Abstract. We present the method of the Extended Aperture Photometry (EAP) that we applied on K2 RR Lyrae stars. Our aim is to minimize the instrumental variations of attitude control maneuvers by using apertures that cover the positional changes in the field of view thus contain the stars during the whole observation. We present example light curves that we compared to the light curves from the K2 Systematics Correction (K2SC) pipeline applied on the automated Single Aperture Photometry (SAP) and on the Pre-search Data Conditioning Simple Aperture Photometry (PDCSAP) data.
The need for custom photometric pipelinesK2 observations suffer from various instrumental effects. The most serious one is the variation in spacecraft attitude that is corrected on a 6-hour timescale. Various methods have been developed to tackle these issues but neither is really optimal for RR Lyrae stars. Main characteristics of RR Lyrae photometry are the ∼6-12 hour periods (in the range of the correction frequency) and sharp features in the light curve (false positives for outliers). Also, we are interested in the full-amplitude variation, and not in a detrended (systematics-corrected) light curve that other pipelines offer. The KepSFF pipeline [5], for example, is confused by the pulsation, and detrended light curves are of lower quality than the raw fluxes, in several cases. The K2SC pipeline [1] removes the position-dependent systematics from the SAP and PDCSAP light curves. PDCSAP data are already free from systematic trends removed by the PDC (Pre-search Data Conditioning) method [4].Here we compare our solution for RR Lyrae stars with the outputs of K2SC pipeline. In many cases, EAP photometry retains the full amplitude much better than SAP/PDCSAP-based light curves, but it is sensitive to contamination from blended sources.
A simple way: Extended Aperture Photometry (EAP)We created new apertures for every star. The apertures were defined to contain the movements of the stars within the target pixel masks, but to avoid contamination from nearby stars whenever possible. Although this method produces a very slightly increased noise from the background pixels, in many cases it also preserves the pulsation e-mail: eplachy@konkoly.hu amplitudes throughout the campaign much better than the SAP and PDCSAP fluxes. After the photometry, we removed all points with SAP_QUALITY flags larger than 0, except for C2, where we also kept the large number of points with flags value of 2 15 (16384).We then ran an automated Fourier analysis script on the light curves to construct an initial fit for sigma clipping [2]. The fit was subtracted from the data, all >3 sigma outliers from the residuals were removed, then the fit was added back. Figure 1 displays four examples of RR Lyrae stars from Campaign 4.
Examples
EPIC 210438688 (K p = 17.387 mag)The aperture used by the Kepler pipeline is clearly undersized. The SAP data indicates that the PSF periodically moved out of the SAP aperture during the second half of the campaign, leading to significant flux l...