{% load staticfiles %} Hawaii Two-0 | Extragalactic Legacy Survey
Hawaii Two-0
Hawaii Two-0

A 20 deg2 deep Subaru Hyper-SuprimeCam imaging and Keck LRIS + MOSFIRE spectroscopic survey
of the two primary Euclid Deep Calibration Fields (NEP and CDFS).

Read the abstract


We propose the Hawaii Two-0 (H20) survey, which will help to put Hawaii at the forefront of galaxy evolution studies for the next decade or more. This will be achieved by carrying out a 20 deg2 deep Subaru Hyper-SuprimeCam (HSC) imaging and Keck LRIS + MOSFIRE spectroscopic survey of the two primary Euclid Deep Calibration Fields (NEP and CDFS). These two 10 deg2 fields are the targets of the largest allocation of Spitzer time ever granted (~5300h) and will be the targets of ultra-deep 1-2 μm photometry and spectroscopy by the Euclid mission scheduled to launch in early 2021. The H20 data will permit immediate science investigations at z < 7 using the HSC+Spitzer data and will give Hawaii priority for z > 7 science once Euclid launches.

The proposed program spans 3 years (6 semesters). The first two years consists of 30 nights of HSC (grizy) imaging to cover both fields. In the final year of the program we request 5 nights of Keck-LRIS and 5 nights of Keck-MOSFIRE time to obtain spectroscopy of high-redshift sources selected from the Imaging Survey. The spectroscopy will be critical for verifying the photometric redshifts, for characterizing rare and interesting sources, and for selecting targets for follow-up with ALMA and JWST. The combined HSC+Spitzer imaging data will provide the largest and only cosmologically significant sample of z > 3 galaxies with stellar mass estimates for at least the next decade. These data will allow us to provide the first definitive constraint on the evolution of massive (M/M > 1010.5) galaxies at 3 < z < 7, estimate the properties of their dark matter haloes through clustering measurements, and find rare "Rosetta stone" objects that are re-ionzing the universe during this key epoch of galaxy evolution.

The Keck spectroscopy will also be used to confirm redshifts and characterize rare/interesting objects, proto-clusters, and quasars as well as measure the strength of the Ly-α, [Oii], and other lines to search for high-escape fraction objects. In addition, any remaining open slits on the masks will be used to obtain spectra for a fully representative sample of galaxies at 3 < z < 7 using methods previously pioneered by members of our group. The H20 Survey will allow all Hawaii graduate students, a TBD Hawaii postdoc, and the Hawaii PIs of this project immediate access to Euclid 0.5-2 μm data when the mission starts in early 2021 (immediately following the completion date of H20). At that point, a set of unprecedented z > 7 galaxy studies will start, with Hawaii having priority for this science based on the data obtained in the H2O Survey.



  • Probe the assembly of massive galaxies by tracing the evolution of their stellar mass function with >500,000 galaxies at 3 < z < 6 and including several thousand at z > 6.
  • Link the stellar mass of galaxies to their dark matter halos via clustering, for testing of cosmological models.
  • Constrain the properties of dark energy using the non-linear power spectrum at high-redshift.
  • Identify the earliest sites of galaxy formation and reionization, and spectroscopically confirm high-escape fraction candidates.
  • Identify and confirm the first galaxies to quench their star formation.

