SPIDERMAN: an open-source code to model phase curves and secondary eclipses

TL;DR

SPIDERMAN is a fast, open-source code that models exoplanet phase curves and secondary eclipses with arbitrary surface brightness distributions, enabling efficient analysis and comparison of different models.

Contribution

The paper introduces SPIDERMAN, a highly efficient and flexible code for modeling exoplanet phase curves and eclipses with arbitrary brightness maps, optimized for rapid MCMC analysis.

Findings

Successfully applied to WASP-43b data, fitting the phase curve with a physical brightness map.

The model suggests the planet's nightside may have clouds or reflected light, reducing the need for a hot nightside.

No significant wavelength dependence found in the brightness map parameters.

Abstract

We present SPIDERMAN, a fast code for calculating exoplanet phase curves and secondary eclipses with arbitrary surface brightness distributions in two dimensions. Using a geometrical algorithm, the code solves exactly the area of sections of the disc of the planet that are occulted by the star. The code is written in C with a user-friendly Python interface, and is optimised to run quickly, with no loss in numerical precision. Approximately 1000 models can be generated per second in typical use, making Markov Chain Monte Carlo analyses practicable. The modular nature of the code allows easy comparison of the effect of multiple different brightness distributions for the dataset. As a test case we apply the code to archival data on the phase curve of WASP-43b using a physically motivated analytical model for the two dimensional brightness map. The model provides a good fit to the data; however, it overpredicts the temperature of the nightside. We speculate that this could be due to the presence of clouds on the nightside of the planet, or additional reflected light from the dayside. When testing a simple cloud model we find that the best fitting model has a geometric albedo of $0.32 \pm0.02$ and does not require a hot nightside. We also test for variation of the map parameters as a function of wavelength and find no statistically significant correlations.