Vetting and False Positive Analysis of TOI 864.01: Evidence for a Likely Hierarchical Eclipsing Binary Masked by Dilution

We present a detailed vetting analysis of the TESS candidate TOI 864.01, initially identified as a potential ultra-short-period (P ~ 0.52 d) Earth-sized planet orbiting an M-dwarf. Using 12 sectors of TESS photometry spanning a multi-year baseline, we recover a robust periodic transit-like signal. While the recovered transit depth is attenuated by detrending (~ 158 ppm), the SPOC pipeline reports an undiluted depth of ~ 640 ppm. Stellar characterization based on Gaia DR3 astrometry yields a nominally single-star solution (RUWE = 1.18), highlighting the limitations of astrometric vetting for tight companions. We performed a statistical validation analysis using TRICERATOPS, aggregating data from all 12 available sectors. The analysis yields a False Positive Probability (FPP) of 0.25 and a Nearby False Positive Probability (NFPP) of < 10^-4. While these metrics ostensibly classify the target as a viable planetary candidate based on Gaia-resolved sources, they fail to account for sub-pixel companions. Crucially, archival high-resolution imaging from the TESS Follow-up Observing Program (TFOP SG1) reveals a stellar companion at a separation of 0.04", unresolved by both Gaia and TESS. When this companion is considered, the signal is best interpreted as a Hierarchical Eclipsing Binary (HEB) on the companion. Bayesian model comparison yields an inconclusive result (Delta ln Z ~ 0.25), consistent with the degeneracy introduced by unresolved blending. Ground-based follow-up photometry further supports significant dilution, with a measured transit depth (~ 0.37 ppt) shallower than predicted (~ 0.64 ppt) and a timing offset of 6.3 minutes. Taken together, we classify TOI 864.01 as a probable False Positive and recommend its retirement from planetary candidate lists.

💡 Research Summary

This paper presents a comprehensive vetting of the TESS object of interest TOI 864.01, initially flagged as an ultra‑short‑period (USP) Earth‑sized planet candidate around an M dwarf. Using twelve TESS sectors spanning more than two years, the authors recover a robust periodic signal with a period of 0.52067 days and an undiluted transit depth of roughly 640 ppm (0.64 ppt). The depth measured after aggressive detrending is only ~158 ppm, but the official SPOC depth is adopted for physical interpretation.

The phase‑folded light curve shows a V‑shaped transit without a flat bottom, a morphology more typical of grazing eclipses or heavily diluted binaries than of a small planet. Centroid analysis on the TESS data shows no significant motion, which only rules out distant background eclipsing binaries; it cannot detect companions closer than ~1″. Gaia DR3 reports a RUWE of 1.18, suggesting a single‑star solution, yet this metric is insensitive to very close, high‑contrast companions.

Statistical validation with the TRICERATOPS package, which uses only Gaia‑resolved sources, yields a False Positive Probability (FPP) of 0.2509 and a Nearby FPP of 7 × 10⁻⁵. By conventional thresholds (FPP < 0.5) the target would be retained as a planetary candidate. However, archival high‑resolution imaging from the TESS Follow‑up Observing Program (TFOP SG1) reveals a stellar companion at a separation of just 0.04″, completely unresolved by both Gaia and TESS. This sub‑pixel companion introduces severe photometric dilution. Ground‑based follow‑up from LCO‑CTIO measured a transit depth of 0.37 ppt—only about 58 % of the expected depth for an undiluted planetary transit—directly confirming the dilution scenario.

Bayesian model comparison using the juliet package fits both a planetary model and an eclipsing‑binary model to the TESS light curve. The log‑evidence difference Δln Z ≈ 0.25 is far below the |Δln Z| = 2 threshold for a decisive preference, indicating that the photometry alone cannot discriminate between a small planet and a diluted binary when a bright contaminant is present.

Additional evidence comes from timing analysis. Multiple ground‑based observations over a ~1.5‑year baseline show transit mid‑times delayed by 6–15 minutes relative to the ephemeris derived from TESS. Such variations are atypical for a stable USP planet but are consistent with light‑travel‑time effects or apsidal precession in a hierarchical eclipsing binary (HEB) system.

Assuming a single‑star scenario, the inferred planetary radius would be ~1.1 R⊕, but this is an artifact of the diluted depth (δ_obs ≈ δ_true/(1 + Dilution)). The true eclipsing object is likely a faint M dwarf companion whose deep eclipses are heavily diluted by the primary star’s flux, making the event appear planetary.

The authors acknowledge limitations: only a single epoch of ground‑based photometry is available, multi‑band observations would better characterize chromaticity; the Bayesian evidence remains inconclusive without resolved light curves for each component; and radial‑velocity measurements are lacking, preventing a definitive exclusion of exotic scenarios. Nonetheless, the convergence of high‑resolution imaging, depth discrepancy, and timing offsets makes the hierarchical eclipsing binary interpretation the most plausible.

Consequently, the paper recommends retiring TOI 864.01 from the list of planetary candidates and classifying it as a probable false positive (VPC). The study underscores a broader lesson for the exoplanet community: statistical validation tools are only as reliable as the input catalogs, and high‑resolution imaging is essential for identifying sub‑arcsecond companions that can masquerade as planetary transits in TESS data.