README FILE TO ACCOMPANY UDF HRC PARALLEL HIGH-LEVEL SCIENCE PRODUCTS ===================================================================== 1. General Information ---------------------- The ACS High Resolution Channel (HRC) offers the possibility to obtain auto-parallels when the WFC is prime. Since the Wide Field Channel (WFC) and the HRC share the same filter wheels, the filter in use by WFC dictates the HRC filter. For the HUDF observations HRC auto-parallels have been obtained with the F220W, F555W and F892N filters and the G800L grism. The pixel size is 0.027arcsec giving high sensitivity to compact sources. The F555W images will provide a deep V band image; the F220W, although not very sensitive will provide a UV colour; the F892N filter is narrow (150A) but is of interest for detecting emission line objects in a narrow range. The G800L grism provides slitless spectra at a dispersion of 24A/pixel, but the resolution depends on the size of the dispersing object. The coverage is 5500-10500A with peak sensitivity about 7500A. The four roll angles of the WFC UDF campaign produce four distinct HRC fields; however the pairs of iamges separated by roll angles of about 4 degrees have overlap of about 50% and were combined. The High Level Products were produced for two fields, referred to as Epoch-1 and Epoch-2. Approximate coordinates for the two fields of ACS HRC parallels are: Epoch ORIENT RA (J2000) DEC (J2000) 1 310 03 32 54 -27 49 07 1 314 03 32 54 -27 49 21 2 40 03 32 32 -27 50 52 2 44 03 32 31 -27 50 45 2. Release Data --------------- 2.1 Direct Images ----------------- High-level drizzled image products in the three filters (F220W, F555W, F892N) and the grism (G800L) for Epoch-1 and Epoch-2 were produced. In these images North is up, east to the left and the pixel size is 30mas. The drizzled images for Epoch-1 are: h_udfhrce1f220drz_img.fits h_udfhrce1f555drz_img.fits h_udfhrce1f892drz_img.fits h_udfhrce1g800drz_img.fits and the drizzle weight images are: h_udfhrce1f220wht_img.fits h_udfhrce1f555wht_img.fits h_udfhrce1f892wht_img.fits h_udfhrce1g800wht_img.fits The drizzled images for Epoch-2 are: h_udfhrce2f220drz_img.fits h_udfhrce2f555drz_img.fits h_udfhrce2f892drz_img.fits h_udfhrce2g800drz_img.fits and the drizzle weight images are: h_udfhrce2f220wht_img.fits h_udfhrce2f555wht_img.fits h_udfhrce2f892wht_img.fits h_udfhrce2g800wht_img.fits 2.2 Object catalogues --------------------- Object catalogues were produced using SExtractor (v. 2.3). The output format is identical to that used for the UDF-WFC catalogue, but different SExtractor configuration files were employed. In particular, one of the standard SExtractor filters (gauss_2.0_5x5.conv) and a smaller DETECT_MINAREA of 5 pixels were used, allowing detection of a number of compact objects which were otherwise missed. The SExtractor configuration files for Epoch-1 are: h_udfhrce1_f220_sex.txt h_udfhrce1_f555_sex.txt h_udfhrce1_f892_sex.txt and the SExtractor configuration files for Epoch-2 are: h_udfhrce2_f220_sex.txt h_udfhrce2_f555_sex.txt h_udfhrce2_f892_sex.txt SExtractor was run in dual-image mode, using the F555W images for object detection. It was verified that the F220W and F892N images did not contain any objects which were undetected in the F555W images. Comparison with the GOODS F606W images showed that all objects detected in the Epoch-1 data were real, while a few spurious detections (located near the edges) were present in the Epoch-2 data (objects 1, 2, 6, 24, 48, 55 and 58). The SExtractor catalogues for Epoch-1 are: h_udfhrce1f220_cat.txt h_udfhrce1f555_cat.txt h_udfhrce1f892_cat.txt and for Epoch-2 are: h_udfhrce2f220_cat.txt h_udfhrce2f555_cat.txt h_udfhrce2f892_cat.txt The first column in these final object catalogues contains object names in IAU-compliant format (JRAMMSS.SS-DEMMSS.S) and was added after the SExtractor reduction. 2.3 Spectra ----------- The object spectra are released as tables in ASCII and FITS format. File names are given according to the following convention: h_udfhrce*id???