摘要：NASA's Kepler, K2, and TESS missions employ simple aperture photometry to derive time-series photometry, where an aperture is estimated for each star, and pixels containing each star are summed to create a single light curve. This method is simple, but in crowded fields, the derived time series can be highly contaminated. The alternate method of fitting a point-spread function (PSF) to the data is able to account for crowding but is computationally expensive. In this paper, we present a new approach to extracting photometry from these time-series missions that fits the PSF directly but makes simplifying assumptions in order to greatly reduce the computation expense. Our method fixes the scene of the field in each image, estimates the PSF shape of the instrument with a linear model, and allows only source flux and position to vary. We demonstrate that our method is able to separate the photometry from blended targets in the Kepler data set that are separated by less than a pixel. Our method is fast to compute and fully accounts for uncertainties from degeneracies due to crowded fields. We name the method described in this work linearized field deblending photometry. We demonstrate our method on the false-positive Kepler target KOI-608. We are able to separate the photometry of the two sources in the data and demonstrate that the contaminating transiting signal is consistent with a small, substellar companion with a radius of 2.67 RJup (0.27 R☉). Our method is equally applicable to extracting photometry from NASA's TESS mission.
