VIPERS-MLS - Survey

W1 field	W4 field

Ks band imaging: The WIRCam coverage is shown by the red regions in the W1 and W4 fields. It fully maps the VIPERS survey (but the VIDEO region) and was extended in southern part in W4 when DXS full area was not yet released. It covers a total area of ~27 sq. deg. (~22.4 sq. deg. after masking)
Observations have been conducted from 2010 to 2013 by combining PI programs from France (Arnouts), Canada (Van Waerbeke, Thanjavur and Willis) and Hawaii (Morrison) for a total integration time of 120 hrs. The Ks band exposure time is 1050 sec split in two OGs of 21 exposures. Some pointings have been repeated due to bad IQs or high sky background as shown in the figure on the left below. Mean IQ per WIRCam pointing is shown in the central figure below and corresponds to < IQ > ~0.6 +/- 0.09 arcsec. All the WIRCam data have been reduced and stacked at Terapix using the astrometric grid of the MegaCam CFHTLS tiles. Terapix performed completeness tests by adding simulated sources (stars and extended sources) in all the final images. The completeness at 80% and 50% for stars and galaxies respectively in all the images are shown in the right figure below.

Number of exposure/pointing	Mean IQ per pointing	Simulated Completeness

GALEX imaging: The GALEX observations are part of the GALEX deep imaging survey (DIS, with exposure time Texp>30,000 sec) in FUV (~1500A) and NUV (~2300A). In addition to the DIS survey, we have observed new pointings in the western and eastern sides of the W1 field in the NUV band only (after the shut-off of the FUV channel) thanks to ~100hrs of discretionary time from the GALEX PI, C. Martin. All the GALEX pointings are shown as cyan and blue circles in the W1 and W4 fields. Unfortunately we did not have a chance to observe the VIPERS region in the W4 field. The total area (W1+W4) imaged by GALEX is ~7.8 sq. deg. in FUV band and ~22 sq. deg. in NUV band.

Strategy: Our catalog relies on the sources detected in the CFHTLS T0007 catalog and based on the gri-chi2 detection image.

The WIRCam photometry, in the Ks stacks (based on the same tiles as CFHTLS), is performed in dual mode with the same SExtractor parameters as T0007. Isophotal magnitudes are used to measure the colors in ugrizK bands while the total fluxes are obtained by applying a scaling factor (total-to-iso based on a weighted mean correction with the g,r,i and Ks bands) for each source as discussed in Moutard et al., 2016a.

The GALEX photometry is perfomed with a dedicated software EMphot to handle the source confusion of deep GALEX images due to the large PSF (FWHM~5arcsec). The basic concept is to use optical priors and convolve them by the GALEX PSF. The PSF's amplitude of each optical prior is obtained via a global maximum likelihood estimation between the GALEX image and the simulated image in small tiles of 4x4arcmin^2. Priors can be considered as Dirac or as extended sources. In this analysis we use the stamps of the u band priors brighter than u=25.5 as illustrated in the images below.
The comparison of the EMphot flux measured with the stamps and the original SExtractor GALEX pipeline is shown in the left Figure below. EM magnitudes agree with SExtractor magnitudes for isolated sources, both for compact (blue dots) and for extended sources (red dots). For the other sources (grey dots) EM magnitudes tend to be fainter by better dealing with the source blending. As shown in right Figure, the EM number counts (blue line) are lower at bright magnitudes (21< NUV <24) than GALEX number counts (black line) while they are consistents for the isolated sources (dashed lines). Due to the use of priors, EM photometry also provides deeper number counts (but with low signal to noise).
Details about the EMphot algorithm, analysis of the GALEX images, quality assessments and simulations are available directly from our GALEX database: GALEX-EMPHOT

GALEX with shape priors	simulated GALEX image	Residual image

EM vs GALEX magnitudes	Number counts EM vs GALEX

Final catalog: The EMphot photometry represents total magnitude. They are combined with the total magnitudes in optical and Ks bands to perform the SED fitting for photometric redshift and star-galaxy separation as described in Moutard et al 2016a,b. Number counts and photometric redshifts are shown in the figures below for the GALEX FUV, NUV and Ks bands.

GALEX Number counts	Ks Number counts	Redshift Distribution