PlanetPack3: a radial-velocity and transit analysis tool for exoplanets

TL;DR

PlanetPack3 is a comprehensive software tool for exoplanet detection and analysis, featuring advanced noise modeling, improved transit timing variation analysis, joint data fitting, and performance enhancements, supporting diverse observational data types.

Contribution

This release introduces Gaussian process noise modeling, enhanced TTV analysis, joint photometric and radial velocity fitting, Rossiter-McLaughlin effect modeling, and speed optimizations, advancing exoplanet data analysis capabilities.

Findings

Gaussian process noise modeling improves data fit accuracy.

Enhanced TTV analysis with limb-darkening and red-noise detection.

Joint fitting of photometric and radial velocity data increases parameter estimation precision.

Abstract

PlanetPack, initially released in 2013, is a command-line software aimed to facilitate exoplanets detection, characterization, and basic dynamical $N$-body simulations. This paper presents the third major release of PlanetPack that incorporates multiple improvements in comparison to the legacy versions. The major ones include: (i) modelling noise by Gaussian processes that in addition to the classic white noise may optionally include multiple components of the red noise, modulated noise, quasiperiodic noise (to be added soon in minor subversions of the 3.x series); (ii) an improved pipeline for TTV analysis of photometric data that includes quadratic limb-darkening model and automatic red-noise detection; (iii) self-consistent joint fitting of photometric + radial velocity data with full access to all the functionality inherited from the legacy PlanetPack; (iv) modelling of the Rossiter-McLaughlin effect for arbitrary eclipser/star radii ratio, and optionally including corrections that take into account average characteristics of a multiline stellar spectrum; (v) speed improvements through multithreading and CPU-optimized BLAS libraries. PlanetPack was written in pure C++ (standard of 2011), and is expected to be run on a wide range of platforms.