Next year, the James Webb Space Telescope (JWST) will devote 206 hours to observe in infrared light one third of the COSMOS field: covering an area larger than the full Moon, this region of the sky was first surveyed by the Hubble Space Telescope and has since been observed by every major telescope both on Earth and in space. The COSMOS field has therefore become an essential reference for studying the evolution of galaxies and large-scale structure of the Universe over cosmic time. Thanks to its large area on the sky, COSMOS-Webb will allow astronomers to see for the first time the distribution of the very first galaxies in the Universe, and will lead to a gold mine of targets for future observations with JWST and other telescopes.

At the beginning of the 20th century, astronomers realized for the first time that distribution of stars in the first sky surveys could provide clues to the shape of our own Milky Way galaxy. Over the next few years, powerful telescopes (such as the Palomar 100” in California) showed that galaxies themselves were spread out across the sky in a pattern that holds profound clues to the nature and eventual fate of the Universe itself. This is the history of astronomy: with each new instrument, astronomers take a step towards a better understanding of our Universe (for an example of what modern measurements and analysis of the large-scale distribution of galaxies can tell us, see “The origin of the large-scale structure of the Universe and dark matter flows in our cosmic neighborhood”).

Next year, using the soon-to-be-launched James Webb Space Telescope (JWST), an international team of astronomers will for the first time map with highly sensitive infrared cameras a vast unexplored region of the very distant Universe: COSMOS-Webb. A key prediction of current theories of the formation of the Universe is that stars and galaxies form inside hot bubbles of ionized hydrogen gas (hydrogen atoms are stripped of their electrons). The distribution of galaxies in these images from JWST will allow astronomers to see for the first time the imprint of these bubbles on the large-scale distribution of these very first stars and galaxies. Comparing the numbers of these distant galaxies with predictions from computer simulations will allow astrophysicists to understand how galaxies form in the very young Universe. And the maps, which will be made freely available to all astronomers, will provide an essential roadmap for future observations.

Composite image of the COSMOS field compared to the size of the Moon (half a degree). Figure 1: Composite image of the COSMOS field compared to the size of the Moon (half a degree). Credits: COSMOS team.

Over the past three decades, the Hubble Space Telescope (HST) has opened up vistas of the distant Universe only dimly glimpsed by ground-based observatories. However, charting out how galaxies are linked together across the sky seemed at first like an impossible task for HST: the telescope's cameras can only see a tiny part of the sky in each exposure (corresponding to the apparent size of a few craters on the Moon). But a team of astronomers ambitiously used HST to survey the COSMOS field, a region of the sky larger than the full moon (see Fig. 1), making it the largest ever contiguous patch of the sky covered by HST. Since then, thousands of scientific papers have delved deep into the COSMOS field, drawing a detailed picture of how galaxies, dark matter, gas and dust interact on large scales and how they have changed over cosmic time. Since those first HST observations, the COSMOS field has been covered by every major telescope on Earth and in space, and these data have been shared in a remarkably open and productive collaboration between many of the world's astrophysicists.

French researchers have played a key role in COSMOS since the project started in 2003, with the continued support of the Centre National d'Etudes Spatiales (CNES). The VIMOS instrument constructed at the Laboratoire d'Astrophysique de Marseille (LAM) and installed at the European Southern Observatory’s (ESO) Very Large Telescope (VLT) measured precise distances for tens of thousands of galaxies in COSMOS. The Megacam and WIRCAM optical and infrared wide-field cameras at the Canada-France-Hawaï Telescope (CFHT) provided early images of the COSMOS field. The creation of the last two COSMOS public catalogues have been led by the IAP and the LAM. In the beginning, wide-field images obtained at different wavelengths were processed at the TERAPIX center at the IAP; they are now processed using the new CANDIDE data center at the IAP, a project dedicated to optical and infrared data in wide-field surveys in support of the Euclid project. From these images, teams at LAM and IAP have created the largest and deepest catalogue of “photometric redshifts”: starting from a small sample of very precise distance measurements, this technique uses brightnesses of galaxies at different wavelength to estimate distances for all galaxies in the COSMOS field, more than a million objects for the most recent catalogues.

