6.1 Data set
Unlike the previous CFHTLS releases, T0007 is immediately public. The T0007 synoptic
table
is an easy way to access all the data and presents a complete summary of the release. These images and
catalogues form the core of the T0007 release. These data are:
- the complete set of individual weight-map images corresponding to the 6378 and 8916 single
images used during the production of the CFHTLS Wide and Deep releases. The Terapix
weight-map images are Multi-Extention FITS (MEF), like the native CFHT images.
- Stacked images with their corresponding weight-maps;
- 855 Wide stacks, corresponding to u*, g, r, i or y and z bands:
- 360 for W1,
- 125 for W2,
- 245 for W3 and
- 125 for W4;
- 96 Deep stacks corresponding to u*, g, r, i, y and z bands, consisting of:
- 24 for D1 (6 D-25 MEDIAN, 6 D-25 SIGWEI, 6 D-85 MEDIAN, 6 D-85 SIGWEI),
- 24 for D2,
- 24 for D3 and
- 24 for D4;
- each stack corresponds to a tile of 1 deg.×1 deg. field (19354×19354 pixels of 0.1860"), centered at the
positions listed in Tables 30 (Wide), and in 16 (Deep);
- the Equatorial coordinate system is J2000 (ICRS) and the projection type is the distorted tangential TAN.
Stacks are produced using a median filter, are weighted accordingly using the weight-map images, and
combined using a Lanczos3 interpolation kernel.
- the “chi2” detection images (g-r-i∕y for the Wide and g-r-i for the Deep data) images with their
weight-map (FITS image).
- the catalogues of sources extracted from each stack, prior to producing catalogues from the chi2 reference
image (855 .ldac Wide and 96 .ldac Deep FITS table);
- single filter catalogues, one per stack. They are creating by running SExtractor in dual-image mode
using the chi2 image as the reference source detection image and the u*,g,r,i,y or a z-stack as a
photometry image (*u/g/r/i/z*.cat ASCII tables). The T0007 single filter catalogues contain all
parameters listed in Tables 27, 28 and 29. There is one source catalogue per chi2 source detection and
per filter, that is 855 Wide and 96 Deep catalogues. So, for each MegaCam stack position,
the Wide (u*, g, r, i∕y, z) and the Deep (u*, g, r, i, y, z) catalogues have the same number of
sources;
- ASCII merged source catalogues (M-SC) are also provided:
- 342 Wide merged 5-filters u*gr[i∕y],z catalogues (either i or y band data inside), and
- 32 Deep merged 6-filters u*griyz catalogues (both i and y band inside),
in ASCII table format. The Wide catalogues are named *urgiz*.cat or *urgyz*.cat; the Deep are all
named *ugriyz*.cat, and contain i and y data. They contain all objects with a restricted list of
parameters from the parent single filter catalogues, but include e(b - v) values for each objects.
They are estimated at each source position from interpolation of the ? maps; Two sets of
catalogues are produced containing either MAG_AUTO magnitudes or SNLS aperture magnitudes
(SNLS and IQ20 magnitudes). Please refer to section 6.3 for a detailed catalogue content
description.
- The ASCII Wide patch merged catalog. These four catalogues (one for each Wide patch) contain all
parameters in all filters for each object detected in the corresponding Wide patch. The parent
catalogues are the Chi2 catalogues which are merged together. For objects in overlapping regions
which are detected on several tiles, the parameters are kept from the largest i-band signal
to noise detection (FLUX_AUTO/FLUX_ERR). Note that these catalogues are huge (40 to 130
GB).
- the DS9 compliant masks (.reg ASCII file), one for each stack. The masks are produced automatically.
The masking uses the USNO source catalogues to locate bright stars in CFHTLS stacks and draw a
polygon that delimits a polluted region. The size of the polygon depends on the magnitude of the stars.
After the stellar source masking, polygons are added automatically to all mask, to exclude the edges of
fields. Masks are then tuned manually with additional polygons that exclude regions with missing
CCDs.
It is important to notice that the mask needs depend on the science goals. The T0007 masks may therefore
be tuned accordingly. We recommend to CFHTLS users to overlay the masks on a DS9 view of each stack
image and have a first look prior to use them.
All quality control files produced during Terapix processing are public. The supplementary data available
are:
- a set of 855 Wide and 96 Deep binary masks images (FITS), one per filter, based on the ASCII
.reg masks;
- the 6378 Wide and 8916 Deep individual QualityFITS evaluation files attached to each input
image (QFITS-in). Only QFITS-in data of images that were selected for stacks in T0007 are
available;
- the astrometric and photometric initial rescaling calibration files attached to each image during the
SCAMP calibration step. There are 6378 Wide and 8916 Deep .ahead and .head (ASCII) files;
- the complete list of MegaCam images contained in each stack (upon request);
- the list of the L99 stack rescaling factors;
- the 855 Wide and 96 Deep individual QualityFITS evaluation files for each final stack
(QFITS-out);
- a series of quality assessment files: tables, images, plots and statistics for any image are available
from the
T0007 synoptic table http://terapix.iap.fr/cplt/T0007/table_syn_T0007.html. They
include:
- all stellar colour-colour plots. There are three color-color plots per Wide stack, and five
color-color plots per Deep stack;
- all three-filter colour (.jpg) images. There are three color images per stack that combines
the five filters in different ways;
- all completeness limit (.png) plots, one per stack. The ascii .dat data tables used to make
these plots are also available, upon request to terapix;
- all galaxy count (.ps) plots, one per stack;
- the complete QualityFITS QFITS-out quality control files (.html page); one per stack;
- the complete set of SCAMP output and quality control files (.html page); between one (all
image together) and six (one per filter) per field;