhaustive study we have conducted dissecting the role of stellar feedback, gravitational instabilities, external turbulence injection via gas accretion and galaxy mergers on galaxies in the EAGLE simulations. I will show these mechanisms dominate in different regimes, and that the gas velocity dispersion is a characteristic of a galaxy rather than a transient property.

    Generally, two techniques are used to determine the redshift of Supernova Ia used in cosmological analysis: direct measurement from spectral features of the host-galaxy, or in-directly from spectral features of the supernova itself. These methods differ in the resulting redshift uncertainty, and the latter technique may additionally introduce unknown biases into the cosmological supernova sample.

    Upon dividing the Pantheon catalog according to the technique used to measure redshift, recent work has indeed shown a discrepancy in the cosmological parameters determined from these two sub-samples.

    I will go through these results and present recent development in our understanding of the problem.

    A key difficulty in linking gamma-ray bursts with their physical origins is that it has not yet been possible to unambiguously classify every GRB as either short or long. This is because there is substantial overlap in the properties of both short and long GRBs, and neither duration nor any other parameter so far considered completely separates the two groups. Recent work showed that a machine-learning dimensionality-reduction algorithm, t-SNE, was able to cleanly separate every object in the Swift GRB sample into two distinct groups. Here, we apply the same methodology to additional samples, providing a catalog that cleanly separates all Swift/BATSE/Fermi GRBs into two distinct groups, corresponding to short and long GRBs.

    Fully hydrodynamical simulations enable us to explore the relationship between the emission from stars and the impact the associated distribution of these stars and intervening dust has on the observed properties. Using the FLARES: First Light And Reionisation Epoch suite of simulations I will talk about the consequences of this, such as the effect on the shape of the UV luminosity function, the attenuation curve, the Balmer decrement and variation across lines-of-sight.


    Coffee break


    Interstellar medium

    Review by Lise Christensen


    > The abstract form seems way too short. I could only fit in a few sentences which makes no sense. I can email you the abstract if that's ok.

    No Vasily, that is not okay. There is a reason for the length restriction you know. Suck it up!

    I will present an overview of my PhD so far. First sharing the results of the project I have been working on for the first half of the PhD, where we have studied a sample of local star-forming galaxies observed with broad- and narrow-band photometry from HST. We use various indicators to obtain the SFR per spatial bin, and find that Paschen-beta traces star-forming regions where the H-alpha and the optical stellar continuum are heavily obscured. The dust-corrected Pa-beta SFR also recovers the 24micron and the infrared luminosity inferred SFRs. Finally, I will introduce the project I will work on next, studying the gravitationally lensed galaxy A1689-zD1 at z=7.13, which is one of the earliest known galaxies where the ISM has been detected. I will study the IRX-beta relation and perform multi-band SED-fitting analysis, both in an integrated and spatially-resolved sense.

    I will present a large statistics of 180 dust continuum sources blindly identified in ALMA Lensing Cluster Survey (ALCS), including the 1.2-mm number counts, the redshift evolution of infrared luminosity function, and the total (=obscured+ unobscured) cosmic star-formation rate density out to z~8.



    Buffet at the Bio-canteen.


    First stars and galaxies

    Review by Pascal Oesch

    The fraction of hydrogen ionizing photons escaping from these first galaxies plays a key role in determining the galaxy population reionizing the Universe, however its dependence on galactic properties remains highly unconstrained. Galaxies showing Lyman-alpha emission, Lyman-alpha emitters (LAEs), can potentially trace the dependence of the ionising escape fraction on the galaxies' mass. I will present my new LAE model that I have incorporated into my semi-numerical model of galaxy evolution and reionization ASTRAEUS, and discuss how sensitive the Lyman-alpha luminosity functions and spatial distribution of LAEs are to the dependence of the ionizing escape fraction on halo mass.

    I will talk about The CAnadian Unbiased NIRISS Cluster Lensing Survey and the science cases we hope to address with it.


    Coffee break



    Review by Anne Hutter

    One of the best trackers of the neutral fraction of the Universe at the epoch of reionization are Lyman Alpha Emitters. We look to get a better understanding of how Lyman Alpha Emission behaves at z~5-6, since these galaxies should have similar properties to those we find at Reionization, but are not as heavily affected by the IGM. In particular, we studied their Lyman Alpha EW, profile shapes, and look for correlations with more easily measurable properties such as Muv and UV Slope.