The SuperCOSMOS Sky Survey (SSS) consists of machine scans of photographic plates - for more information see http://www-wfau.roe.ac.uk/sss The Sloan Digital Sky Survey (SDSS) consists of more recent CCD camera scans of similar areas of sky with better image quality and greater depth (ie. sensitivity) than the older photographic data - for more information see http://www.sdss.org/ This archive file contains the following files: ReadMe - this file edr.dat - SDSS data for this field UKJ823/sssedrpair.dat - Paired SDSS/SSS object data from the blue passband UKR823/sssedrpair.dat - " " " " " " red " UKI823/sssedrpair.dat - " " " " " " infrared PAE0319/sssedrpair.dat - " " " " " " red " UKJ823/iam.ascii - SSS data (unpaired) UKR823/iam.ascii - SSS data (unpaired) UKI823/iam.ascii - SSS data (unpaired) PAE0319/iam.ascii - SSS data (unpaired) (the PAE files contains early epoch photo data taken in the 1950s whereas the first three use more recent photographic data). In each case, the data are taken from one SSS field (# 823) which is at co-ordinates (0,0) in Right Ascension & Declination. In the paired files, the fields are as follows: Fld Description / units 1 SSS Right Ascension / radians x 10^8 2 SSS Declination / radians x 10^8 3 SSS Object left extent (xmin) / 0.01 micron 4 SSS Object right extent (xmax) / 0.01 micron 5 SSS Object bottom extent (ymin) / 0.01 micron 6 SSS Object top extent (ymax) / 0.01 micron 7 SSS Total area / pixels (pixels are 10x10 micron^2) 8 SSS Brightest intensity above sky / arbitrary intensity units 9 SSS COSMAG instrumental magnitude / magnitude x1000 10 SSS Sky intensity at object centroid / arbitrary intensity units 11 SSS Intensity weighted X centroid / 0.01 micron 12 SSS Intensity weighted Y centroid / 0.01 micron 13 SSS Unweighted semi-major axis / 0.01 micron 14 SSS Unweighted semi-minor axis / 0.01 micron 15 SSS Unwieghted orientation / degrees 16 SSS Weighted semi-major axis / 0.01 micron 17 SSS Weighted semi-minor axis / 0.01 micron 18 SSS Weighted orientation / degrees 19 SSS Classification flag: 1=galaxy, 2=star, 3=unclassified, 4=noise 20 SSS Celestial position angle / degrees 21 SSS Area above areal profile threshold 1 / pixels 22 SSS Area above areal profile threshold 2 / pixels 23 SSS Area above areal profile threshold 3 / pixels 24 SSS Area above areal profile threshold 4 / pixels 25 SSS Area above areal profile threshold 5 / pixels 26 SSS Area above areal profile threshold 6 / pixels 27 SSS Area above areal profile threshold 7 / pixels 28 SSS Area above areal profile threshold 8 / pixels 29 SSS Blend flag: 0=isolated; -n = parent of n objects; +n=nth child of blend 30 SSS Quality flag 31 SSS N(0,1) stellar profile statistic x1000 32 SSS Instrumental profile magnitude / magnitude x1000 33 SSS Calibrated COSMAG (field 9) / magnitude 34 SSS Calibrated profile magnitude (field 32) / magnitude 35 SDSS Right Ascension / decimal degrees 36 SDSS Declination / decimal degrees 37 SDSS g' magnitude 38 SDSS r' magnitude 39 SDSS i' magnitude 40 SDSS classification code / 6 = star, 3 = galaxy In the unpaired files, the fields are as for fields 1 thru 34 in iam.ascii files and as for fields 35 thru 40 for the file edr.dat Fields irrelevant to classification of images in the SSS data are: 1,2,3,4,5,6,11,12,19,20 and probably fields 21 thru 28, since these are coded up in the statistic in field 31. Combinations of fields that will be useful for classification include ellipticity e = 1 - b/a (where b = field 17; a = field 16) where anything with e > 0.33333 will not be an isolated star (but deblended stars can have e > 0.33333 when near much brighter objects...) and for bright galaxies (probably not a major consideration for this project) filling factor G = area / (pi*a_u*b_u) (where area=field 7; a_u=13; b_u=14) where bright objects will never be stars if G < 0.6 and are potentially galaxies when G > 0.9 (bright means calibrated magnitude < 17 in this context). Note also that for bright objects, the appropriate magnitude to use (either field 33 or 34) is class-dependent: if the object is stellar-like, then field 34 is the one to use; if the object is not stellar like, then 33 is more appropriate. This is not a major consideration for faint objects, since the calibrations converge. To see how good our default classification is, simply compare the SDSS class (field 40) with the SSS class (field 19). For more information on the SSS data, see the set of papers at http://www-wfau.roe.ac.uk/sss/docs.html - particularly Paper II.