COSMOS-Web (PID: 1727) is a 255 hour wide-field Cycle 1 JWST treasury program that maps a contiguous 0.6 deg² area with deep NIRCam imaging in 4 filters (F115W, F150W, F277W, and F444W) and a non-contiguous 0.2 deg² area with MIRI in parallel. COSMOS-Web builds on the rich heritage of multiwavelength observations and data products available in the COSMOS field, and will contain about a million galaxies across cosmic time. For more information, see survey Overview Paper or contact the survey PIs: Jeyhan Kartaltepe (jeyhan [at] astro [dot] rit [dot] edu) and Caitlin Casey (cmcasey [at] utexas [dot] edu).

COSMOS-Web Primary Science Goals

The design of COSMOS-Web is motivated by three primary science goals:

1. Map Cosmic Reionization

COSMOS-Web will revolutionize our understanding of reionization's spatial distribution, environments, and drivers at early stages by detecting thousands of galaxies in the epoch of reionization (6<z<11) on scales large enough to mitigate cosmic variance.

Constraining the physics of reionization requires identifying and characterizing the galaxies that are embedded deep within the predominantly neutral Universe at z > 8. COSMOS-Web will grow the census of Epoch of Reionization (EoR, 6<z<11) galaxies beyond what is currently known pre-JWST by a factor of ~5 and quantify the evolution of the UV luminosity function (UVLF), stellar mass function (SMF), and star formation histories of galaxies across a wide redshift range. The large survey area will capture reionization on scales larger than its expected patchiness, eliminating the effect of cosmic variance, and forge the detection of hundreds of intrinsically bright galaxies at 8<z<11 embedded in the neutral IGM, which likely trace the highest-density peaks from which the reionization process was likely to begin.

2. Trace Massive Galaxy Evolution

COSMOS-Web will enable the identification and characterization of massive galaxies in the first two billion years, including the rarest quiescent galaxies at z > 4, and allow us to constrain the formation and evolution of early massive galaxies.

The growing census of quiescent galaxies at early epochs (M_* > 10¹⁰ M_⊙ out to z ~ 3–4) has presented a strong challenge to theoretical models. In order to build up their significant stellar masses and quench their star formation so early in the Universe’s history, these galaxies must have formed their stars at incredible rates (≫100 M_⊙ yr⁻¹ at very early times) and then abruptly shut down the production of stars well within the Universe’s first billion years. The quiescent galaxy luminosity function beyond z ~ 4 is unconstrained, though such galaxies are expected to be very rare and particularly difficult to separate from dusty star forming galaxies that can mimic the same red colors. COSMOS-Web will allow us to take the first census of massive galaxies from EoR to the peak of galaxy assembly, study the evolution of their morphologies and sizes, and distinguish between massive star-forming galaxies and the first massive quiescent systems.

3. Link Dark Matter to Visible Matter

We will be able to directly measure the evolution of the stellar mass to halo mass relation (SMHR) out to z ~ 2.5 and its variance with galaxies' star formation histories and morphologies.

The link between galaxies’ dark matter halos and their baryonic content (visible matter) is of fundamental importance to cosmology. It is thought that galaxies’ halo mass growth should be independent of their baryonic processes (such as star formation and quenching) and if measurable, could provide a direct path to constraining galaxy growth and quenching mechanisms. Obtaining direct measurements of halo masses not only helps us to constrain the astrophysics of galaxies but also provides independent measurements of cosmological parameters. Weak gravitational lensing is the only tool that can be used as a direct probe of halo masses for a large sample of galaxies across cosmic time. COSMOS-Web will enable these weak lensing measurements, allowing us to directly measure galaxies’ halo masses (and therefore, constrain the SMHR) out to z ~ 2.5 down to M_* > few ×10⁹ M_⊙ with weak lensing (down to 10⁸ M_⊙at z ~ 1).

