Aims. We present and release photometric redshifts for a uniquely large and deep sample of 522286 objects with i AB ≤ 25 in the Canada-France Hawaii Telescope Legacy Survey (CFHTLS) "Deep Survey" fields D1, D2, D3, and D4, which cover a total effective area of 3.2 deg 2 . Methods. We use 3241 spectroscopic redshifts with 0 ≤ z ≤ 5 from the VIMOS VLT Deep Survey (VVDS) as a calibration and training set to derive these photometric redshifts. Using the "Le Phare" photometric redshift code, we developed a robust calibration method based on an iterative zero-point refinement combined with a template optimisation procedure and the application of a Bayesian approach. This method removes systematic trends in the photometric redshifts and significantly reduces the fraction of catastrophic errors (by a factor of 2), a significant improvement over traditional methods. We use our unique spectroscopic sample to present a detailed assessment of the robustness of the photometric redshift sample. Results. For a sample selected at i AB ≤ 24, we reach a redshift accuracy of σ ∆z/(1+z) = 0.029 with η = 3.8% of catastrophic errors (η is defined strictly as those objects with |∆z|/(1 + z) > 0.15). The reliability of our photometric redshifts decreases for faint objects: we find σ ∆z/(1+z) = 0.025, 0.034 and η = 1.9%, 5.5% for samples selected at i AB = 17.5-22.5 and 22.5-24 respectively. We find that the photometric redshifts of starburst galaxies are less reliable: although these galaxies represent only 22% of the spectroscopic sample, they are responsible for 50% of the catastrophic errors. An analysis as a function of redshift demonstrates that our photometric redshifts work best in the redshift range 0.2 ≤ z ≤ 1.5. We find an excellent agreement between the photometric and the VVDS spectroscopic redshift distributions at i AB ≤ 24. Finally, we compare the redshift distributions of i selected galaxies on the four CFHTLS deep fields, showing that cosmic variance is still present on fields of 0.7-0.9 deg 2 . These photometric redshifts are made publicly available at http://terapix.iap.fr (complete ascii catalogues) and http://cencos.oamp.fr/cencos/CFHTLS/ (searchable database interface).Key words. galaxies: distances and redshifts -galaxies: photometry -methods: data analysis Article published by EDP Sciences and available at http://www.edpsciences.org/aa or http://dx

This paper presents the "First Epoch" sample from the VIMOS VLT Deep Survey (VVDS). The VVDS goals, observations, data reduction with the VIPGI pipeline and redshift measurement scheme with KBRED are discussed. Data have been obtained with the VIsible Multi Object Spectrograph (VIMOS) on the ESO-VLT UT3, allowing us to observe 600 slits simultaneously at a spectral resolution R 230. A total of 11 564 objects have been observed in the VVDS-02h and VVDS-CDFS "Deep" fields over a total area of 0.61 deg 2 , selected solely on the basis of apparent magnitude 17.5 ≤ I AB ≤ 24. The VVDS efficiently covers the redshift range 0 < z ≤ 5. It is successfully going through the "redshift desert" 1.5 < z < 2.2, while the range 2.2 < z < 2.7 remains of difficult access because of the VVDS wavelength coverage. A total of 9677 galaxies have a redshift measurement, 836 objects are stars, 90 objects are AGN, and a redshift could not be measured for 961 objects. There are 1065 galaxies with a measured redshift z ≥ 1.4. When considering only the primary spectroscopic targets, the survey reaches a redshift measurement completeness of 78% overall (93% including less reliable flag 1 objects), with a spatial sampling of the population of galaxies of ∼25% and ∼30% in the VVDS-02h and VVDS-CDFS respectively. The redshift accuracy measured from repeated observations with VIMOS and comparison to other surveys is ∼276 km s −1. From this sample we are able to present for the first time the redshift distribution of a magnitude-limited spectroscopic sample down to I AB = 24. The redshift distribution N(z) has a median of z = 0.62, z = 0.65, z = 0.70, and z = 0.76, for magnitudelimited samples with I AB ≤ 22.5, 23.0, 23.5 and 24.0 respectively. A high redshift tail above redshift 2 and up to redshift 5 becomes readily apparent for I AB > 23.5, probing the bright star-forming population of galaxies. This sample provides an unprecedented dataset to study galaxy evolution over ∼90% of the life of the universe.

Abstract. We investigate the evolution of the galaxy luminosity function from the VIMOS-VLT Deep Survey (VVDS) from the present to z = 2 in five (U, B, V, R and I) rest-frame band-passes. We use the first epoch VVDS deep sample of 11 034 spectra selected at 17.5 ≤ I AB ≤ 24.0, on which we apply the Algorithm for Luminosity Function (ALF), described in this paper. We observe a substantial evolution with redshift of the global luminosity functions in all bands. From z = 0.05 to z = 2, we measure a brightening of the characteristic magnitude M * included in the magnitude range 1.8−2.5, 1.7−2.4, 1.2−1.9, 1.1−1.8 and 1.0−1.6 in the U, B, V, R and I rest-frame bands, respectively. We confirm this differential evolution of the luminosity function with rest-frame wavelength from the measurement of the comoving density of bright galaxies (M ≤ M * (z = 0.1)). This density increases by a factor of around 2.6, 2.2, 1.8, 1.5, 1.5 between z = 0.05 and z = 1 in the U, B, V, R, I bands, respectively. We also measure a possible steepening of the faint-end slope of the luminosity functions, with ∆α ∼ −0.3 between z = 0.05 and z = 1, similar in all bands.