TY - JOUR
T1 - The kepler pixel response function
AU - Bryson, Stephen T.
AU - Tenenbaum, Peter
AU - Jenkins, Jon M.
AU - Chandrasekaran, Hema
AU - Klaus, Todd
AU - Caldwell, Douglas A.
AU - Gilliland, Ronald L.
AU - Haas, Michael R.
AU - Dotson, Jessie L.
AU - Koch, David G.
AU - Borucki, William J.
PY - 2010
Y1 - 2010
N2 - Kepler seeks to detect sequences of transits of Earth-size exoplanets orbiting solar-like stars. Such transit signals are on the order of 100 ppm. The high photometric precision demanded by Kepler requires detailed knowledge of how the Kepler pixels respond to starlight during a nominal observation. This information is provided by the Kepler pixel response function (PRF), defined as the composite of Kepler's optical point-spread function, integrated spacecraft pointing jitter during a nominal cadence and other systematic effects. To provide sub-pixel resolution, the PRF is represented as a piecewise-continuous polynomial on a sub-pixel mesh. This continuous representation allows the prediction of a star's flux value on any pixel given the star's pixel position. The advantages and difficulties of this polynomial representation are discussed, including characterization of spatial variation in the PRF and the smoothing of discontinuities between sub-pixel polynomial patches. On-orbit super-resolution measurements of the PRF across the Kepler field of view are described. Two uses of the PRF are presented: the selection of pixels for each star that maximizes the photometric signal-to-noise ratio for that star, and PRF-fitted centroids which provide robust and accurate stellar positions on the CCD, primarily used for attitude and plate scale tracking. Good knowledge of the PRF has been a critical component for the successful collection of high-precision photometry by Kepler.
AB - Kepler seeks to detect sequences of transits of Earth-size exoplanets orbiting solar-like stars. Such transit signals are on the order of 100 ppm. The high photometric precision demanded by Kepler requires detailed knowledge of how the Kepler pixels respond to starlight during a nominal observation. This information is provided by the Kepler pixel response function (PRF), defined as the composite of Kepler's optical point-spread function, integrated spacecraft pointing jitter during a nominal cadence and other systematic effects. To provide sub-pixel resolution, the PRF is represented as a piecewise-continuous polynomial on a sub-pixel mesh. This continuous representation allows the prediction of a star's flux value on any pixel given the star's pixel position. The advantages and difficulties of this polynomial representation are discussed, including characterization of spatial variation in the PRF and the smoothing of discontinuities between sub-pixel polynomial patches. On-orbit super-resolution measurements of the PRF across the Kepler field of view are described. Two uses of the PRF are presented: the selection of pixels for each star that maximizes the photometric signal-to-noise ratio for that star, and PRF-fitted centroids which provide robust and accurate stellar positions on the CCD, primarily used for attitude and plate scale tracking. Good knowledge of the PRF has been a critical component for the successful collection of high-precision photometry by Kepler.
UR - http://www.scopus.com/inward/record.url?scp=77950588328&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=77950588328&partnerID=8YFLogxK
U2 - 10.1088/2041-8205/713/2/L97
DO - 10.1088/2041-8205/713/2/L97
M3 - Article
AN - SCOPUS:77950588328
SN - 2041-8205
VL - 713
SP - L97-L102
JO - Astrophysical Journal Letters
JF - Astrophysical Journal Letters
IS - 2 PART 2
ER -