Proposed Research

Click on an image to read the caption
Dark matter density map at 4.3 < z < 5.3 over 20 deg<sup>2</sup> , from the Millennium Simulation (Springel et al. 2005), while small rectangle and circle are comparable to CANDELS and COSMOS respectively. Only H20 has the statistical power to study the rare overdensity peaks (dark orange) and cosmic voids (dark purple) as well as characterizing the overall density field needed for cosmology.
Galaxy stellar mass functions with 1σ statistical errors from existing (hatched) and proposed (solid) data. H20 will improve the constraints by a factor of > 10, thus allowing us to make a more definitive measurement of the overall mass function, link it to the dark matter via clustering, and characterize differences as a function of local environment.
Three proposed models of the galaxy SMF at z~6 are shown, along with current data (shaded area, Grazian et al. 2015) and proposed constraints (red circles, with error bars including expected uncertainties from Poisson noise, cosmic variance, and SED fitting). The three models (solid, dashed, dotted lines) have radically different implications for galaxy evolution in the early universe (e.g. Davidzon et al. 2017). H20 will clearly differentiate between these models. When combined with the proposed clustering measurements it will also directly measure the duty cycle of star formation (e.g. determining typical star formation histories).
Sensitivity limits in the HSC filters of this proposal (blue), along with the IRAC ch. 1 and 2 from our ongoing SLS program (red) and the designed NIR filters of Euclid (green). Light and dark grey lines show spectral energy distributions (SEDs) of two galaxies at z ~ 2.5 and 7 respectively, which the deep H20 imaging will be able to disentangle. To confirm and characterize this differentiation the proposed Keck follow-up is essential.
 Stellar-to-halo mass ratio (SHMR) at z~5 from state-of-the-art analyses. The 2 deg2 of the COSMSOS field (blue line, Coupon et al. in prep.) still suffer large statistical uncertainties at z ~ 5 (shaded area). HSC-SSP Wide estimates (Harikane et al. 2016) are apparently more precise, but they rely on MUV converted into stellar mass through average ad-hoc assumptions. Both samples cannot constrain the most massive regime while in the cosmic volume probed by H20 we expect to find at least ~ 40 halos > 7 × 10<sup>12</sup> M<sub>⊙</sub>. Moreover our data will consolidate the SHMR at M<sub>halo</sub> ~ 10<sup>12</sup> M<sub>⊙</sub>, to pin down the peak of efficiency and determine weather it evolves from z ~ 0 (dashed line) to z = 4-6.
 SED fitting to our photometric baseline (g, r, i, z, y, [3.6], [4.5]) recovers the redshift of simulated galaxies with an error (normalized median absolute deviation) of σz /(1 + z) . 0.03 at z < 1.5 and z > 3, with the 1.5 < z < 3 range reaching similar performance once Eculid data are avaialable. The fraction of outliers (Δz > 0.15 σz ) is expected to be < 10%; the actual value will be quantified using our spectroscopic sample.
Difference between the intrinsic stellar mass of 3 < z < 6 galaxies and those recovered with SED fitting. When IRAC data are not available estimates have over 1.5 dex of scatter, making impossible to constrain stellar mass assembly as a function of cosmic time. The HSC-SSP survey lacks sufficiently deep IRAC data (> 0.5 dex scatter bellow M<sub>☆</sub>) except in the two 1.8 deg ultra-deep fields which are too small for the proposed science and suffer from large cosmic variance.

Synergy with other surveys

Proof of Concept Exercise

High-z Dropouts Selection

To confirm the ability of the Hawaii Two-0 (20 sq. deg.) survey (as designed) to robustly identify high-redshift galaxies (z≳4) using our H20 dropout selection technique, we used the UH+SSP HSC data in the COSMOS (2 sq. deg.) field to select high redshift galaxy candidates.

Note that r-band dropouts are sources that "just begin to appear" in the r-band, are clearly absent in the g-band, and robustly detected in the i-band. The above figure clearly illustrates the potential for H20 observations to detect high-redshift dropout galaxies.

Hawaii Two-0 Predictions

Band Redshift Detected Sources
COSMOS 2 sq. deg
Expected Sources
H20 20 sq. deg
g ~4 72,098 720,000
r ~5 3,814 38,000
i ~6 262 2,600
z ~7 46 460


Due to earthquake delays, our first H20 data are not expected to arrive before early January 2019.

Original Observing Plan

2018A — NEP

  • 4 nights Subaru HSC
  • 2018B — CDF-S

  • 5 nights Subaru HSC
  • 2019A — NEP

  • 5 nights Subaru HSC
  • 1 night Keck LRIS
  • 1 night Keck MOSFIRE
  • 2019B — CDF-S

  • 5 nights Subaru HSC
  • 1 night Keck LRIS
  • 1 night Keck MOSFIRE
  • 2020A — NEP

  • 5 nights Subaru HSC
  • 2 night Keck LRIS
  • 1 night Keck MOSFIRE
  • 2020B — CDF-S

  • 5 nights Subaru HSC
  • 1 night Keck LRIS
  • 2 night Keck MOSFIRE
  • Subaru HSC Imaging Depths

    g r i z y J,H,K ch.1 ch.2
    27.5 27.5 27 26.5 26 26 24.8 24.7

    ☆ From Euclid Deep (operations start in 2021). WFIRST will also cover this area at 26 mag 10σ (operations start mid-2020s).
    † From the Euclid/WFIRST Spitzer Legacy Survey (ADS)

