TRAP: a temporal systematics model for improved direct detection of exoplanets at small angular separations

Abstract

Context. High-contrast imaging surveys for exoplanet detection have shown that giant planets at large separations are rare. Thus, it is of paramount importance to push towards detections at smaller separations, which is the part of the parameter space containing the greatest number of planets. The performance of traditional methods for the post-processing of pupil-stabilized observations decreases at smaller separations due to the larger field-rotation required to displace a source on the detector in addition to the intrinsic difficulty of higher stellar contamination. Aims. Our goal is to develop a method of extracting exoplanet signals, which improves performance at small angular separations. Methods. A data-driven model of the temporal behavior of the systematics for each pixel can be created using reference pixels at a different positions, on the condition that the underlying causes of the systematics are shared across multiple pixels, which is mostly true for the speckle pattern in high-contrast imaging. In our causal regression model, we simultaneously fit the model of a planet signal “transiting” over detector pixels and non-local reference light curves describing the shared temporal trends of the speckle pattern to find the best-fitting temporal model describing the signal. Results. With our implementation of a spatially non-local, temporal systematics model, called TRAP, we show that it is possible to gain up to a factor of six in contrast at close separations (<3λ∕D), as compared to a model based on spatial correlations between images displaced in time. We show that the temporal sampling has a large impact on the achievable contrast, with better temporal sampling resulting in significantly better contrasts. At short integration times, (4 s) for β Pic data, we increase the signal-to-noise ratio of the planet by a factor of four compared to the spatial systematics model. Finally, we show that the temporal model can be used on unaligned data that has only been dark- and flat-corrected, without the need for further pre-processing.

Auto-RSM: An automated parameter-selection algorithm for the RSM map exoplanet detection algorithm

“…The most popular PSFsubtraction methods include ADI median-subtraction, locally optimised combination of images (LOCI, Lafreniere et al 2007), principal component analysis (PCA/KLIP, Soummer et al 2012;Amara & Quanz 2012), non-negative matrix factorisation (NMF, Ren et al 2018), and the local low rank plus sparse plus Gaussian decomposition (LLSG, Gonzalez et al 2016). Other algorithms such as ANDROMEDA (Cantalloube et al 2015), KLIP FMMF (Pueyo 2016;Ruffio et al 2017), PACO (Flasseur et al 2018), and TRAP (Samland et al 2021) exploit the inverse problem approach, which, for HCI, consists in tracking a model of the expected planetary signal in the set of frames included in the ADI sequence. All these methods rely on signal-to-noise ratio (S/N) maps to detect planetary candidates.…”

BEAST begins: sample characteristics and survey performance of the B-star Exoplanet Abundance Study

“…Thus the three approaches are as follows: (1) a classical ADI approach; (2) a TLOCI approach; and (3) a Karhunen-Loève Image Projection (KLIP, see Soummer et al 2012;Pueyo 2016) Principal Component Analysis (PCA)-based approach as an alternative to the TLOCI algorithm. Beyond this standardized procedure, several efforts are in progress to analyse the BEAST data with alternative reduction schemes, including Pyn-Point (Stolker et al 2019), ANDROMEDA (Cantalloube et al 2015), and TRAP (Samland et al 2021), but this paper focusses on the standardized procedure.…”

BEAST begins: Sample characteristics and survey performance of the B-star Exoplanet Abundance Study

“…Thus the three approaches are as follows: (1) a classical ADI approach; (2) a TLOCI approach; and (3) a Karhunen-Loève Image Projection (KLIP, see Soummer et al 2012;Pueyo 2016) Principal Component Analysis (PCA)-based approach as an alternative to the TLOCI algorithm. Beyond this standardized procedure, several efforts are in progress to analyse the BEAST data with alternative reduction schemes, including Pyn-Point (Stolker et al 2019), ANDROMEDA (Cantalloube et al 2015), and TRAP (Samland et al 2020), but this paper focusses on the standardized procedure.…”

“…It would most likely be possible to enhance the contrast further still on a star-by-star basis through alternative high-contrast algorithms (e.g. Samland et al 2020) or through adjusting the parameters in the SpeCal algorithm. As mentioned in Sect.…”