_spc.#### with * as 1 or 2 to denote spectra from Epoch-1 or Epoch-2 respectively. #### is either 'txt' or 'fits' to indicate the data format. ??? is the object number (column 2) in the SExtractor catalogues 'h_udfhrce1f555_cat.txt' for Epoch-1 and 'h_udfhrce2f555_cat.txt' for Epoch-2. Some spectra were extracted based on the GOODS CDFS coordinates (see section 2.4) and for these spectra ??? is set to > 100. The FITS tables and the ASCII tables have identical keywords and data columns. Important keywords are: ID_IAU : the identification following IAU guidelines RA2000 : the right ascension in J2000 coordinates DEC2000 : the declination in J2000 coordinates EXPTIME : the maximum coadded exposure time in [S] (see 4.2) NCOADDS : the number of coadded images DIRIMAGE: the direct image name X_IMAGE : the x-position on direct image in [pixel] Y_IMAGE : the y-position on direct image in [pixel] ID_GOODS: the GOODS CDFS name for sources from the GOODS catalogue Other keywords are derived from the FITS header of the raw images. The object spectra are stored in the six table columns: LAMBDA : the wavelength in [ANGSTROM] COUNTS : the object flux in [ELECTRONS/S] ERR_CTS : the error for the value in column 'COUNTS' in [ELECTRONS/S] FLUX : the object flux in [ERG/CM^2/S/A] ERR_FLUX: the error for the value in column 'FLUX' in [ERG/CM^2/S/A] CONTAM : either '1' or '0' to show whether there a contamination problem at this wavelength or not (see 4.3) The ASCII tables have the first column as ROW : row number followed by the above six columns. 2.4 GOODS CDFS Detected Objects ------------------------------- The following lists show the objects taken from the GOODS CDFS catalogue which were added to the UDF HRC F555W catalogue objects to extract all possible spectra from the grism images. The GOODS CDFS object name (ID_IAU) is followed by the object number (see 2.3) used to index the spectra. To discriminate the GOODS objects from the objects found on the UDF HRC F555W image, the object numbers are indexed from 101 and the keyword ID_GOODS is set to the name from the GOODS CDFS catalogue. Epoch 1: -------- GOODS CDFS name Obj. Num J033252.17-274923.8 101 J033252.27-274929.8 102 J033252.58-274927.5 103 J033252.64-274914.2 104 J033252.68-274931.0 105 J033252.75-274934.2 106 J033253.20-274911.6 107 J033253.35-274943.7 108 J033253.50-274859.4 109 J033253.53-274859.1 110 J033253.65-274930.4 111 J033253.72-274942.1 112 J033253.84-274943.9 113 J033253.90-274901.4 114 J033254.10-274916.0 115 J033254.11-274854.5 116 J033254.30-274934.1 117 J033254.30-274927.5 118 J033254.32-274935.4 119 J033254.47-274916.4 120 J033254.79-274925.1 121 J033254.88-274911.1 122 J033255.04-274922.3 123 J033255.06-274912.8 124 J033255.09-274916.2 125 J033255.16-274923.5 126 J033255.18-274921.0 127 J033255.21-274905.3 128 J033255.31-274926.4 129 J033255.38-274917.0 130 J033255.39-274920.3 131 J033255.49-274915.3 132 J033255.59-274911.4 133 J033255.71-274915.6 134 J033255.81-274907.4 135 J033255.94-274912.3 136 Epoch 2: -------- GOODS CDFS name Obj. Num J033229.10-275038.6 101 J033229.29-275050.8 102 J033229.33-275038.2 103 J033229.37-275053.9 104 J033229.38-275038.9 105 J033229.47-275053.2 106 J033229.47-275053.9 107 J033229.51-275032.3 108 J033229.64-275102.4 109 J033229.76-275034.0 110 J033229.80-275104.5 111 J033229.85-275105.9 112 J033229.86-275031.9 113 J033229.92-275055.2 114 J033230.03-275026.8 115 J033230.16-275108.0 116 J033230.17-275034.1 117 J033230.18-275025.6 118 J033230.29-275052.1 119 J033230.37-275030.2 120 J033230.45-275027.3 121 J033230.68-275109.1 122 J033230.79-275044.5 123 J033230.97-275106.5 124 J033231.15-275111.3 125 J033231.21-275054.5 126 J033231.22-275108.9 127 J033231.25-275041.2 128 J033231.28-275050.0 129 J033231.40-275114.3 130 J033231.49-275045.6 131 J033231.51-275056.7 132 J033231.51-275058.7 133 J033231.56-275054.8 134 J033231.77-275040.9 135 J033231.80-275110.4 