An Independent Analysis of Kepler-4b through Kepler-8b

TL;DR

This paper presents independent, improved analyses of Kepler-4b to Kepler-8b, detecting secondary eclipses, eccentricities, and transit timing variations, advancing understanding of these exoplanets' properties.

Contribution

It introduces new analysis methods, refines planetary parameters, and reports the first detection of secondary eclipses and potential orbital eccentricities for several Kepler planets.

Findings

Detection of secondary eclipse for Kepler-7b with high significance.

Marginal detection of eccentricity for Kepler-4b suggesting an eccentric orbit.

Evidence for transit timing variations in Kepler-6b indicating possible additional bodies.

Abstract

We present two independent, homogeneous, global analyses of the transit light curves, radial velocities and spectroscopy of Kepler-4b through Kepler-8b, with numerous differences over the previous methods. These include: 1) improved decorrelated parameter fitting set, 2) new limb darkening coefficients, 3) time-stamps modified to BJD for consistency with RV, 4) two different methods for compensating for the integration-time of Kepler LC-data, 5) best-fit secondary-eclipse depths and excluded upper limits, 6) fitted mid-transit times, durations, depths and baseline fluxes for individual transits. We make several new determinations: 1) We detect a secondary eclipse for Kepler-7b of depth (47+/-14)ppm and significance 3.5-sigma. We conclude reflected light is a more plausible origin than thermal emission and determine a geometric albedo of Ag=(0.38+/-0.12). 2) An eccentric-orbit model for the Neptune-mass planet Kepler-4b is marginally detected to 2-sigma confidence with e=(0.25+/-0.12). If confirmed, this would place Kepler-4b in a similar category as GJ436b and HAT-P-11b as an eccentric, Neptune-mass planet. 3) We find marginal (2-sigma) evidence for Kepler-5b's secondary eclipse of depth (26+/-17)ppm. The most plausible explanation is reflected light caused by a planet with Ag=(0.15+/-0.10). 4) A 2.6-sigma peak in Kepler-6b TTV periodogram is detected and is not easily explained as an aliased frequency. We find that a resonant/non-resonant perturber, Trojan or exomoon all provide inadequate explanations and the most likely source is stellar rotation. 5) We find different impact parameters relative to the discovery papers in most cases, but generally self-consistent. 6) We constrain the presence of resonant planets, exomoons and Trojans using transit timing.