Some general information about T0002
There are several important differences in release T0002 with respect to T0001 (December 2004). First of all, the T0002 deep stacks contain more data with respect to T0001 (for T0001, data were included from June 1st. 2003 to July 22, 2004 , wheras for T0002 the cut off date is now Nov. 22, 2004). For example, the D1-u data is now around 0.6 magnitudes deeper than the T0001 version. For the first time, this release also includes data from the "wide" survey: in total 42 stacked images of 1 deg2 are provided. The size of all output images has been fixed to be exactly 19354x19354 with a pixel scale of 0.186'', the native Megaprime/Megacam sampling. This is the TERAPIX T0002 "image quanta".
Secondly a major change is that new terapix tool "SCAMP" has been used to compute the global astrometric and photometric solutions, in contrast to "astrometrix" and "photometrix" which were used in T0001. For this release, astrometric and photometric solutions were computed using all available images within a radius of 2.5 degree of each field centre (and these field centres follow the survey definition, to be found here). All images enclosed with this radius are then selected to match the TERAPIX image quality criteria (namely, a seeing better than 1.1'' in all bands). Photometric images, used the computation of the photometric solution, were defined as those marked as photometric by CFHT's queue scheduling program.
For all stacks comprising at least three filters for, both deep and wide components of the survey, zero points have been adjusted by a small amount (usually less than 0.05 magnitudes) so that all colour-colour plots of stars occupy the same location in colour-colour space. (This should facilitate the computation of photometric redshifts and number counts of objects in particular regions of colour-colour space). In all cases the "reference" field has been arbitrarily chosen to be the D2 deep field.
All SExtractor catalogues have been re-extracted with these offsets applied and are supplied as part of the release. No change has been made to the deep object masks with respect to T0001: for some applications, this may be sub-optimal as with the increased exposure time the contrast of bright haloes around stars has risen accordingly. Star-galaxy separation was performed in an identical fashion to T0001, namely using the adaptive location of the stellar locus on each i-band image.
For the deep stacks, some slight changes were made to SExtractor's detection parameters which should reduce the number of spurious objects near the data's completeness limit with respect to T0001. As in T0001, in all cases detection was carried out on a "chisquared" detection image prepared from the corresponding g, r and i images. For each field, as before, merged catalogues are provided containing all total magnitudes available for each galaxy in each filter. The merged catalogues also contain the measured E(B-V) at the position of each object which may be used to correct the supplied magnitudes for the effects of galactic extinction (although in most fields the variation across the field is small). These merged catalogues contain a subset of parameters from the full catalogues, and these are also supplied as part of this release.
Notes on the catalogue names and contents for the deep fields
Two types of catalogues are provided in this release. In the first type, which have names like "d1[ugriz]_chisq_t02_wo.cat", all the sextractor parameters derived from the images are provided. There are at most five of these for each image quanta corresponding to each of the five MEGACAM filters. (The 'wo' suffix indicates, "with offset", that is a small zero-point offset was applied to the magnitude zero-point before extraction to align the colour-colour plots to the D2 field, as explained above. There are no 'wo' files for the D2 field as this, by definition, the reference field). The second type of catalogue, for the deep fields, have names like d[1,2,3,4]_ugriz_flag_t02_wo.dat] for each of the four deep fields. These catalogues contain a subset of the full chisquared catalogues presented above, but also contain the following additional information
Whether or not an object lies inside a masked region
Whether or not an object has been morphologically classified as a star based on the i-band image
What the local value of the E(B-V) is based on the schlegel dust maps.
The format of thde Deep catalogue is (Mag ugriz = MAG_AUTO):
id , x , y , ra , dec , r2 , flag , u , g , r , i , z , uerr , gerr , rerr , ierr , zerr , e(b-v)
The format is the same for the wide catalogues, expect that the e(b-v) is not included.
The meaning of the 'flag' parameter is identical to T0001., "0" and "1" corresponds to galaxies and stars not inside the masked regions, respectively. #Flag explanation :
# - 0/1 galaxy/star
# - 0/4 normal/in masked region
# - 0/8 normal/saturated
So "good" object (non-saturated objects not in masked regions) have: # flag = 0 (galaxy) or flag = 1 (star)
Back to the T0002 summary table