CANDELS: The Cosmic Assembly Near-infrared Deep Extragalactic Legacy Survey

TL;DR

CANDELS addresses how galaxies assemble over the first half of cosmic time by combining a wide area (Wide) and deep (Deep) HST survey across five fields, enabling large, statistically robust samples of high-redshift galaxies and AGN. The study integrates theory-driven mock catalogs with multi-wavelength data (HST optical/NIR, Spitzer IRAC, and others) and a SN Ia component to probe dark energy evolution, galaxy morphology, and the growth of structure from z ≈ 8 to 1.5. The three-tier wedding-cake design, with careful management of cosmic variance and a coordinated observing strategy (including UVIS, grism, and SN-follow-up), yields transformative constraints on the UV luminosity function, stellar masses, star-formation histories, and the co-evolution of galaxies and black holes, while providing a rich public data resource for the community. By pushing rest-frame optical morphology measurements to high redshifts and extending SN Ia observations into the dark-matter-dominated era, CANDELS advances our understanding of galaxy assembly, AGN activity, and the expansion history of the Universe.

Abstract

The Cosmic Assembly Near-infrared Deep Extragalactic Legacy Survey (CANDELS) is designed to document the first third of galactic evolution, over the approximate redshift (z) range 8--1.5. It will image >250,000 distant galaxies using three separate cameras on the Hubble Space Telescope, from the mid-ultraviolet to the near-infrared, and will find and measure Type Ia supernovae at z>1.5 to test their accuracy as standardizable candles for cosmology. Five premier multi-wavelength sky regions are selected, each with extensive ancillary data. The use of five widely separated fields mitigates cosmic variance and yields statistically robust and complete samples of galaxies down to a stellar mass of 10^9 M_\odot to z \approx 2, reaching the knee of the ultraviolet luminosity function (UVLF) of galaxies to z \approx 8. The survey covers approximately 800 arcmin^2 and is divided into two parts. The CANDELS/Deep survey (5σ point-source limit H=27.7 mag) covers \sim 125 arcmin^2 within GOODS-N and GOODS-S. The CANDELS/Wide survey includes GOODS and three additional fields (EGS, COSMOS, and UDS) and covers the full area to a 5σ point-source limit of H \gtrsim 27.0 mag. Together with the Hubble Ultra Deep Fields, the strategy creates a three-tiered "wedding cake" approach that has proven efficient for extragalactic surveys. Data from the survey are nonproprietary and are useful for a wide variety of science investigations. In this paper, we describe the basic motivations for the survey, the CANDELS team science goals and the resulting observational requirements, the field selection and geometry, and the observing design. The Hubble data processing and products are described in a companion paper.

Figures (23)

• Figure 1: Four-orbit images of HUDF galaxies from ACS vs. 2-orbit images from WFC3/IR illustrate the importance of WFC3/IR for studying distant galaxy structure. WFC3/IR unveils the true stellar mass distributions of these galaxies unbiased by young stars and obscuring dust. The new structures that emerge in many cases inspire revised interpretations of these objects, as indicated.
• Figure 2: Limiting magnitudes of CANDELS photometry compared to existing photometry and to the SEDs of model blue and red galaxies. HST observations are shown in two-toned triangles. Each triangle encodes three brightness limits. The bottoms of the light-colored inverted triangles are $5\sigma$ point-source limits for Deep data. The bottoms of the darker triangles show the approximate $5\sigma$ limits for aperture photometry of a fiducial Lyman-break galaxy with a half-light radius of $0.25"$, also in Deep data. The tops of the triangles show the approximate $5\sigma$ limits for LBGs in the CANDELS/Wide survey using the exposure times from Table \ref{['tab:areadepth']} (except for F814W, see below). The green triangles at left show the GOODS/ACS data ($BViz$), the red triangles are for WFC3/IR ($YJH$). The yellow triangles denote CANDELS F814W, with the triangle bottoms corresponding to the fiducial 28 ks exposure in CANDELS/Deep and the top corresponding to a 1-orbit exposure. The points of the solid green triangles in the IR show point-source depths (valid for distant galaxies) for the GOODS VLT ISAAC $JHK$ observations 2010AA...511A..50R and the GOODS Spitzer IRAC data at 3.6$\micron$ and 4.5$\micron$2003mglh.conf..324D. The blue SEDs are for $L^*$ Lyman-break galaxies at redshifts $z=3,5,$ and 7; the red SED shows a maximally old galaxy at $z=3$ with rest-frame $M_V = -21$ mag. For comparison, we also plot the $5\sigma$ sensitivities of the deepest existing HST images (gray triangles), covering the much smaller areas of the HUDF ($BViz$) and the HUDF09 ($YJH$) as reported by 2011arXiv1105.2297O.
• Figure 3: Clumpy $z \approx 2$ galaxies from WFC3 Early Release Science (ERS) images taken in F160W (2-orbit depth) versus F775W (3-orbit depth). Object IDs are from the FIREWORKS 2008ApJ...682..985W and MUSYC 2010ApJS..189..270C catalogs. Galaxies are arranged in two groups. On the left are objects for which the two images look quite similar, while objects on the right look substantially different. Cases like W4934, W4973, and W5609 on the right appear to have underlying disks composed of older stars and may be candidates for inherently regular galaxies in which clump formation has occurred via in situ disk instabilities. Cases on the left like W4032, W5201, and W5353 lack such underlying disks and are more likely to be separate galaxies now merging. F160W also highlights the presence and location of central potential wells, as in W4934, W5160, and W5229.
• Figure 4: Top panels show simulated rest-frame $urz$ images of a gas-rich major merger. The image was created using a hydrodynamic simulation of a binary galaxy merger including star formation, black hole accretion, and AGN feedback 2005Natur.433..604D. The simulation results were post-processed through the dust and radiative transfer code Sunrise 2010MNRAS.403...17J. Successive merger stages are labeled (1--6). The middle and bottom panels show the predicted IR luminosity and rest-frame optical asymmetry vs. merger stage for the same simulation. This picture predicts that the maximal asymmetry should be measured during the relatively early stages of the merger, while the object would be identified as an obscured or optical QSO in the late stages, when the two nuclei may have largely coalesced and faint tidal features may be the only remaining signs of morphological disturbance (see also 2008ApJS..175..356H).
• Figure 5: A simplified footprint of the CANDELS observations in the GOODS-S field with WFC3/IR. The "Wide" portion of the CANDELS observations (green) and the pre-existing observations of the WFC3 ERS (brown) are reproduced faithfully. However the "Deep" portion of the field, represented here as a $3\times5$ raster (blue), will in practice be observed across several epochs with slightly varying geometries. The individual epochs, as well as the footprints of the ACS parallel observations, are illustrated in Figures \ref{['fig:gdsepa']} and \ref{['fig:gdsepb']}.