SCAN data reduction workflow

Processing SCAN data is more complicated than DREAM and STARE data. The challenge is to make an image when the data taken at each bolometer position do not line up exactly with the output map pixels. The fundamental difference is that the primary mirror is scanning the array continuously on the sky. The pixels in the map can therefore be chosen arbitrarily, and bear no relationship with the array of detectors [7,9,8].

SMURF supports two methods of creating a map:

1. A simple rebinning algorithm to assign data to output map pixels, sometimes referred to as a `naive map';
2. An iterative method which solves for a succession of model components assumed to represent the raw data.

The rebinning method may be adequate for bright compact sources (bright enough to be detected in one sample and smaller than the field of view of the array). The rebin method is also suitable for obtaining a `quick-look' image and vital when you wish to make maps of some of the models generated by the iterative map-maker. The SMURF task makemap supports this map-making method.

In practice, the dynamic iterative map maker (DIMM; see Section ) will generally give superior results, though it is more computationally intensive. makemap supports the DIMM.