4.1 Overview

The T0007 release of the Wide survey covers 171 MegaCam pointings (or tiles) divided into the W1, W2, W3 and W4 fields and is complete in all filters. In total there are 855 u*,g,r,i∕y and z stacks and 171 chisquared images (2052 including the weight-maps). The positions and geometry of the four Wide fields comprising T0007, namely W1, W2, W3 and W4, are shown on Fig. 19 and summarized in Table 3. Table 6 lists the effective area, after masking, of the four Wide patches in each of the filters.

The image selection criteria applied to the 6378 CFHTLS Validated images for the production of the T0007 CFHTLS Wide survey are the following:

The parent sample of images after the QualityFITS selections is composed of 1000 u*, 1059 g, 1728 r, 1064 i, 291 y and 1236 z-band images. 25% of the sample are short photometric or Pre-Wide exposures that are only used for calibration but are not combined into stacks.

All stacks have the same pixel scale and cover exactly 1×1 deg2 (19354×19354 pixels of 0.1860"). The stacks only combine images that are part of a CFHTLS Wide observing sequence and that are within 3 arcminutes of the CFHTLS pre-defined stack center


Figure 19: Positions and geometry of the CFHTLS Wide fields. The black thick contours show the total field of view composing the u*,g,r,i∕y and z stacks of the T0007 release. The fields located outside these regions are supplementary astrometric calibration data but are not included in the stacks. W1, W2, W3 and W4 do not cover the same field of view, so the four Wide regions are not shown with the same scale in this figure. The blue squares show a typical MegaCam field of view and indicate the positions of the reference center field. The MegaCam images included in the release are in green. They reveal the tiling and mosaicing of each CFHTLS Wide area. The small rectangles indicate individual MegaCam CCDs. These plots have been produced by SCAMP during the calibration step of W1, W2, W3 and W4.


Figure 20: Positions, identification and naming conventions of stacks in the CFHTLS Wide survey. The figure shows the positions of each “tile” (blue square corresponding to the field of view of MegaCam) covering the W1 and W2 Wide fields. Each tile is labelled by its Cartesian name on top (see http://terapix.iap.fr/cplt/oldSite/Descart/summarycfhtlswide.html) and the official CFHTLS below. Field W*(0,0) (in red) at the approximate center field of each survey patch, before the survey strategy changes.

Reference center
Total Total sky Filters Comment
field ra dec pointings coverage
name [J2000] [J2000] [n×p] [deg2]

W1 02:18:00 -07:00:00 9×8 8.5×7.5 u*,g,r,i∕y,z 16 y
W2 08:57:49 -03:19:00 5×5 4.8×4.7 u*,g,r,i∕y,z New center; 2y
W3 14:17:54 +54:30:31 7×7 6.7×6.6 u*,g,r,i∕y,z 7 y
W4 22:13:18 +01:19:00 25 23.3 u*,g,r,i∕y,z se-nw elongated; 7 y

Table 3: Overview of the CFHTLS Wide fields. The pointings “n×p” refer to the numbers of tiles along the horizontal and vertical axes of MegaCam (i.e. the e-w and n-s axes). Sky coverage is expressed along ra and dec axes.

Filter Median number Median exposure time Mean limiting magnitude Mean image quality
of exposures [seconds] [MegaCam AB system] [′′]

u* 5 3000 25.2 0.85
g 5 2500 25.5 0.78
r 4 2000 25.0 0.71
i∕y 7 4300 24.8 0.64
z 6 3600 23.9 0.68

Table 4: Mean properties of the CFHTLS Wide survey. The i and y are not split into two sub-samples. The Mean limiting magnitude corresponds to the 80% completeness limit for point-like sources. The Mean image quality is the mean FWHM of stellar sources over all stacks.

Each stack comprises a sequence of medium exposures a few hundred seconds long each. After each exposure the telescope is shifted by a small amount (typically a few tens of arcseconds, see the right panel of Fig. 21) in order to fill the gaps between CCDs and enable the rejection of bad pixels and cosmic rays during subsequent processing steps.

The adjacent fields have an overlapping region of several arcminutes width to constrain the field-to-field astrometric and photometric calibrations. The size of the overlaps is 4 arcmin in DEC and 3 arcmin in RA (see Fig. 21). The overlaps between fields reduce the sky coverage of the 171 CFHTLS Wide tiles to approximately 155 deg2.


Figure 21: Offsets between adjacent pointings (left) and dithering (right) inside a pointing of the CFHTLS Wide survey. Each pointing (or tile) overlaps along the ra and dec directions with its nearest neighbors. In addition to the large offsets, the observations at each CFHT Wide reference position ("Ref" and green filled circles) are split into an observing sequence of exposures, with a small dither between each. The dithering pattern depends on the number of exposure per sequence and is different for each filter (see Table 4). In all configurations the dithered positions (black filled circles) are among those located inside a 30"×180" ellipse centered at the reference position of a Wide pointing. This is illustrated on the right hand of the figure (drawn from the CFHT dithering pattern plot: http://www.cfht.hawaii.edu/Instruments/Imaging/MegaPrime/specsinformation.html. The details of this figure are given at this URL). The offsets and the dithers along the DEC direction are sufficiently large to fill the central horizontal gap (83′′high) between the CCDs (see Fig. 2) with a fraction of exposures of each sequence. The dash-dot-dot blue squares on the left show the positions of Pre-Wide images. They are shifted by a half-MegaCam field of view, in both directions, in order to provide very large overlaps between pointings and ease the astrometric calibration. The Pre-Wide images are short exposures (3 mn) only done in r-band.

