In order to maintain long-term stable pointing upon a single field, the Kepler spacecraft is in an Earth-trailing, 372 day heliocentric orbit. Kepler’s 0.95-m aperture Schmidt telescope carries a photometer with an array of 42 CCD chips (Koch et al. 2010). Each CCD has a direct neighbor, and collectively they are referred to as a “module”, of which there are 21. Each module has 4 output nodes. Therefore the CCD array and targets falling upon silicon are often mapped according to module and output numbers. Alternatively, mission documentation also refers to output nodes by “channel” number, which ranges from 1–84. The module, output and channel locations are provided for each target within the archived meta-tables at the Multi-Mission Archive at STScI (MAST). The Kepler Instrument Handbook (Van Cleve & Caldwell 2009) maps module, output and channel numbers to the detector array. The field of view and pixel scale were designed to maximize the number of resolvable stars brighter than Kp = 15. Kp refers to the AB magnitude (Brown et al. 2011) across the Kepler 425-900 nm bandpass. The Kepler field of view spans 115.6 square degrees over 94.6 million 3.98 × 3.98 arcsec detector pixels, with a 3.1-7.5 arcsec full-width half-maximum point spread function. The Kepler field contains 10 million stars brighter than the confusion limit of Kp = 20-21. The camera takes one 6.02s image across the full field every 6.54 s. Exposures are summed onboard and stored at either long cadence (29.4 min; see Jenkins et al. 2010) or short cadence (58.85 s; see Gilliland et al. 2010). Science data are downloaded approximately once per month when Kepler leaves the field temporarily to point its high-gain antenna towards Earth (Haas et al. 2010). The number of pixels collected and transmitted is a trade-off between maximizing the number of targets delivered and minimizing the length of the data gaps. For long cadence observations, a maximum of 5.4 × 10⁶ pixels are stored onboard in the spacecraft Solid State Recorder, and the number of targets typically range from 160,000 to 170,000. Short cadence data are limited to 512 targets. Download requires approximately a 24-hour hiatus in data collection.

The long and short cadence pixels equate to less than 6% of the detector plane, and the remaining pixel data are not stored. The stored pixels are chosen strategically to provide postage stamp images centered on the positions of Kepler targets (Batalha et al. 2010). The critical concept for understanding instrumental artifacts is that in order to maximize the number of targets collected, the postage stamp sizes and shapes are chosen to maximize the per-target photometric signal-to-noise on the 3-12 hour timescales of exoplanet transits. The size of a postage stamp increases with target brightness. The postage stamps do not contain all the flux from a target because the collection of the target’s faint PSF wings degrades signal-to-noise by the inclusion of more sky background. The pixels within a postage stamp are defined by a calculation that combines the photometry and astrometry within the Kepler Input Catalog (KIC; Brown et al. 2011) and an analytical pixel response model for the detector and optics (Bryson et al. 2010). The postage stamps are fixed within the pixel array; they do not evolve over a 93-day observation period, or “quarter”. A new target list is uploaded to the spacecraft after each quarterly roll. Changes in the target list occur due to new detector geometry, operational developments to the exoplanet survey, and community-led science programs. Any time-dependent variation in the position of the target or the size of the point spread function will result in a redistribution of flux within the postage stamp pixels. The spacecraft pointing stability is good to typically 20 milliarcsec over 6.5 hours but the high precision light curves can contain systematic noise that manifests from thermally-driven focus variations, pointing offsets, and differential velocity aberration (Christiansen et al. 2011; Van Cleve & Caldwell 2009). Many of the systematics within the archived light curves are the result of time-dependent light losses as the target wings fall out of the pixel apertures and time-dependent contamination by neighboring sources falling into the pixel apertures. All collected data are stored and propagated to the community by the MAST. Technical manuals are:

• Kepler Instrument Handbook - describes the design, operation, and in-flight performance of the Kepler spacecraft, telescope, and detector.
• Kepler Data Processing Handbook - provides a description of the algorithms and pipelined data processing performed upon collected data.
• Kepler Archive Manual - describes the content and format of archived Kepler data products and the available archive search and retrieval resources.
• Kepler Data Characteristics Handbook - defines the causes and provides illustrative examples of characteristics within the Kepler time-series data and systematic artifacts found therein.
• Kepler Data Release Notes - provide an impact assessment of systematic artifacts and spacecraft events upon Kepler time-series data.

Within the MAST archive, Kepler data are stored as files designed to the format and conventions of the Flexible Image Transport System (FITS; Pence et al. 2010). Short cadence target data files each contain one month of observations, and long cadence data files contain one quarter of observations. Instructions for the MAST data search and retrieval tools are provided in the Kepler Archive Manual. The three primary forms of Kepler science data stored within the archive are:

• Full-Frame Images (FFIs) - using all the pixels on each detector channel, the Kepler spacecraft accumulates one 29.4 min image comprised of a sequence of 270 consecutive 6.02s exposures co-added onboard, once per month, before each data download. The 84 channels are stored within a single FITS file. FFIs are collected primarily for engineering purposes but provide a scientific resource in their own right – high-precision photometry of the entire Kepler field of view on a 1-month cadence. Additionally, FFIs can be employed to assess the target’s flux level and sources of contamination from nearby objects within a target’s pixel aperture.
• Target Pixel files (TPFs) - each TPF stores a time-stamped sequence of uncalibrated and calibrated postage stamp pixel images of a Kepler target over a long cadence quarter or short cadence month.
• Light curve files - each file contains time-series photometry for an individual Kepler target, derived from an optimized subset of pixels contained within the associated TPF.

Previous: Introduction Up: PyKE Primer Next: Target Pixel Files