travaux    Wiki restricted area  

regaldis logo   Recherche et étude des premières galaxies dans l’univers   ANR
Identification and study of the first galaxies in the universe

  Contents:   Coordinateur du Projet: Roser PELLO
Laboratoire d'Astrophysique de Toulouse-Tarbes
Observatoire MidiPyrénées
14 Avenue Edouard Belin
email roser [at]
Tél * 05 61 33 28 12
Fax 05 61 33 28 40

Responsables des noeuds:
Marseille: Jean-Gabriel CUBY
Paris: Yannick MELLIER
Toulouse: Roser PELLO

Participants au projet:
Frederic BOONE   Astronome Adjoint   Observatoire de Paris
Benjamin CLEMENT   thésard   Laboratoire d'Astrophysique de Marseille
Thierry CONTINI   CR1   Laboratoire d'Astrophysique de Toulouse-Tarbes
Jean-Gabriel CUBY   Astronome   Laboratoire d'Astrophysique de Marseille
Matthews HAYES   Post-Doc   Observatoire de Genève
Patrick HUDELOT   IR   Institut d'Astrophysique de Paris
Jean-Paul KNEIB   CR1   Laboratoire d'Astrophysique de Marseille
Nicolas LAPORTE   Thésard  Laboratoire d'Astrophysique de Toulouse-Tarbes
Jean-François LE BORGNE   Astronome   Laboratoire d'Astrophysique de Toulouse-Tarbes
Alexandre MAIZY   Thésard  Laboratoire d'Astrophysique de Toulouse-Tarbes
Henry McCRACKEN   Astronome Adjoint   Institut d'Astrophysique de Paris
Yannick MELLIER   Astronome   Institut d'Astrophysique de Paris
Roser PELLO   Astronome   Laboratoire d'Astrophysique de Toulouse-Tarbes
Johan RICHARD   Post-doc   Durham
Daniel SCHAERER   CR1   Laboratoire d'Astrophysique de Toulouse-Tarbes
        (Actuellement en détachement à Observatoire de Genève)

Scientific Rationale

Constraining the abundance and the properties of the first galaxies is an important challenge for modern cosmology. Distant star-forming sources at redshifts z~7-12 could have been responsible for a significant part of the cosmic reionization. Detailed studies of these “primordial” systems require the use of new ground-based and space facilities becoming available between 2008 and 2009 (e.g. Herschel, VLT/Hawk-I, GTC/EMIR, HST/ WFC3), in preparation for future surveys with ALMA. The goal of our project is precisely to take advantage of our privileged access to these key facilities (recently granted or guaranteed time) to build up the first representative spectroscopic sample of galaxies at redshift z>7, together with an additional sample of candidates based on photometric redshifts and wide multi-wavelength coverage.

The method proposed is similar to the one successfully developed in our pilot program with ISAAC/VLT (Pello et al. 2004, Richard et al. 2006) and HST/ACS+NICMOS (Kneib et al 2004, Richard et al 2008), using lensing clusters as natural gravitational telescopes. We propose here a systematic exploration of this new observationally unknown territory. Multi-wavelength coverage should allow us to determine the physical properties of extremely distant sources, such as star formation rate, extinction, stellar population, and possibly also metallicity and IMF. Various other inferences (e.g. on luminosity functions and luminosity density) will also be made from the pure photometric observations of high-z galaxy candidates. The expected results, probing the Universe at ages of ~ 450-800 Myr, will have profound implications on our knowledge of distant possibly even primeval galaxies, galaxy formation, cosmological reionization, and the early Universe.

New deep imaging data of blank fields and lensing clusters will be used, coming from our observing programs recently accepted with Spitzer, Herschel, LABOCA/APEX, CFHT/WIRCAM (WUDS Survey) and VLT/HAWK-I observations. We have also access to Chandra X-ray observations, in addition to a large database, including HST images, from our own observing programs and archives. The project also benefits from the new generation of near-IR spectrographs (Flamingos2/Gemini-S and EMIR/GTC) for the precise redshift determination and subsequent emission-line studies.

Galaxy candidates at z>7 will be selected from deep optical and near-IR imaging. Other multi-wavelength data (including X-rays, mid and far IR, mm bands) will be used to further characterize them. Strong lensing clusters are more efficient than present blank fields in identifying more distant sources (thanks to the gravitational magnification) and for detailed studies of galaxies in the z~7-12 domain. However, both blank and lensing fields are needed in order to constrain the complete Luminosity Function, allowing us to achieve a complete view of the first star-forming galaxies. Present day ultra-deep surveys (either blank or lensing fields) are dramatically small in terms of effective surface. For this reason, increasing the number of lensing and blank fields with ultra-deep near-IR photometry, as proposed here, is essential to get tighter constraints on the abundance and physical properties of z>7 starburst galaxies. Gathering a representative sample of galaxies at z~7-12 is also of great interest for future observations of primeval galaxies with ALMA.

