2025ApJ...984L..30T

In an era when we are charting multiple planets per system, one might wonder the extent to which “missing” (or failing to detect) a planet can skew our interpretation of the system architecture. We address this question with a simple experiment: starting from a large, homogeneous catalog, we remove planets and monitor how several well-defined metrics of the system architecture change. We first perform this test on a catalog of observed exoplanets. We then repeat our test on a catalog of synthetic planetary systems with underlying hyperparameters that have been fit to reproduce the observed systems as faithfully as possible (though imperfectly). For both samples, we find that the failure to detect one or more planets tends to create more irregularly spaced planets, whereas the planet mass similarity and coplanarity are essentially unaffected. One key difference between the synthetic and observed data sets is that the observed systems have more evenly spaced planets than the observation-bias-applied synthetic systems. Since our tests show that detection bias tends to increase irregularity in spacing, the even spacing in the observed planetary systems is likely astrophysical rather than the result of the Kepler missions' inherent detection biases. Our findings support the interpretation that planets in the same system have similar sizes and regular spacing and reinforce the need to develop an underlying model of planetary architectures that reproduces these observed patterns.