Beyond these primary science driver’s, COSMOS-Web’s legacy value will extend to many subfields of extragalactic astronomy and beyond. We anticipate that COSMOS-Web will have potential impact on the detailed study of galaxy morphologies, using spatially resolved SEDs to measure galaxy properties, placing constraints on the dust attenuation law, identifying and characterizing galaxy protoclusters, finding strong gravitational lenses, identifying direct collapse black hole candidates, studying the co-evolution of supermassive black holes and their host galaxies, searching for z > 10 pair instability supernovae, and identifying ultracool sub-dwarf stars in the Milky Way’s halo.

Observing Strategy

The COSMOS-Web mosaic consists of 152 visits over 255 hours and will be taken over three epochs. The first epoch of data consists of six visits covering ∼77 arcmin² with NIRCam and was observed on 5-6 January 2023. Currently, 77 pointings are scheduled for April/May 2023 (roughly half the field), and the remaining 69 pointings in December 2023/January 2024.

The first epoch of COSMOS-Web NIRCam observations obtained on Jan. 5-6, 2023, including the F115W, F150W, F277W, and F444W filters as a color composite. These data cover six out of a total of 152 visits. The total area covered is ∼77 square arcmin. The relative position of this mosaic in the survey is shown at upper left. At lower left are several zoomed-in 10′′ × 10′′ cutouts and one 16′′ × 16′′ cutout showing specific galaxies selected from these first data. Image credit: COSMOS-Web / Kartaltepe / Casey / Franco / Larson / RIT / UT Austin / CANDIDE. Full resolution image.

COSMOS-Web will obtain near-infrared imaging using the NIRCam instrument in four filters, reaching depths determined by the number of NIRCam exposures at a given position (ranging 1-4 exposures). The four filters include two short wavelength bands, F115W and F150W, and two long wavelength bands, F277W and F444W.

*COSMOS-Web NIRCam depths in each filter*
NIRCam Exposures [#]	Exposure Time [s]	Short Wavelength Area [arcmin²]	F115W Depth [5 σ]	F150W Depth [5 σ]	Long Wavelength Area [arcmin²]	F277W Depth [5 σ]	F444W Depth [5 σ]
1	257.68	71.3	26.87	27.14	17.8	27.71	27.61
2	515.36	991.6	27.13	27.35	978.0	27.99	27.83
3	773.05	60.0	27.26	27.50	24.4	28.12	27.94
4	1030.73	805.2	27.45	27.66	904.3	28.28	28.17

The first epoch of COSMOS-Web MIRI observations obtained on Jan. 5-6, 2023. The MIRI data are distributed in six non-overlapping tiles and include data from both the MIRI imager and Lyot Coronograph field of view. At left is a comparison between Spitzer IRAC channel 4 (8μm) data and MIRI 7.7μm data in a 40′′ × 40′′ zoom-in panel. Image credit: COSMOS-Web / Kartaltepe / Casey / Harish / Liu / RIT / UT Austin / CANDIDE. Full resolution image.

COSMOS-Web will obtain mid-infrared imaging using the MIRI instrument in parallel in one filter, F770W. The depth in this filter varies by the number of MIRI exposures at a given position (ranging 2-8 exposures).

*COSMOS-Web MIRI F770W Depths*
MIRI Exposures [#]	Exposure Time [s]	Area Covered [arcmin²]	F770W Depth [5 σ]
2	527.26	80.5	25.33
4	1054.52	430.4	25.70
6	1581.77	30.8	25.76
8	2109.03	146.1	25.98

Community Downloads

Note that the Cycle 1 program, PRIMER (PID: 1837) observes the COSMOS-CANDELS region and is therefore embedded within COSMOS-Web.

Plans for Data Product Release

COSMOS-Web is being observed in three epochs (Epoch 1: Jan 5/6, 2023; Epoch 2: April/May 2023; Epoch 3: December 2023 / January 2024). Our tream is committed to the public release of initial mosaics 6 months after each epoch and catalogs after roughly one year.

Our first data was taken in January - see below for links to jpg images of these first data!

NIRCam: Full resolution image with and without logo

MIRI: Full resolution image with and without logo