Alleviating the transit timing variation bias in transit surveys

Transit timing variations (TTVs) can provide useful information for systems observed by transit, as they allow us to put constraints on the masses and eccentricities of planets, or even constrain existence non-transiting companions. However, TTVs also act a detection bias that prevent small planets in transit surveys would otherwise be detected standard algorithms such Boxed Least Square algorithm if their orbit was not perturbed. This is especially present with long baseline, Kepler , some TESS sectors, upcoming PLATO mission. Here we introduce method robust large TTVs, illustrate its use recovering confirming pair resonant super-Earths ten-hour around Kepler-1705 (prev. KOI-4772). The based neural network trained recover tracks low-signal-to-noise-ratio (S/N) perturbed river diagrams. We parameters these candidates fitting light curve. individual S/N Kepler-1705b c are about three times lower than all previously known 3 h more, pushing boundaries recovery small, dynamically active planets. Recovering this type object essential obtaining complete picture planetary systems, solving often taken into account statistical studies exoplanet populations. In addition, means mass estimates which studying internal structure discovered surveys. Finally, show due strong orbital perturbations, it possible spin outer planet trapped sub- super-synchronous spin–orbit resonance. have important consequences climate because non-synchronous implies flux star spread over whole surface.