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 to z ≈ 2, reaching the knee of the ultraviolet luminosity function of galaxies to z ≈ 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 ∼125 arcmin 2 within Great Observatories Origins Deep Survey (GOODS)-N and GOODS-S. The CANDELS/Wide survey includes GOODS and three additional fields (Extended Groth Strip, COSMOS, and Ultra-deep Survey) and covers the full area to a 5σ pointsource limit of H 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.

This paper describes the Hubble Space Telescope imaging data products and data reduction procedures for the Cosmic Assembly Near-IR Deep Extragalactic Legacy Survey (CANDELS). This survey is designed to document the evolution of galaxies and black holes at z ∼ 1.5 − 8, and to study Type Ia SNe beyond z > 1.5. Five premier multi-wavelength sky regions are selected, each with extensive multiwavelength observations. The primary CANDELS data consist of imaging obtained in the Wide Field Camera 3 / infrared channel (WFC3/IR) and UVIS channel, along with the Advanced Camera for Surveys (ACS). The CANDELS/Deep survey covers ∼ 125 square arcminutes within GOODS-N and GOODS-S, while the remainder consists of the CANDELS/Wide survey, achieving a total of ∼ 800 square arcminutes across GOODS and three additional fields (EGS, COSMOS, and UDS). We summarize the observational aspects of the survey as motivated by the scientific goals and present a detailed description of the data reduction procedures and products from the survey. Our data reduction methods utilize the most up to date calibration files and image combination procedures. We have paid special attention to correcting a range of instrumental effects, including CTE degradation for ACS, removal of electronic bias-striping present in ACS data after SM4, and persistence effects and other artifacts in WFC3/IR. For each field, we release mosaics for individual epochs and eventual mosaics containing data from all epochs combined, to facilitate photometric variability studies and the deepest possible photometry. A more detailed overview of the science goals and observational design of the survey are presented in a companion paper.

We analyze star formation (SF) as a function of stellar mass (M ⋆ ) and redshift z in the All Wavelength Extended Groth Strip International Survey (AEGIS). For 2905 field galaxies, complete to 10 10 (10 10.8 )M ⊙ at z < 0.7(1), with Keck spectroscopic redshifts out to z = 1.1, we compile SF rates (SFR) from emission lines, GALEX, and Spitzer MIPS 24µm photometry, optical-NIR M ⋆ measurements, and HST morphologies. Galaxies with reliable signs of SF form a distinct "main sequence (MS)", with a limited range of SFR at a given M ⋆ and z (1σ ±0.3 dex), and log(SFR) approximately proportional to log(M ⋆ ). The range of log(SFR) remains constant to z > 1, while the MS as a whole moves to higher SFR as z increases. The range of SFR along the MS constrains the amplitude of episodic variations of SF, and the effect of mergers on SFR. Typical galaxies spend ∼ 67(95)% of their lifetime since z = 1 within a factor of 2(4) of their average SFR at a given M ⋆ and z. The dominant mode of the evolution of SF since z ∼ 1 is apparently a gradual decline of the average SFR in most individual galaxies, not a decreasing frequency of starburst episodes, or a decreasing factor by which SFR are enhanced in starbursts. LIRGs at z ∼ 1 seem to mostly reflect the high SFR typical for massive galaxies at that epoch. The smooth MS may reflect that the same set of few physical processes governs star formation prior to additional quenching processes. A gradual process like gas exhaustion may play a dominant role.

The DEEP2 and COMBO-17 surveys are compared to study luminosity functions of red and blue galaxies to z $ 1. The two surveys have different methods and sensitivities, but nevertheless results agree. After z $ 1, M à B has dimmed by 1.2Y1.3 mag for all colors of galaxies, à for blue galaxies has hardly changed, and à for red galaxies has at least doubled (our formal value is $0.5 dex). Luminosity density j B has fallen by 0.6 dex for blue galaxies but has remained nearly constant for red galaxies. These results imply that the number and total stellar mass of blue galaxies have been substantially constant since z $ 1, whereas those of red galaxies (near L à ) have been significantly rising. To explain the new red galaxies, a ''mixed'' scenario is proposed in which star formation in blue cloud galaxies is quenched, causing them to migrate to the red sequence, where they merge further in a small number of stellar mergers. This mixed scenario matches the local boxy-disky transition for nearby ellipticals, as well as red sequence stellar population scaling laws such as the color-magnitude and Mg-relations (which are explained as fossil relics from blue progenitors). Blue galaxies enter the red sequence via different quenching modes, each of which peaks at a different characteristic mass and time. The red sequence therefore likely builds up in different ways at different times and masses, and the concept of a single process that is ''downsizing'' (or upsizing) probably does not apply. Our claim in this paper of a rise in the number of red galaxies applies to galaxies near L à . Accurate counts of brighter galaxies on the steep part of the Schechter function require more accurate photometry than is currently available.

We characterize the mass-dependent evolution in a large sample of more than 8,000 galaxies using spectroscopic redshifts drawn from the DEEP2 Galaxy Redshift Survey in the range 0.4 < z < 1.4 and stellar masses calculated from K-band photometry obtained at Palomar Observatory. This sample spans more than 1.5 square degrees in four independent fields. Using restframe (U − B) color and [OII] equivalent widths, we distinguish star-forming from passive populations in order to explore the nature of "downsizing"-a pattern in which the sites of active star formation shift from high mass galaxies at early times to lower mass systems at later epochs. Over the redshift range probed, we identify a mass limit, M Q , above which star formation appears to be quenched. The physical mechanisms responsible for downsizing can thus be empirically quantified by charting the evolution in this threshold mass. We find that M Q decreases with time by a factor of ≈3 across the redshift range sampled according to M Q ∝ (1 + z) 3.5 . We demonstrate that this behavior is quite robust to the use of various indicators of star formation activity, including morphological type. To further constrain possible quenching mechanisms, we investigate how this downsizing signal depends on local galaxy environment using the projected 3 rd -nearest-neighbor statistic D p,3 which is particularly well-suited for large spectroscopic samples. For the majority of galaxies in regions near the median density, there is no significant correlation between downsizing and environment. However, a trend is observed in the comparison between more extreme environments that are more than 3 times overdense or underdense relative to the median. Here, we find that downsizing is accelerated in overdense regions which host higher numbers of massive, early-type galaxies and fewer late-types as compared to the underdense regions. Our results significantly constrain recent suggestions for the origin of downsizing and indicate that the process for quenching star formation must, primarily, be internally driven. By quantifying both the time and density dependence of downsizing, our survey provides a valuable benchmark for galaxy models incorporating baryon physics.