Images taken with modern detectors require calibration via flat fielding to obtain the same flux scale across the whole image. One method for obtaining the best possible flat fielding accuracy is to derive a photometric model from dithered stellar observations. A large variety of effects have been taken into account in such modelling. Recently, Moehler et al. (2010) discovered systematic variations in available flat frames for the European Southern Observatorys FORS instrument that change with the orientation of the projected image on the sky. The effect on photometry is large compared to other systematic effects that have already been taken into account. In this paper, we present a correction method for this effect: a generalization of the fitting procedure of Bramich & Freudling (2012) to include a polynomial representation of rotating flat fields. We then applied the method to the specific case of FORS2 photometric observations of a series of standard star fields, and provide parametrised solutions that can be applied by the users. We found polynomial coefficients to describe the static and rotating large-scale systematic flat-field variations across the FORS2 field of view. Applying these coefficients to FORS2 data, the systematic changes in the flux scale across FORS2 images can be improved by ~1% to ~2% of the total flux. This represents a significant improvement in the era of large-scale surveys, which require homogeneous photometry at the 1% level or better.
Download