An exomoon survey of 70 cool giant exoplanets and the new candidate Kepler-1708 b-i

Abstract

Exomoons represent a crucial missing puzzle piece in our efforts to understand extrasolar planetary systems. To address this deficiency, we here describe an exomoon survey of 70 cool, giant transiting exoplanet candidates found by Kepler. We identify only one exhibiting a moon-like signal that passes a battery of vetting tests: Kepler-1708 b. We show that Kepler-1708 b is a statistically validated Jupiter-sized planet orbiting a Sun-like quiescent star at 1.6 au. The signal of the exomoon candidate, Kepler-1708 b-i, is a 4.8σ effect and is persistent across different instrumental detrending methods, with a 1% false-positive probability via injection-recovery. Kepler-1708 b-i is ~2.6 Earth radii and is located in an approximately coplanar orbit at ~12 planetary radii from its ~1.6 au Jupiter-sized host. Future observations will be necessary to validate or reject the candidate.

Keywords: Exoplanets; Rings and moons.

Fig. 1. Orbital properties of the 70 cool giants.

A comparison of the derived orbital eccentricities from this work (y axis) versus the orbital periods (previously known) for our planetary-candidate sample. We use colour (legend) to depict the dynamical ‘temperature’ via the inferred absence/presence of TTVs. Source data

Fig. 2. Transit light curves of Kepler-1708 b.

The left/right column shows the first/second transit epoch, with the maximum-likelihood planet–moon model overlaid in solid red. The grey line above shows the contribution of the moon in isolation. Lower panels show the residuals between the planet–moon model and the data, as well as the planet-only model. BJD, barycentric Julian date; UTC, coordinated universal time. Source data

Fig. 3. FPP calculation for Kepler-1708 b-i.

Histogram of the log Bayes factor between a planet–moon and a planet-only model from 200 fake planet-only signals injected into the light curve. Two signals pass the threshold (=log_e10) and have positive radii, indicating a 1% FPP. Source data

Extended Data Fig. 1. Probability distribution of the cool giant’s eccentricities.

Left: We extract a random draw from the eccentricity posterior distribution of each planet and apply a smooth kernel density estimator (KDE) to the sample with a Gaussian kernel. Each line represents 1 of 100 such realisations. Right: Credible intervals evaluated using 10⁵ such samples as computed in the left panel. Source data

Extended Data Fig. 2. Transit light curves of KIC-8681125.01 for the first (left) and second (right) epochs.

Top: Each panel shows the method marginalised detrended photometry centred on the times of transit, with the maximum likelihood planet–moon fit overlaid in solid black. Model comparison statistics are provided within the inset box. Bottom: Same as above but for a model with a single planet and variable blend factor between the two epochs. This model substantially outperforms the planet–moon model. Source data

Extended Data Fig. 3. Transit light curves of KIC-5351250.06/Kepler-150f for the first (left) and second (right) epochs.

Each row shows a different model fit to the same data. Whilst the planet–moon model is clearly a better fit than the planet-only model, a 2-spot model is able to out-perform either and is well motivated from the activity levels observed in the out-of-transit light curve. Source data

Extended Data Fig. 4. Detrended transit light curves of KIC-7906827.01 for the first (left) and second (right) epochs.

Each row shows a different combination of light curve detrending method and input data, which are combined to build the method marginalised product. For each, we overlay the maximum a posteriori planet–moon model as conditioned upon the method marginalised light curve, and a comparison of how much better it matches the data versus the planet-only model, in a chi-squared sense. Source data

Extended Data Fig. 5. Pixel-level comparison the two transits of KIC-7906827.01.

Left: Pixel log-intensity is shown for the postage stamp downloaded for KIC-7906827 from the Kepler spacecraft, for epochs 1 (top) and 2 (bottom). The black solid outline shows the optimal aperture selected by the Kepler pipeline. Middle: Same as the left, except we show the signal to noise ratio (SNR) of the planetary transit signal in each pixel. As expected, the transit signal is colocated with the brightest source in view. Right: Same as the middle, except we show the chi-squared improvement of the planet–moon model over the planet-only model in each pixel light curve after local detrending. As expected, the moon signal appears colocated with the target. Source data