Each tile is centered at a well-defined position. The coordinates of a stack center position are identical for all filters. W1, W2, W3 and W4 are therefore composed of a complete set of u*,g,r,i,y and z adjacent square tiles of 1 deg2 each.

The integer part of the center position of each tile is used to name the final stacks and other data products. This naming convention was agreed by CADC, Terapix and the CFHTLS Steering Group and is valid for all CFHTLS releases.

In addition to the official astronomical CFHTLS naming convention, Terapix defined in 2002 a shorter naming convention based on a Cartesian grid coordinate system, where the increment unit is a MegaCam-size field and the center positions of the grid points are the center positions of the tiles21 .

The reference center of each Wide field of view is defined as the reference W[1,2,3,4](0,0) and the Cartesian field names increase toward the East and North. The tiling and field naming conventions are drawn on Fig. 20 and the complete list of the T0007 Wide stacks is given in Tables 30.

The total exposure time per filter is approximately the same for each tile (see Table 30 and the synoptic table). For some tiles, the exposure time is longer than the nominal value, because exposures that were considered out of spec by QSO and hence immediately reobserved have been later on salvaged at the post processing time. It produces very sharp exposure time distributions for all filters (see Fig. 23). Some observations have been adjusted to take into account unexpected events (like absorption by clouds/cirrus or technical problems). There are a few fields that have deeper u*-band and z-band exposures which explains the small tails shown in the inset of Fig. 23. These longer exposures were taken before it was decided to cut the u* and z exposure times by a factor of two at the survey mid-term review (2005).

In addition to the original CFHTLS exposures, the VIPERS consortium took complementary exposures to fill in the holes in W1 which were caused by malfunctioning CCDs later in 2003. These images (distributed in 6 pointings and three filters u,g and y) have been added to T0007. As a result, these pointing are deeper than the rest of the survey. A detailed description of the added VIPERS-DDT images can be found in Table 5. Their position on the W1 patch are shown in Figure 22.

Figure 22: Position of the additional VIPERS-DDT exposures used to fill in the holes in the survey. Top left: u*-band, top-right: g-band, lower left: y-band. lower-right: filter mask coverage of the W1 Wide patch (See Section 6 for the meaning of the flags). Seven fields still remain with half-CCD holes in one or more filters due to the malfunctioning of some CCDs late 2003. These anomalies are listed in Tables 14 and 15.

Wide position Filter Exposure time (per image) Number of exposures

W1+2+2 u.MP9301 600 6
W1+4+2 g.MP9401 290 7
W1+3+1 g.MP9401 290 5
W1+3+1 g.MP9401 290 5
W1+4+1 g.MP9401 290 5
W1+2+1 i.MP9702 300.096 7
W1+3+2 i.MP9702 300.092 5

Table 5: Description of the 40 additional VIPERS-DDT images.

All r-band stacks combine data taken during two epochs separated by at least two years and totalling 2×500 s each. In all filters except u*, observations were carried out only when the seeing (FHWM) was below 0.95". For the u*-band, this rule was relaxed to 1.3".

The complete list of CFHTLS input images that were combined into the Wide stacks is given in the Appendix A.2, in Table 31.


Figure 23: Distribution of exposure times over the Wide fields. The inset is a close-up view of the long exposure tails.

The global properties of the CFHTLS Wide survey are summarized in Table 4 and in more detail in Table ??. Overall, they meet the survey specifications but the homogeneity of the survey over a Wide patch scale must be assessed carefully. The Wide field-to-field and MegaCam tile-to-tile scatters will be explored in detail in the following sections. The depth, the seeing, the photometric errors analysis are presented in the next parts of the document.

Finally, the complete set of configuration files used for the T0007 release is presented in the Appendix. They are intended for users who would prefer to re-process the pre-processed or the stacked images with parameters tuned for their own scientific projects. The configuration files are available on request from Terapix.

Field Parameter u* g r i y z ugriz coverage

W1 Nb stacks 72 72 72 56 16 72 72
Raw observed area - - - - - - 64.16
Final unmasked area 58.47 (210500) 58.41 (210267) 58.49 (210581) 45.41 (163490) 13.08 (47096) 58.47 (210504) 58.33 (209975)
W2 Nb stacks 25 25 25 23 2 25 25
Raw observed area - - - - - - 22.22
Final unmasked area 18.60 (66952) 18.51 (66624) 18.60 (66952) 17.19 (61878) 1.43 (5152) 18.60 (66952) 18.50 (66622)
W3 Nb stacks 49 49 49 42 7 42 49
Raw observed area - - - - - - 41.91
Final unmasked area 37.04 (133351) 37.04 (133351) 37.04 (133351) 31.74 (114277) 5.56 (20018) 37.04 (133351) 37.04 (133340)
W4 Nb stacks 25 25 25 18 7 25 25
Raw observed area - - - - - - 22.09
Final unmasked area 19.66 (70777) 19.66 (70777) 19.66 (70777) 14.10 (50754) 5.62 (20218) 19.66 (70777) 18.50 (66622)
Wide Nb stacks 171 171 171 129 32 171 25
Raw observed area - - - - - - 22.09
Final unmasked area 133,77 (481580) 133.616 (481019) 133.795 (481661) 108.44 (390399) 25.69 (92484) 133.78 (481584) 132.38 (476559)

Table 6: The raw observed area is the surface covered by MegaCam the sky. The Final unmasked area is the surface observed by the CFHTLS which is not masked (due to foreground star and instrumental defects). The surface is given in square degrees and square arcminutes.