ALDERAAN: A Python pipeline that hunts for tiny planetary timing variations in Kepler data, using Gaussian processes and nested sampling to detect the gravitational whispers of hidden worlds.

Transit timing variations (TTVs) are one of astronomy’s most elegant detective stories. When planets orbit their stars, they occasionally tug on each other gravitationally, causing subtle shifts in their transit schedules—sometimes just minutes early or late. These tiny deviations can reveal hidden planets, measure masses, and map entire planetary systems. But extracting these signals from noisy spacecraft data requires sophisticated statistical archaeology.

ALDEROAN tackles this challenge head-on with a three-stage pipeline optimized for TTV detection in Kepler, K2, and TESS photometry. It combines narrow bandstop filters with Gaussian Process regression to handle autocorrelated noise from both stellar activity and instrumental effects, then applies either Dynamic Nested Sampling or Hamiltonian Monte Carlo with umbrella sampling for robust parameter estimation. Built on a foundation of proven astronomy libraries—batman for transit modeling, celerite for fast GP computations, and exoplanet for modern probabilistic inference—it transforms raw FITS files from MAST into publication-ready transit fits.

While currently focused on Kepler data, ALDERAAN represents the kind of specialized tooling that pushes exoplanet science forward. For researchers studying multi-planet systems or hunting for non-transiting companions, this pipeline offers the precision needed to extract meaningful science from increasingly subtle signals in the era of precision photometry.

⭐ Stars: 7
💻 Language: Python
🔗 Repository: gjgilbert/alderaan