Direct exoplanet detection and characterization using the ANDROMEDA method: Performance on VLT/NaCo data

TL;DR

This paper presents ANDROMEDA, a maximum likelihood-based algorithm for direct exoplanet detection using ADI, validated on VLT/NaCo data, demonstrating high sensitivity and automatic detection capabilities comparable to existing methods.

Contribution

The paper introduces and experimentally validates ANDROMEDA, a novel maximum likelihood ADI method for exoplanet detection that requires minimal parameter tuning and improves detection and characterization.

Findings

Successfully detected and characterized point sources in real VLT/NaCo data.

Achieved sensitivity comparable to state-of-the-art algorithms.

Demonstrated robustness and automation in processing large survey data.

Abstract

Context. The direct detection of exoplanets with high-contrast imaging requires advanced data processing methods to disentangle potential planetary signals from bright quasi-static speckles. Among them, angular differential imaging (ADI) permits potential planetary signals with a known rotation rate to be separated from instrumental speckles that are either statics or slowly variable. The method presented in this paper, called ANDROMEDA for ANgular Differential OptiMal Exoplanet Detection Algorithm is based on a maximum likelihood approach to ADI and is used to estimate the position and the flux of any point source present in the field of view. Aims. In order to optimize and experimentally validate this previously proposed method, we applied ANDROMEDA to real VLT/NaCo data. In addition to its pure detection capability, we investigated the possibility of defining simple and efficient criteria for automatic point source extraction able to support the processing of large surveys. Methods. To assess the performance of the method, we applied ANDROMEDA on VLT/NaCo data of TYC-8979-1683-1 which is surrounded by numerous bright stars and on which we added synthetic planets of known position and flux in the field. In order to accommodate the real data properties, it was necessary to develop additional pre-processing and post-processing steps to the initially proposed algorithm. We then investigated its skill in the challenging case of a well-known target, $\beta$ Pictoris, whose companion is close to the detection limit and we compared our results to those obtained by another method based on principal component analysis (PCA). Results. Application on VLT/NaCo data demonstrates the ability of ANDROMEDA to automatically detect and characterize point sources present in the image field. We end up with a robust method bringing consistent results with a sensitivity similar to the recently published algorithms, with only two parameters to be fine tuned. Moreover, the companion flux estimates are not biased by the algorithm parameters and do not require a posteriori corrections. Conclusions. ANDROMEDA is an attractive alternative to current standard image processing methods that can be readily applied to on-sky data.