The COSMOS catalogues have since become an essential reference point for studying the evolution of galaxies over cosmic time and to ask many new questions about the Universe. The scientific articles describing them are amongst the most highly-cited in astronomy. Articles led by French researchers describe how the masses of galaxies have built up across cosmic time, and how galaxies are distributed on the largest scales. In recent years, the COSMOS catalogues have also played a crucial role in preparing for European Space Agency's (ESA) Euclid mission by providing a detailed census of the numbers and types of distant galaxies. The original COSMOS HST images have also been crucial in making realistic simulations of what the Euclid satellite will see.

Of course, astronomers would always like to see further. Data from the COSMOS field have hinted at very distant, massive galaxies out of reach of most current surveys. The finite speed of light is a time-machine which means that the most distant objects lie at the very beginnings of the Universe. But the expanding Universe means their light becomes redder and redder as it speeds towards us. To see so very far back, one needs a telescope with not only a mirror large enough to collect the rare photons arriving from the early epochs of the Universe, but also one that is sensitive to the infrared light emitted by these distant galaxies. An even greater challenge is that every ordinary object on Earth also emits infrared photons; these would drown out the faint signal from such distant galaxies. To detect this ancient light, the telescope needs be cooled to a few degrees above the absolute zero and placed far from terrestrial sources. The ten-billion-dollar JWST, with its golden 6.5-meter foldable mirror, is optimized for the study of these faint distant objects. It is due to be launched in October 2021 and sent to one of the coldest places in the solar system, the second Lagrange point. A single exposure from JWST will see deeper than the decade-long ground based infrared surveys carried out within the COSMOS field.

JWST is a remarkable instrument, perhaps the most sensitive telescope ever constructed. Its cameras can record light from objects infinitely fainter than anything that had ever been observed before. But where does one point such a state-of-the-art and sensitive observatory, if it is looking deep into a country that no one has ever visited before? The best thing to do, right at the start, is to survey a wide swath of this new territory to guide the astronomers. As with HST, such a survey seemed an impossible task with the tiny field-of-view of the JWST cameras. Nevertheless, the COSMOS team proposed an ambitious program, COSMOS-Webb, to cover one third of the COSMOS field with JWST’s infrared and near-infrared cameras. Such a survey requires more than 200 hours of observations, an enormous amount for the highly competitive first year of JWST’s operations: astronomers around the world requested more than four times the amount of time than was actually available. The results of this competition were announced in March 2021: the COSMOS-Webb proposal has been accepted, making it the largest survey to be carried out during the first year of JWST’s observations.

The program comprises 206 hours with JWST's NIRCam instrument and 80 hours taken simultaneously with the longer-wavelength MIRI camera. From these observations, the COSMOS-Webb program will produce a new publicly available map of the very distant Universe. This map will contain hundreds of thousands of very faint and distant galaxies, signposts for follow-up observations with JWST’s other very sensitive instruments, as well as a precise census of the early Universe. Astronomers will see for the first time the structure and form of galaxies in the early Universe. These observations may also help them to understand how the mysterious dark matter helped galaxies to form at the earliest times. Moreover, these observations will produce a new generation of public COSMOS catalogues which probably lead to discoveries not imagined today.


puce The Astrophysical Journal Supplement Series article: Scoville et al. (2007), “The Cosmic Evolution Survey (COSMOS): Overview”

puce American Astronomical Society meeting article: Weaver et al. (2021), “COSMOS2020: A next-generation catalog to explore the 1 < z < 8 Universe”

Writing and contact

Other contact

Web writing: Valérie de Lapparent
Layout and iconography: Jean Mouette

May 2021

Institut d'Astrophysique de Paris - 98 bis boulevard Arago - 75014 Paris