A false positive transit candidate for EPIC 211101996 from K2 and TESS data identified as background eclipsing binary Gaia DR3 66767847894609792

Transiting planets around young stars are hard to find due to the enhanced stellar activity. Only a few transiting planets have been detected around stars younger than 100 Myr. We initially detected a transit-like signal in the K2 light curve of a very cool M dwarf star (EPIC 211101996) in the Pleiades open cluster, with an estimated age of about 100 Myr. Our detailed analysis of the per-pixel light curves, detrending with the W\={o}tan software and transit search with the Transit Least Squares algorithm showed that the source of the signal is a contaminant source (Gaia DR3 66767847894609792) 20"west of the target. The V-like shape of its phase-folded light curve and eclipse depth of ~15% suggest that it is a grazing eclipsing binary. The contaminant has hitherto been listed as a single star, which we now identify as an eclipsing stellar binary with a period of about 6 days.


INTRODUCTION
In their search for transiting planets in 702 light curves of ultra-cool dwarf stars (and possibly brown dwarfs) in K2 data, Sestovic & Demory (2020a) did not find any new planets in addition to the previously known TRAPPIST-1 system.In their search, they used the Box Least Squares (BLS) algorithm (Kovács et al. 2002), which has been shown to be less sensitive to shallow transits than the Transit Least Squares (TLS) algorithm (Hippke & Heller 2019;Heller et al. 2019).We thus repeated the search for transiting objects around those 702 ultra-cool targets, hoping to find small planets that had previously been missed.

METHODS
We took the K2 light curves that had been generated by Sestovic & Demory (2020a) from K2 target pixel files using their custom pipeline for flux extraction, systematic noise removal, and Gaussian progress regression (Sestovic & Demory 2020b).Then we applied TLS to search for transits, the total computing time of which amounted to just a few hours on a standard laptop.TLS detected a new transit-like signal in the K2 light curve of EPIC 211101996 (TIC 35155735) with a period of 5.98796 (±0.01094) d and a depth of about 3.6 %.The signal detection efficiency (SDE) of 47.9 exceeded the commonly used threshold value for a detection of 7 by a lot.
Using spectral fitting of the Gaia and 2MASS photometry, Sestovic & Demory (2020a) estimate a mass of 0.1 M ⊙ , a radius of 0.127 R ⊙ , an effective temperature of 2850 K, and a spectral type of M6V for EPIC 211101996.The planetto-star radius ratio that we initially measured implied a planet radius of about 2.6 R ⊕ for our new transit candidate.
The sky position of EPIC 211101996 and its parallax of 7.36 mas from Gaia Data Release 3 (DR3) (Gaia Collaboration et al. 2023), equivalent to a distance of about 136 pc, strongly suggest that it is a member of the Pleiades open star cluster, which has an estimated age of just about 100 Myr (Basri et al. 1996;Ushomirsky et al. 1998).
For our vetting of the transiting planet candidate, we used the nominal K2 light curve generated by the EVEREST v2.0 pipeline (Luger et al. 2018) and applied various detrending methods provided by the Wōtan software package (Hippke et al. 2019).None of our detrending methods resulted in a detection of the signal that we found in the light curve provided by Sestovic & Demory (2020a).We also tested other corrections of the K2 light curve from systematic effects such as k2sc, k2sff, k2varcat, polar and ktwo, which did not result in a confirmation of the signal.
From August to November 2021, new TESS observations of EPIC 211101996 were obtained in Sectors 42, 43, and 44.The TESS data validation report summary showed a clear transit signal with a period of 11.97649 (± 0.00578) d, which is 2.000095 times the period that we found in the K2 data.More suspiciously, however, the transit depth was now a whopping 25 %, suggesting a radius for the transiting object of 50 % of the primary radius, and thus a possible stellar binary rather than a transiting planet.

