README for MEGAPRIME cluster data of RXJ1347-1145 ================================================= Last file update: 29.07.2009 (Thomas Erben) =========================================== VERY short log on new data: - There are newly processed images with version number V2.0A. - The set now contains the five colours ugriz - The set now contains an image ...sum.fits which gives, at each pixel position, the number of input images contributing to that pixel. - Astrometry was done w.r.t. 2MASS --------------------------------------------------- Original file: ============== I processed CFHT/MGAPRIME data for the cluster RXJ1347-1145 with the THELI pipeline (see Erben et al., AN 326,432) for details. MEGAPRIME has a f.o.v. of 1 sq.degree with a pixel scale of 0.186 arcsec. This implies that the science data are quite big images; nearly 2GB per frame with about 20kx20k pixels!! The cluster has deep, sub-arcsecond data in the 'r'-band (suitable for lensing) and supplementary observations in the g, i and z bands. The lensing 'r'-band data were obtained in 'two blocks' with a separation of about 6 arcmin in DEC. In total I created three stacks for the 'r'-bands: - One 'deep' stack containing all available (and usable) data - One stack for each of the 'blocks' (called POS1 and POS2) that contain only images with the same declination This allows us to perform independent shape measurements on the two blocks and to estimate systematics. We also can decide whether it is safe to use the deep stack for shape measurements at the end. For the supporting colors I only created the deep stack. For each data set there are three images: RXJ1347_g(riz)_V1.5A(_POS1/2).swarp.fits RXJ1347_g(riz)_V1.5A(_POS1/2).swarp.weight.fits RXJ1347_g(riz)_V1.5A(_POS1/2).swarp.flag.fits - RXJ1347: cluster name - 'g', 'r', 'i' or 'z': filter - V1.5A: internal version number - POS1/2: the two blocks 1 or 2; no 'POS' identifier for the deep stacks - .swarp.fits: SCIENCE image co-added with 'swarp' - .swarp.weight.fits: WEIGHT image. To use it in SExtractor you need to use the 'WEIGHT_TYPE' 'MAP_WEIGHT' - .flag.fits: FLAG image: It contains a '1' for pixels that have a zero in the WEIGHT image. This concerns mainly the image borders and saturated pixels. When using it in SExtractor it allows you to get rid of objects which have pixels falling into these 'bad' areas. The image headers contain the keys GAIN, MAGZP and SEEING which you need to feed into your SExtractor config file. Magnitudes are measured in instrumental AB magnitudes. The pixel values have the unit ADU/s, i.e. they are normalised to an exposure time of 1 sec. With this the total (apparent) magnitudes of objects are calculated with 'mag = MAGZP-2.5*log(total flux)'. This is what you directly get out of SExtractor when you give MAGZP as magnitude zeropoint. Astrometry is done w.r.t. the USNOB1 catalogue. All images of a particular cluster share the same CRVAL1 and CRVAL2 values. Hence you can create images that precisely match on the pixel level if you cut them w.r.t. the pixel values given in the CRPIX1, CRPIX2 header keywords.