The Possible Moon of Kepler-90g is a False Positive

TL;DR

This study investigates a potential exomoon signal near Kepler-90g, finds it to be a false positive likely caused by instrumental effects, and emphasizes the need for careful analysis in exomoon detection.

Contribution

The paper introduces a novel pixel-level analysis method called the 'transit centroid' to distinguish true exomoon signals from false positives caused by instrumental artifacts.

Findings

The candidate exomoon signal is likely a false positive.

Pixel-level analysis reveals the signal occurs across nearly all pixels.

Instrumental effects such as SPSD can mimic exomoon signals.

Abstract

The discovery of an exomoon would provide deep insights into planet formation and the habitability of planetary systems, with transiting examples being particularly sought after. Of the hundreds of Kepler planets now discovered, the seven-planet system Kepler-90 is unusual for exhibiting an unidentified transit-like signal in close proximity to one of the transits of the long-period gas-giant Kepler-90g, as noted by Cabrera et. al. (2014). As part of the 'Hunt for Exomoons with Kepler' (HEK) project, we investigate this possible exomoon signal and find it passes all conventional photometric, dynamical and centroid diagnostic tests. However, pixel-level light curves indicate that the moon-like signal occurs on nearly all of the target's pixels, which we confirm using a novel way of examining pixel-level data which we dub the 'transit centroid'. This test reveals that the possible exomoon to Kepler-90g is likely a false positive, perhaps due to a cosmic ray induced Sudden Pixel Sensitivity Dropout (SPSD). This work highlights the extreme care required for seeking non-periodic low-amplitude transit signals, such as exomoons.