It is generally assumed that charge coupled devices (CCDs) are linear detectors, i.e., that the number of detected counts (after bias subtraction) is proportional to the number of incident photons. However, in many cases the assumption of linearity may not be valid, depending on the science program and the properties of the CCD system.
Correcting for non-linearities is important whenever quantitative intensity measurements need to be made, such as when doing photometry of images or obtaining equivalent widths from spectra. The aim of this paper is to encourage CCD users to consider whether their data could be improved by using a non-linearity correction and to describe methods of determining the correction, with particular application to detectors at Mount Stromlo Observatory, Australia.