CARS: the CFHTLS-Archive-Research Survey. I. Five-band multi-colour data from 37 sq. deg. CFHTLS-wide observations

International audienceContext: We present the CFHTLS-Archive-Research Survey (CARS). It is a virtual multi-colour survey that is based on public archive images from the Deep and Wide components of the CFHT-Legacy-Survey (CFHTLS). Our main scientific interests in the CFHTLS Wide-part of CARS are optical searches for galaxy clusters from low to high redshift and their subsequent study with photometric and weak-gravitational lensing techniques. Aims: As a first step in the CARS project, we present multi-colour catalogues from 37 sq. degrees of the CFHTLS-Wide component. Our aims are first to create astrometrically and photometrically well-calibrated co-added images from publicly available CFHTLS data. Second, we offer five-band (u^*g'r'i'z') multi-band catalogues with an emphasis on reliable estimates for object colours. These are subsequently used for photometric redshift estimates. Methods: We consider all those CFHTLS-Wide survey pointings that were publicly available on January 2008 and that also have five-band coverage in u^*g'r'i'z'. The data were calibrated and processed with our GaBoDS/THELI image processing pipeline. The quality of the resulting images was thoroughly checked against the Sloan-Digital-Sky Survey (SDSS) and already public high-end CFHTLS data products. From the co-added images we extracted source catalogues and determined photometric redshifts using the public code Bayesian Photometric Redshifts (BPZ). Fifteen of our survey fields directly overlap with public spectra from the VIMOS VLT deep (VVDS), DEEP2 and SDSS redshift surveys, which we used for calibration and verification of our redshift estimates. Furthermore we applied a novel technique, based on studies of the angular galaxy cross-correlation function, to quantify the reliability of photo-z's. Results: With this paper we present 37 sq. degrees of homogeneous and high-quality five-colour photometric data from the CFHTLS-Wide survey. The median seeing of our data is better than 0.9 arcsec in all bands and our catalogues reach a 5σ limiting magnitude of about i'_AB≈ 24.5. Comparisons with the SDSS indicate that most of our survey fields are photometrically calibrated to an accuracy of 0.04 mag or better. This allows us to derive photometric redshifts of homogeneous quality over the whole survey area. The accuracy of our high-confidence photo-z sample (10-15 galaxies per sq. arcmin) is estimated with external spectroscopic data to σΔ_z/(1+z)≈ 0.04-0.05 up to i'_AB Based on observations obtained with MegaPrime/MegaCam, a joint project of CFHT and CEA/DAPNIA, at the Canada-France-Hawaii Telescope (CFHT), which is operated by the National Research Council (NRC) of Canada, the Institut National des Sciences de l'Univers of the Centre National de la Recherche Scientifique (CNRS) of France, and the University of Hawaii. This work is based in part on data products produced at TERAPIX and the Canadian Astronomy Data Centre (CADC) as part of the Canada-France-Hawaii Telescope Legacy Survey, a collaborative project of NRC and CNRS