Determining the impact of post-main-sequence stellar evolution on the transiting giant planet population

TL;DR

This study establishes that close-in giant planets orbiting post-main-sequence stars are less common as their hosts evolve, with a pronounced deficit at shorter orbital periods. By leveraging TESS Full-Frame-Images and a homogeneous pipeline, the authors measure a post-main-sequence giant-planet occurrence rate of $0.28\pm0.04\%$ for $1\le P\le12$ d and $8\le R_P\le22\,R_\oplus$, and find a pronounced drop from $0.35\%$ in sub-giants to $0.11^{+0.06}_{-0.05}\%$ in early red giants. The period-dependence and the comparison to main-sequence populations support tidal-decay as the dominant mechanism sculpting this population, in line with theoretical tidal dissipation predictions. These results imply significant tidal stripping of close-in giants during early post-main-sequence evolution and provide a statistically robust platform for future mass measurements and demographic studies of the evolving planetary system around aging stars.

Abstract

The post-main sequence evolution of stars is expected to impact the exoplanets residing on close-in orbits around them. Using photometric data from the TESS Full-Frame-Images we have performed a transit search for exoplanets with post-main sequence hosts to search for the imprints of these impacts on the giant planet population. We detect 130 short period planets and candidates, thirty-three of which are newly discovered candidates, from a sample of 456,941 post-main sequence stars spanning the evolutionary stages from the end of the main sequence to the bottom of the red giant branch. We measure an occurrence rate of $0.28 \pm 0.04$% for short period giant planets orbiting post-main sequence stars. We also measure occurrence rates for two stellar sub-populations, measuring values of $0.35 \pm 0.05$% for a sub-population representing the earliest stages of post-main sequence evolution and $0.11^{+0.06}_{-0.05}$% for a sub-population of more evolved stars. We show that the giant planet occurrence rate decreases with increasing stellar evolution stage, with a larger occurrence rate decrease observed for shorter period planets. Our results are clear evidence that the population of short period giant planets is being sculpted by the post-main sequence evolution of the host stars, and we conclude that this is most likely through the destruction of these giant planets through the increased strength of planet-star tidal interactions resulting in the rapid tidal decay of the planets' orbits.

Figures (17)

• Figure 1: TESS post-main sequence stellar population used as the input sample for this study. Left panel: The stars included in this survey are plotted as the magenta points. The black points show a selection of stars from the TIC across all evolutionary stages as reference. The orange dashed lines show the various stellar tracks and parameter criteria used to select the sample (see Section \ref{['sub:evolstars']}). Right panel: We show the two sub-populations considered in this work. The stars in our sub-giant sample are plotted as the blue points and the stars in our early red giant sample are plotted as the green points. The orange dashed line gives the $\text{EEP} = 465$ boundary used to separate the two samples. See the text in Section \ref{['sub:evolstars']} for details on this boundary. The number of stars in each sub-population are given by the annotations.
• Figure 2: An example BLS candidate detection. Top: TESS QLP flux phase-folded at the orbital period reported by BLS, with the cyan squares showing the data binned in phase. The orange line shows the best-fit BLS box model. Note that we show a zoomed in view around phase 0 for visual clarity of the transit event. The full phase-folded flux data set extends beyond the edges of the plot. Bottom: The BLS periodogram for this candidate. The solid orange vertical line shows the best orbital period reported by BLS, with the dashed lines showing integer multiples and fractions of this period.
• Figure 3: Host stars of the 130 planet candidates detected using our planet search and vetting pipeline. The 117 host stars within our sub-giant sub-population are plotted as the blue squares, and the 13 host stars within our early red giant sub-population are plotted as the green diamonds. Filled in symbols highlight host stars of confirmed planets (according to the NASA exoplanet archive), and open symbols highlight those stars which host planet candidates which are yet to be confirmed. Our full stellar sample is plotted as the black points and the orange line shows the $\text{EEP} = 465$ boundary we use to separate the two sub-populations (see Section \ref{['sub:evolstars']}).
• Figure 4: Our post-main sequence sample of planets and planet candidates. As with Figure \ref{['fig:planet_sample']}, planets and candidates around stars in our sub-giant sub-population are plotted as the blue squares and those around stars in our early red giant sub-population are plotted as the green triangles. Confirmed planets (according to the NASA exoplanet archive) are plotted as the filled symbols and open symbols show as yet unconfirmed planet candidates. The black points show the population of known planets from the NASA exoplanet archive with a planet mass and radius measured to better than 40 % precision.
• Figure 5: Detection efficiency of stages of our planet search and vetting pipeline. The three panels show the detection efficiencies for: left: our full planet search and automated light curve vetting (see Sections \ref{['sub:bls']} and \ref{['sec:FPs']}); middle: our complete transit fitting analysis (see Section \ref{['sec:fitting']}); right: the combined performance of the complete planet search, vetting, and fitting analysis pipeline. The colour of the grid cells gives the detection efficiency (%) for planets within the cell. The numbers in each box also give the percentage detection efficiency within each grid cell.