RESULTS
We tested multiple aperture masks around EPIC 211101996 using EVEREST to extract the K2 light curves, then detrended them with Tukey's biweight filter implemented in Wōtan and searched for the origin of the transit-like signal with TLS.We identified the contaminant Gaia DR3 66767847894609792, which is about 20 ′′ west of the target, as the source of the signal (Fig. 1).TLS found an SDE value of ∼ 20 for the light curve that we extracted (see Fig. 1c), which is substantially lower than the signal that we originally detected in the light curve provided by Sestovic & Demory (2020a).
Gaia DR3 66767847894609792 has hitherto been classified as a single stellar object in Gaia DR3 with a parallax of 0.3967 (± 0.0428) mas.The resulting distance of 2520 +305 −245 pc and its proper motion µ α = 2.454 (± 0.045) mas/yr, µ δ = −2.896(± 0.040) mas/yr suggest that the astrophysical source of the transit-like signal is in a chance alignment far behind the Pleiades cluster.
The V-like shape and ∼15 % depth of the phase-folded light curve (see Fig. 1b) plus visual indication of a shallow secondary eclipse indicate an eclipsing stellar binary as the source of the signal.Moreover, the Markov chain Monte Carlo multiple-star classifier from Gaia DR3 (not shown) exhibits a bimodal posterior distribution with primary stellar parameters T eff,1 ∼ 5550 K, log(g 1 ) ∼ 4.4, secondary stellar parameters T eff,2 ∼ 5100 K, log(g 2 ) ∼ 4.75, a combined metallicity of [M/H] ∼ − 0.7 dex, and a distance of about 1730 pc.This revised distance estimate is based on lowresolution BP/RP spectra and the parallax estimate from Gaia (Gaia Collaboration et al. 2023), still suggesting a position far behind the Pleiades from our perspective.
The TESS Science Processing Operations Center (SPOC) derived an alias of the true orbital period due to gaps in the observations that masked several transits.SPOC also falsely attributed the signal to EPIC 211101996, partly because the TESS pixels are much larger than the K2 pixels.A detailed comparison of the TESS and K2 aperture masks and the resulting depths of the eclipse signals was beyond the scope of our study.
The transit-like periodic signal in the K2 and TESS light curves of EPIC 211101996 (TIC 35155735) is a false positive caused by a background contaminant, which we identify as an eclipsing stellar binary.The source of the signal is Gaia DR3 66767847894609792, which should be considered as a stellar binary rather than a single source.
RH acknowledges support from the German Aerospace Agency (Deutsches Zentrum für Luft-und Raumfahrt) under PLATO Data Center grant 50OO1501.JVH is funded by the DFG priority programme SPP 1992 "Exploring the Diversity of Extrasolar Planets (SM 486/2-1)".The authors thank Matthias Ammler-von Eiff for helpful comments.
Facilities: Kepler (K2), TESS, Gaia Software: TLS (Hippke & Heller 2019), Wōtan (Hippke et al. 2019), lightkurve (Lightkurve Collaboration et al. 2018) in the same figure and the light curve of the eastern contaminant in Figure4.6.In the light curve of the western contaminant the target signal of 5.99 d is visible, while it cannot be found by Transit Least Squares for both the target and the eastern contaminant.However, both at the start and at the end of the data range the flux of the western contaminant shows high levels of noise, while within the range of 2240-2275 d the target signal is visible to the human eye.The signal of 5.99 d can be found by Transit Least Squares only when the flux is constrained to the range of 2236-2288 d, missing three transits.

Figure 4
Figure 4.3: Target pixel file of EPIC211101996 from K2 data.Inside each pixel is shown the flux within the interval 2247-2269 d generated by the threeby-one aperture centered around that pixel.Within the interval, four transits should be visible, which only applies to the western contaminant, Gaia DR3 66767847894609792.In red is marked the K2 aperture as reference.

Figure 4 . 3 :
Figure 4.3: Target pixel file of EPIC211101996 from K2 data.Inside each pixel is shown the flux within the interval 2247-2269 d generated by the threeby-one aperture centered around that pixel.Within the interval, four transits should be visible, which only applies to the western contaminant, Gaia DR3 66767847894609792.In red is marked the K2 aperture as reference.

Figure 4 . 7 :
Figure 4.7: Output from Transit Least Squares.Top: the relative flux folded over the phase with the transit fit curve in red.Middle: the relative flux plotted over time with the transit fit curve in red.Bottom: the signal detection e ciency (SDE) plotted over the period with integer multiples of the period with maximum SDE (blue vertical lines).Left column: results for the western contaminant.Right column row: results for the target EPIC211101996.

Figure 4 . 7 :Figure 1 .
Figure 4.7: Output from Transit Least Squares.Top: the relative flux folded over the phase with the transit fit curve in red.Middle: the relative flux plotted over time with the transit fit curve in red.Bottom: the signal detection e ciency (SDE) plotted over the period with integer multiples of the period with maximum SDE (blue vertical lines).Left column: results for the western contaminant.Right column row: results for the target EPIC211101996.