Binning is sinning: morphological light-curve distortions due to finite integration time

TL;DR

Finite integration times in light-curve observations distort transit shapes, leading to inaccurate system parameters; this paper provides analytic tools to estimate and correct these effects, improving exoplanet characterization.

Contribution

The paper introduces analytic expressions and numerical techniques to quantify and correct morphological distortions caused by finite integration times in transit light-curves.

Findings

Long integration times systematically underestimate stellar density.

Analytic formulas enable quick error estimation for light-curve fitting.

Correcting for integration effects yields consistent transit parameters across datasets.

Abstract

We explore how finite integration times or equivalently temporal binning induces morphological distortions to the transit light-curve. These distortions, if uncorrected for, lead to the retrieval of erroneous system parameters and may even lead to some planetary candidates being rejected as ostensibly unphysical. We provide analytic expressions for estimating the disturbance to the various light-curve parameters as a function of the integration time. These effects are particularly crucial in light of the long-cadence photometry often used for discovering new exoplanets by, for example, Convection Rotation and Planetary Transits (COROT) and the Kepler Mission (8.5 and 30 min). One of the dominant effects of long integration times is a systematic underestimation of the light-curve-derived stellar density, which has significant ramifications for transit surveys. We present a discussion of numerical integration techniques to compensate for the effects and produce expressions to quickly estimate the errors of such techniques, as a function of integration time and numerical resolution. This allows for an economic choice of resolution before attempting fits of long-cadence light-curves. We provide a comparison of the short- and long-cadence light-curves of TrES-2b and show that the retrieved transit parameters are consistent using the techniques discussed here.