DELIVERY AND ARCHIVING OF THRESHOLD-CROSSING-EVENT AND KEPLER-OBJECT-OF-INTEREST TABLES FOR THE EXTENDED KEPLER MISSION

The Kepler spacecraft collects nearly continuous time-series photometry from a single field-of-view for the purpose of identifying and characterizing transiting planet candidates. The increasing baseline of Kepler data provides increasing sensitivity to smaller planets with longer orbital periods. However, the longer baseline also increases the number of astrophysical and non-astrophysical artifacts impeding detection of the planets in the most interesting regions of parameter space. The process of distinguishing planetary candidates from false positive detections requires careful analysis of both the flux time series and pixel data provided by Kepler (see Batalha et al. 2012 for examples). This process produces well-vetted catalogs of Kepler Objects of Interest (KOIs) in refereed publications, but introduces substantial delays between the time when data is gathered and when the results are distributed.

In order to substantially reduce the time between data gathering and the distribution of results, the transition between the Kepler primary mission (2009-2012) and the Kepler extended mission (2012-) implements a major change in the data delivery process and community participation. The previous primary mission practice of delaying release of a KOI catalog until the accompanying catalog paper has been submitted to a refereed journal has ended. The new extended mission process is designed to eliminate most of the delay between data gathering and release of potential planet candidates detected with Kepler. To do so, the Kepler project will conduct the process of KOI identification, disposition into planetary candidates and false positives, and characterization of physical planet parameters openly within the NASA Exoplanet Archive. Community scientists can follow the process in real time from the initial delivery of pipeline products, through identification of KOIs from the pipeline products, dispositioning of KOIs into planet candidates and false positives, and uniform data modeling.

The earliest form of a planet candidate results from the identification of a potential transit signal by the Kepler pipeline in the form of a Threshold Crossing Event (TCE). Each TCE is designated by the target hosting the potential planet candidate and its ephemeris. During the extended mission, the project will release comprehensive tables of TCEs identified by periodic pipeline runs. The statistical thresholds passed by a TCE can differ between pipeline runs and the user of the TCE tables should consult the relevant TCE release notes for details. In addition to identifying the TCEs, the pipeline performs many automated statistical tests on the Kepler flux time series and pixel-data (with the Data Validation (DV) module of the pipeline) in order to provide information that helps to discriminate between TCEs that are planet candidates and those that are false positives. The DV data products are delivered at the same time as the TCE tables.

The first examination of new TCEs, known as triage, is designed to quickly eliminate the obvious instrumental false positives and to identify previously known KOIs. Once these have been separated out, the residual population of astrophysically interesting TCEs require more scrutiny, so they are declared new KOIs. The second step focuses on these KOIs and performs a comprehensive analysis to disposition them into planet candidates and false positives. A planetary candidate is a KOI that has passed all prior tests of the null hypothesis that the target is a planetary candidate, although this does not a priori mean that all possible tests have been conducted. False positives typically occur when i) the KOI is in reality an eclipsing binary star, ii) the Kepler light curve is contaminated by a background eclipsing binary, iii) stellar variability is confused for coherent planetary transits, or iv) instrumental artifacts are confused for coherent planetary transits. The new pipeline product deliveries in the extended mission are:

The delivery of these pipeline products to the NASA Exoplanet Archive is a requirement of the Kepler extended mission. The backend of the pipeline producing these products will be run 1-2 times per year. Products will be delivered to the archive without delay, before the TCE table has been analyzed by the Kepler Project. The first extended mission delivery of a TCE table and DV reports will be from Q1-Q12 pipeline processing. Typically, each pipeline run produces 10,000-20,000 TCEs. Users of the Q1-Q6 KOI table and Q1-Q8 KOI tables at the Exoplanet Archive compiled during the primary mission should note that the pipeline products from which the previously reported KOIs were derived will not be available.

Following the delivery of pipeline products to the Exoplanet Archive, the Kepler Project will begin searching the new list of TCEs for new and existing KOIs. New and old KOIs will be delivered to a corresponding new KOI table at the archive on a weekly basis, without disposition into planetary candidates or false positives. This exercise is known as "triage".

As vetting tests for the null hypothesis that a TCE is a planet are performed, the "disposition" of each KOI as either a planet candidate or false positive will be updated and, most importantly, may change over time. It is therefore critical that the scientific community not conduct sample completeness studies on KOI tables that remain "active". Active tables do however provide the latest information for community scientists interested in follow-up observations and disposition activities. After a period of activity, the classification of the KOI table will changed from "active" to "done" when all dispositions are judged as final and all model parameters have been updated appropriately. This will typically occur after a new delivery of TCEs to the archive based upon a longer data baseline. Statistical studies should not be attempted upon KOI samples until the table has been classified as "done" and the specific nature of the sample is understood. In particular the multiple goals of the Q1-Q8 KOI table make certain statistical sample analyses difficult.

Scientific opportunities for the community include:

1. Identifying new transit-like events not currently included in the delivered TCE tables from the light curve and pixel repository at the MAST.
2. Identifying new planetary candidates from the TCE tables and DV products.
3. Dispositioning planetary candidates and false positives from the KOI tables.
4. Identifying targets of interest for ground- and space-based follow-up observations from the TCE and KOI tables.
5. Harvesting new yields of TCEs, KOIs, planetary candidates and planetary false positives independent of the Kepler Project. Independent yields are critical for assessing the reliability of the Kepler pipeline and community work.