Dark Monitor (6/1/02)




During normal operations, IRAC will periodically (every six or twelve hours) take single images in all of the exposure modes allowed by the AOT of a dark region of the sky. This will be used to monitor array-wide and readout channel-wide bias drifts. To process this data:
  1. Identify the DCEs. They will be generated with an IER. Make all the frames for a given channel an ensemble.
  2. The data should be processed like science data, except that there is to be no dark subtraction, dark drift, linearization, or flat-fielding applied. Only the basic formatting steps of the pipeline should be run.
  3. Using DETECT, generate a mask file that masks all of the objects found in the longest frame time for a given channel. The IST will provide the appropriate namelists. We will always know what the longest frame time is, since the design of the IER is fixed.
  4. Using the bitmask so generated from the longest frame time, measure the median value of all non-masked bits in a given readout channel for each image. This will produce 4 numbers per image in DN. Also derive the standard deviation.
  5. For every image write these values to a text file in ascii format. A master text file should be stored somewhere in the system so that the IST can always access it, and this text file should be continuously added to during the mission. For pipeline purposes there should also be a database table. The database will be used by a possible future module, whereas the text file is for use by the IST because we already have tools to sort through and examine it. The file should contain the following columns:

A mechanism will also be needed to implement corrections based on this database. This would be a module that takes 4 numbers as input and additively applies them to each readout channel of a DCE. By default such a module would not be turned on in the BCD pipeline, but it may be neccessary pending analysis of the data from the measurement module.