Chris, Here is a description of the SCUBA2 Cosmology Legacy Survey, very closely following the form of the text Douglas wrote for SASSY. I could make this a lot shorter if you would prefer. Thanks. Mark. _______________________________________________________________ The SCUBA2 Cosmology Legacy Survey CANFAR project ------------------------------------------------ Abstract -------- The SCUBA2 Cosmology Legacy survey is a large and thorough survey of structure formation and star formation history. It has been awarded over 1,100 hours of observing time, including over half of the anticipated best weather time in the frist two years. This good weather time is required to perform the 450 micron parts of the survey. It is not an exaggeration to state that the major reason that SCUBA2 has been built is to perform this survey. Our survey will probe sufficiently large volumes, and will yield a sufficiently large robust sample of sources with meaningful redshift information to allow the clustering strength of submm galaxies to be accurately measured, and the link between sub-mm galaxies and dark-matter halos to be established as a function of redshift. Our survey will allow the detailed study of the high-redshift progenitors of the richest structures seen in the present-day universe (e.g. the Coma cluster). The survey provides enough dynamic range in luminosity, a sufficient number of sources, adequate redshift information, and the necessary SED information for calculation of bolometric luminosities to allow the cosmological evolution of the sub-mm luminosity function to be delineated. This in turn will allow us to refine the Lilly/Madau diagram, and establish the link between obscured and visible starformation over cosmic history. Lastly, the survey allows the exploitation of a wide range of multi-frequency supporting datasets to allow the detailed study of the properties of sub-mm selected galaxies, and hence crucial tests of semi-analytic/numerical models of galaxy formation. The success of this survey requires coordination among a large science team spread over the world, facilities for easy comparison to data collected at other wavelengths and that map-making, calibration and astrometry be caried out with a level of precision not previously achieved with submillimetre data. ------------- The Cosmology Legacy Survey is a collaboration of 100 scientists located in Canada, the UK, the Netherlands, the US, Mexico, Germany, Austria and Australia. The survey will be conducted using SCUBA2 at the JCMT, and raw data will be archived at the CADC. We anticipate perhaps 100 Tera-bytes of raw data from the first two years. The primary data product of the survey is two-colour maps of the fields. These and other data products will also be accesible through the CADC. Becasue of the very large amount of raw data required to make comparitively few maps, a collaborative but centralized map-making computational programm will be required. Developing this collaboration is one of the tasks we hope this CANFAR program will support. Many of the archiving and computational issues are similar to those of SASSY, and so here we will just highlight some of the differences. First, the short wavelength camera, working at 450 microns, is central to many of the survey science goals and a lot of effort and observing time will be dedicated to producing 450 micron maps. Systematic effects are expected to be appreciable, the atmosphere is bright, and the array is undersampled. Map making will require new tools and very different rules for data editing, etc., than those we have become used to at 850 microns. The smaller beam size will require much more care with astrometry than has been given to SCUBA maps to date. Second, both at 450 microns and at 850, we will make very deep maps, with many many visits to each part of the sky. SCUBA2 offers, and requires, a new level of care and accuracy in map-making and there will be a substantial effort by the team to quantify the effectiveness of the data reduction pipeline. This sort of work has been attempted by comparitively small teams such as those working on WMAP or BLAST. This cosmology survey persents a new challenge in effective collaboration becasue of the size of the data set, the size of the science team and the care required. Third, unlike SASSY which is largely aimed at finding objects, a large part of the cosmology survey goals are met through detailed study of the objects found. In the submillimetre, this has been shown to require cross comparison to data taken at almos all other wavelengths, from radio to x-ray. We will need to develope techniques for astrometry and calibration which work well over the large areas of our maps. Time awarded ------------ In the first two years of SCUBA2 operation, the Cosmology Legacy Survey has been awarded 1120 hours. 490 hours of this is allocated to take place during the best 10% of the weather comprising 2/3 of the good weather time anticipated. The proposal contained plans for five years of operation which have not been adjudicated yet, but we anticipate operations at a similar level after the initial two year period. Expected resources ------------------ Like SASSY, the Cosmology Legacy Survey will require substantial computing resources. After some calibration and data editing steps, we anticipate that map making will compress 10s of TB of raw data into maps which are "only" a few hundred MB in size. This is likely to be performed iteratively, and in a single iteration each data point only must be read once. We will need to be able to make simple maps with only a few iterations and perhaps some filtering of the input time stream, fairly rapidly so that the science team can understand progress of the survey. Our best maps are likely to reqire that we not filter the data heavily, but instead iterate the map-makers many times. Over 100 iterations is common. This will be slow, even with a large computer cluster. The goal is to develope a system which can converge a factor of several times more rapidly than we collect data. Making a high fidelity map in several weeks is adequate. Experience with BLAST and WMAP indicate that the team will need to make more maps with simulated data than with real data in order to develope and understand the process. Major processing steps ---------------------- The final map making process for any field in the survey takes place after all observations of the field are complete, perhaps ten or 20 nights of observing. However, the survey team will need prompt access to the data and some interim processing, including map making, in order to monitor the survey and adjust the strategy. APPENDIX Science Use Cases for CLS --------------------------- This appendix describes some example tasks survey scientists need to carry out to achieve the goals of the Cosmology Legacy Survey. 1. Example Science Use Case (SUC): Timely access to data -------------------------------------------------------- - Access to raw data within 12 hours of completed observations - Transfer raw data to dedicated DR machine with tools for analysis - A mechanism is needed for approval by multiple users - A mechanism is needed for rapid sharing of this prompt analysis and comments on it with team memers. 2. Example SUC: Monitor survey progress --------------------------------------- - Ability to determine list of completed observations: - Show relation between survey field ID and JCMT data files - Tabulate completion statistics - Visual representation of completion statistics - Visual representation of progress: - ability to plot completion on map of sky - ability to plot region of planned 2/5 year coverage - ability to plot/identify imminently observable fields - completion map should indicate number of times a field has been observed. -Ability to compare data quality to coverage goals: - make maps of anticipated variance in the collected data - make maps of coverage versus weather conditions or detector performance metrics - ability to track trends in various performance metrics. -ability to search large quantities of the data for anomolies. 3. Example SUC: Monitor and document the quality of the collected raw data. --------------------------------------------- - Need to develope a set of tools to assess data quality - Compare photometric calibration to detector model. - Model pointing offsets as a function of time, elevation, atmosphere,.. - Ability to register feedback/notes/comments regarding data quality - Contact survey coordinators/data handlers - Notify survey mailing list of updates/new data The Science Use Cases 5 and 6 listed for SASSy are required here too. -- Mark Halpern UBC Physics and Astronomy (604)822-6709