The science goals for ground-based large-area surveys, such as the Dark Energy Survey, Pan-STARRS, and the Large Synoptic Survey Telescope, require calibration of broadband photometry that is stable in time and uniform over the sky to precisions of a per cent or better. This performance will need to be achieved with data taken over the course of many years, and often in less than ideal conditions. This paper describes a strategy to achieve precise internal calibration of imaging survey data taken in less than photometric conditions, and reports results of an observational study of the techniques needed to implement this strategy. We find that images of celestial fields used in this case study with stellar densities of order one per arcmin-squared and taken through cloudless skies can be calibrated with relative precision of 0.5 per cent (reproducibility). We report measurements of spatial structure functions of cloud absorption observed over a range of atmospheric conditions, and find it possible to achieve photometric measurements that are reproducible to 1 per cent in images that were taken through cloud layers that transmit as little as 25 per cent of the incident optical flux (1.5 magnitudes of extinction). We find, however, that photometric precision below 1 per cent is impeded by the thinnest detectable cloud layers. We comment on implications of these results for the observing strategies of future surveys.
Download