    Data Reduction Plan

    Co-I Chambers will ensure the team has access to the latest Pan-STARRS data for astrometric and flux calibration. The work of generating the mosaics will be done by Co-I Repp and a TBD Hawaii graduate student funded by the FIELDS grant. Computation for this work is available from several sources, namely the supercomputing facility at UH, the Euclid Science Data Centre at IPAC, and compute clusters at UCR. Co-I Capak has participated in the commissioning of the HSC pipeline and is well aware of its current limitations. We have a "work-around" pipeline in place that does the basic de-trending with the HSC pipeline then uses the Terapix tools (SExtractor, SWARP, PSFex, etc.) to mask and remove artefacts, model and homogenize the PSF and measures science quality photometry suitable for the proposed work. IRAC photometry will be measured using IRClean that was successfully applied in COSMOS and SPLASH Laigle et al. (2016).


    Due to earthquake delays, our first H20 data are not expected to arrive before early January 2019.

    Allocated Observing Runs

    Subaru HSC — Lost 4 nights due to earthquakes
    Subaru HSC — Lost 3 nights due to earthquakes
            2 allocated nights remaining (Jan 1&2)
    Subaru HSC — 4 nights awarded, May 29&30 (2nd half), July 26—30 (1st half)

    Data Reduction

    NEP HSC Photometery


    CDF-S HSC Photometery


    Keck Spectrscopy



    Semester 1 (S18A)

      Kick-off meeting to decide observing strategy for Hawaii Two-0 project
      Consult with IfA transient experts on optimizing survey strategy for supernova detection
      Prepare Spitzer SLS data in order to construct full SEDs with HSC data
      Setup HSC pipeline on UH Cray
      Secure funding for two IfA grad students to work on H20 program
      Preliminary image reduction and catalog
    Prepare and execute first Subaru HSC observations in NEP
    Preliminary NEP photo-z catalog including stellar masses
    First selection of high-z candidates in NEP for follow-up spectroscopy

    Semester 2 (S18B)

      Present IfA talk on first semester progress and data availability
      Confirm depth requirements of the H20 Observations using COSMOS HSC data
      Identify z=4-7 dropouts in COSMOS HSC data to estimate H20 final outcomes
      Finalize Spitzer data reduction
    Prepare and execute first Subaru HSC observations in NEP
    Preliminary NEP photo-z catalog including stellar masses
    First selection of high-z candidates in NEP for follow-up spectroscopy
    Finalize first data release for NEP
     Prepare and execute first Subaru HSC observations in CDFS
     Produce preliminary stacked images and photometric catalogs for CDFS
     Produce preliminary photo-z and stellar mass catalog for CDFS
     First selection of high-z candidates in CDFS for follow-up spectroscopy

    Semester 3 (S19A)

      Prepare and execute first Subaru HSC observations in NEP
      Preliminary NEP photo-z catalog including stellar masses
      First selection of high-z candidates in NEP for follow-up spectroscopy
     Execute first round of CDFS spectroscopic follow-up
     Finalize first data release for CDFS
     Prepare updated images and catalogs for NEP

    Semester 4

      Finalize first data release for NEP
     Continue spectroscopic follow-up
     Finalize second data release for NEP
     Prepare updated data release for CDFS

    Semester 5

     Continue spectroscopic follow-up
     Finalize second data release for CDFS
     Prepare final release candidate for NEP images and catalog

    Semester 6

     Conduct final spectroscopic follow-up
     Prepare final release candidate for CDFS images and catalog

    Semester 6 + 1 yr

     Produce final public data release of NEP and CDFS


    All H20 data will be hosted through this website. The links above will be made active once data collection has begun.

    IfA astronomers are welcome to use any of the survey data.
    However, please notify one of the PIs before publishing or distributing the data.



    Dave Sanders

    Co-PI, IfA



    Istvan Szapudi

    Co-PI, IfA



    Josh Barnes

    Co-PI, IfA



    Ken Chambers

    Co-PI, IfA



    Andrew Repp

    Co-I, IfA


    Conor McPartland

    Co-I, IfA




    Peter Capak

    Co-I, IPAC/Caltech


    Bahram Mobasher

    Co-I, UC Riverside


    Sune Toft

    Co-I, DAWN/University of Copenhagen


    Konrad Kuijken

    Co-I, Leiden Observatory, Euclid Consortium Complementary Observations Group lead