Multiscale Systematic Error Correction via Wavelet-Based Bandsplitting in<i>Kepler</i>Data

The previous presearch data conditioning algorithm, PDC-MAP, for the Kepler data processing pipeline performs very well for the majority of targets in the Kepler field of view. However, for an appreciable minority, PDC-MAP has its limitations. To further minimize the number of targets for which PDC-MAP fails to perform admirably, we have developed a new method, called multiscale MAP, or msMAP. Utilizing an overcomplete discrete wavelet transform, the new method divides each light curve into multiple channels, or bands. The light curves in each band are then corrected separately, thereby allowing for a better separation of characteristic signals and improved removal of the systematics.