International Background

Considerable advances have been made during the last years in the exploration of the early Universe with the discovery of galaxies at z~6-7, close to the end of reionization epoch (e.g. Hu et al. 2002, Kodaira et al. 2003, Cuby et al. 2003, Kneib et al. 2004, Stanway et al. 2004, Bouwens et al. 2004, 2008a, 2008b, Iye et al. 2006, Stark et al. 2007, Bradley et al. 2008). Extending the searches beyond z~ 6.5 and back to ages where the Universe was being re-ionized (cf. Fan et al. 2002) requires extremely deep observations in the near-IR bands. Indeed, astounding depths can be reached in ultra-deep fields, such as demonstrated with the NICMOS imaging of the UDF (Thompson et al. 2005; Bouwens et al. 2005, 2006, 2008a) from which faint (H(AB)~27) candidates at z~7-10 have been identified.

Our collaboration conducted a pilot program based on ESO/VLT observations in the near-IR, a deep survey of two lensing clusters to find star-forming galaxies at z~6-12 (see Schaerer et al. 2006). This project is based on the photometric pre-selection of candidates making use of the natural magnification provided by foreground lensing clusters. This technique, also first referred to as the ``gravitational telescope'' by Zwicky (1937), has proven highly successful in identifying a large fraction of the most distant galaxies known today thanks to magnifications by typically 1-3 magnitudes (e.g. Ellis et al. 2001, Hu et al. 2002, Kneib et al. 2004, Bradley et al. 2008). In this context of international competition, our team is one of the pioneers on the use of the ``gravitational telescope'' (hereafter GT) for the exploration of faint and distant galaxies, and the photometric selection of galaxies at high-redshift (e.g. Pello et al. 1999; Bolzonella et al. 2000; Kneib et al. 2004; Pello et al. 2004ab). We have started a systematic multi-wavelength follow up of our sample of lensing clusters to characterize these high-z sources in different wavelength domains including X-rays, IR, and sub-mm observations. Some interesting results have already been published by our team based on multi-band analysis (e.g. Schaerer & Pello 2005, Schaerer et al. 2007, Boone et al. 2007). A summary of results from our VLT Pilot Program and follow up observations is presented in Section 2.1.

The abundance of z>7 galaxies was discussed by Bouwens et al. (2006ab), and recently updated by different authors (e.g. Bouwens et al. 2008a, and Richard et al. 2008 based on NICMOS/HST data in the core of 6 lensing clusters; Fig. 1). Their conclusion is that strong evolution exists between z~7-8 and z~3-4, the Star Formation Rate (SFR) density being smaller at very high-z up to the limits of the different surveys. The Luminosity Function (hereafter LF) derived from these photometric surveys exhibit a turnover at the bright end of the LF relative to z~3-5, although the extent of this effect is still a matter of debate. One of the main issues, together with field-to-field variance, is the fact that all existing surveys, either space or ground-based, either in lensing or in blank fields, are still dramatically small in terms of effective surface (and corresponding surveyed covolume at high-z) compared to the needs in order to derive statistically significant results. Given the implication of these first results on future instrumental (both ground-based and space) and scientific developments, it is urgent to increase the size and the depth of the surveyed fields, as well as the wavelength coverage, to set strong constraints on the star-formation at redshifts beyond z ≥ 7.

The identification and study of the first galaxies require an extensive multi-wavelength coverage of their Spectral Energy Distribution (SED). In addition, detailed studies of spectral signatures for z ≥ 7 sources, including a precise redshift determination, require near-IR observations at wavelengths beyond 1μm. Most feasibility studies during the last years have been motivated by JWST, which should be able to observe these objects up to redshifts of the order of z ~20. However, the delay of JWST and the availability of well suited near-IR facilities in ground-based 8-10m telescopes fasten the development of observational projects targeting z ≥ 7 sources. The first studies on the physical properties of these sources can be started nowadays with instruments such as ISAAC/VLT (as proposed in our former pilot project) and pursued with improved efficiency using the future multi-object spectrographs to come, such as EMIR at GTC (also available to the ESO community in 2010), or KMOS for the second generation of VLT instruments (~ 2013). In other words, the efficiency of this research is closely related to the instrumental developments in the near-IR domain