• Downloaded 5 1-minute raw dark frames
  • run scripts available in svn repository to unpack, split into single chip images, find the median for each chip
  • subtract medians from single epoch dark frames
  • run SExtractor on median-subtracted single epoch dark frames with high thresholds, small minimum area and high deblending minimum contrast (see config file in svn)
    • blink median-subtracted darks and SExtractor "OBJECT" check images to see whether this really catches all CRs and little else
  • run supermongo macro to read in object catalogs, remove groups of >10 objects with identical XMIN and XMAX (because these are typically bad columns)
  • count remaining objects in each chip; make plot of count average per chip and histogram of counts; compare to expectation of ~90 events of cosmic origin at zenith

Results and Interpretation

Running this on 5 dark frames from 09/09/2012, the attached plots show the distribution of counts. Mean counts are approximately 40 / chip / minute when pointing the telescope to the zenith with dome closed. The quoted rate of cosmic origin events is 5 / cm^2 / minute corresponding to 90 / chip / minute. We don't know how the rate of cosmic origin events changes with the dome closed, but since the requirement is that the non-cosmic event rate should be lower than this cosmic event rate that's not actually recessary.