136 J033231.82-275110.6 137 3. Data Reduction ----------------- 3.1 Filter Data --------------- The raw images were processed with the CALACS standard pipeline, and exposures for each epoch were then drizzled onto common reference frames using the stsdas/multidrizzle task. All images were drizzled with North up and East to the left using an output pixel size of 30 mas, the same as that used for the UDF WFC frames (although not projected on the identical tangent plane). The dropsize ("pixfrac") was set to 0.5 and the final output images have dimensions of 2000 x 2000 pixels. The World Coordinate System header keywords (CRVAL1 and CRVAL2) in the drizzled images were adjusted to provide coordinates consistent with the GOODS CDFS catalogues. The offsets between the raw UDF/HRC data and the GOODS catalogs were less than 1 arcsec. The photometric keywords (PHOTPLAM, PHOTFLAM and PHOTBW) in the image headers were updated with the latest values published on the ACS Web Pages. 3.2 Spectra ------------ The extraction of the spectra from the grism images is a complex process which involves various software packages. Here only a broad overview over the various steps that are necessary to derive the released data are given. Additional details can be found on the preview pages of the UDF HRC Parallels at http://www.stecf.org/UDF/HRCpreview.html. The main steps of the data reduction are: a) standard CALACS pipeline reduction b) cosmic ray detection and sky background removal c) extraction of the 2D spectra from the individual images d) drizzling of the 2D spectral images to coadded images with a common wavelength scale e) 1D extraction of the spectra from the drizzled coadded images f) creation of the preview a) Standard Pipeline Reduction ------------------------------ The raw images were processed with the CALACS standard pipeline. b) Cosmic Ray Detection and Background Subtraction -------------------------------------------------- To flag the cosmic rays, the stsdas/multidrizzle task was run on the _flt files using the default configuration. As a by-product multidrizzle detects cosmic rays on each image and flags those pixels. Mask files of the flagged pixels were transferred to the data quality extension of the corresponding _flt file. To perform a global background subtraction on each grism image, a master sky image was constructed from a combination of all ACS HRC grism images (864) from the UDF and GOODS programmes. Additional information concerning the master sky is available on request. c) Extraction of the 2D Spectra from the Individual Images ---------------------------------------------------------- The 2D spectra were extracted from the individual images using the aXe spectral extraction software. This software package was specifically designed to handle large format spectroscopic slitless images such as from the ACS. Details about aXe can be found at http://www.stecf.org/software/aXe/index.html. In order to extract the object spectra, the UDF HRC F555W catalogue (see Sect. 2.2) was used together with a list of objects from the GOODS CDFS catalogue (see sect 2.4). The GOODS sources were included in the extraction list either if they were not found in the (shallower) UDF HRC F555W image, or only part of their first spectral order (but not their direct image) fell in the area covered by the UDF HRC data. For all objects in the extraction catalogue, grism stamp images were produced using standard aXe tasks with the addition of tasks newly developed for the drizzle extension of the aXe software. d) Drizzling of the Stamp Images to Coadded Images -------------------------------------------------- One way to derive the coadded spectrum of an object from several grism images is to extract the object spectrum from each image individually and then coadd the single 1D spectra to a combined, deep spectrum. For the UDF HRC grism data, a reduction scheme was developed that forms a coadded, deep 2D spectral image from all the