Data reduction pipelines

Automated data reduction is a key to successfully exploit the wealth of data returned by autonomous observatories. This section excels at photometric data reduction, in particular data obtained by standard CCD (Charge Coupled Device). The principle path from raw CCD data to scientific useful data can be best depicted in the following flow chart:


Principle steps in reducing CCD data. Top row of steps are applicable to any optical CCD data, bottom row are specific to imaging, though flatfielding and gain adjustment may also be found in, e.g., spectroscopic data reduction.

Not all of the steps outlined above are necessary for all detectors and all use cases. Some reduction steps may include different algorithms or may work on a different set of calibration data. The idea behinde a pipeline is to automize all this steps, to get from raw data to right-away useable reduced data without manual interference.

Though a demanding tasks on it's own, we may dare to ask the question: If there are more than a single route to the achievement of a single subtask, why not let the user specify the method used? In an automated environment, the use may be offered to request certain data reduction steps on observing block upload. In a more user-driven scenario, it may be left to the user to specify (or even try out) different algorithms when prompted with her raw data. The latter approach was picked up by ESO in their Reflex environment - an environment that let the user specify different reduction steps in a graphical user interface.

Last update: 25. August 2021