The Monitor project: Data processing and lightcurve production

Abstract

We have begun a large-scale photometric survey of nearby open clusters and star-forming regions, the Monitor project, aiming to measure time-series photometry for >10,000 cluster members over >10 sq.deg of sky, to find low-mass eclipsing binary and planet systems. We describe the software pipeline we have developed for this project, showing that we can achieve peak RMS accuracy over the entire data-set of better than 2 mmag using aperture photometry, with RMS < 1% over 4 mag, in data from 2 and 4m class telescopes with wide-field mosaic cameras. We investigate the noise properties of our data, finding correlated `red&#39; noise at the 1 - 1.5 mmag level in bright stars, over transit-like timescales of 2.5 hours. An important source of correlated noise in aperture photometry is image blending, which produces variations correlated with the seeing. We present a simple blend index based on fitting polynomials to these variations, and find that subtracting the fit from the data provides a method to reduce their amplitude, in lieu of using techniques such as point spread function fitting photometry which tackle their cause. Finally, we use the Sysrem algorithm to search for any further systematic effects.