spectra in the individual images and then extracts a single spectrum from that 2D spectral image. This reduction scheme has several advantages, e.g. it allows the correct treatment of cosmic rays and has enhanced flexibility in the combination of sampled data. To create the coadded, deep 2D spectral images, the stsdas/drizzle task was used. For the grism spectrum of each object from the single grism images, transformation coefficient were derived to drizzle those individual 2D spectra onto a single, common drizzle image per object. The transformation coefficients are computed such that the combined drizzle images resemble long slit spectra with the dispersion direction parallel to the x-axis, a constant wavelength scale, and the extraction direction perpendicular to the dispersion direction. This new reduction scheme will form the main part of the extensions to the aXe-package which will be released as aXe-1.4 in summer 2004. e) Extraction of the Deep Spectra from the Drizzled Images ---------------------------------------------------------- The final extraction of the 1D spectra from the 2D drizzled images is done using the existing aXe tasks and a modified configuration that takes into account the modified spectrum of the drizzled images. f) Creation of the preview -------------------------- As a final step an html preview for all spectra was created using the aXe visualization tool aXe2web. The preview pages are available at http://www.stecf.org/UDF/HRCpreview.html, and details concerning aXe2web are given at http://www.stecf.org/software/aXe/index.html#axe2html. 4. Remarks ---------- 4.1 Maximum exposure times -------------------------- The exposure time given in the header of the extracted spectra is the sum of the exposure times of all images that contributed to the spectrum of that object. This number is not the exposure time for each resolution element of the spectrum, but rather the maximum exposure time of the drizzled image, from which the spectrum is then extracted. The exclusion of pixels affected by cosmic rays and the incomplete coverage of spectra in individual frames decreases that maximum exposure time. While the former affects all pixels in a random manner, the latter usually occurs at the lower and upper wavelength end of a spectrum and is reflected in an increase of the corresponding errors. The ultimate test whether a particular object has a uniform exposure time or whether an increase in the error is a result of a patchy coverage is to analyze the weight maps 'h_udfhrce1g800wht_img.fits' (for Epoch-1) and 'h_udfhrce2g800wht_img.fits' (for Epoch-2) at the corresponding position. 4.2 Contamination flag ---------------------- In contrast to conventional (slit) spectroscopy, which cuts out small areas on the sky using slits, slitless spectroscopy delivers many spectra over a whole field of view. This occasionally results in the partial overlap of spectra from two different objects. The row 'CONTAM' in the spectrum tables gives an indication whether there is such a problem at the corresponding wavelength of a spectrum. The CONTAM value in that row is set to '1' if pixels which fall into that wavelength bin were also used to form the extracted spectrum of other objects. Flagged entries do not necessarily have to be discarded as bad data. If, for example, the contaminant spectrum is derived from a fainter object, its influence will be marginal and may be neglected. In the case that the spectrum of a fainter object is completely dominated by the brighter one however, then the contamination-flagged points should be discarded from analysis. To estimate whether a reported contamination has an effect on the spectrum or not, the spectrum of the object must be investigated on the drizzled grism images 'h_udfhrce1g800drz_img.fits' for Epoch-1 and 'h_udfhrce2g800drz_img.fits' for Epoch-2. 5. Further details ------------------ For enquiries on the details of the data reduction, please contact: acshelp@eso.org