Applying a temporal systematics model to vector Apodizing Phase Plate coronagraphic data: TRAP4vAPP

¹ Leiden Observatory, Leiden University, PO Box 9531, 2300 RA Leiden, The Netherlands
² SUPA, Institute for Astronomy, University of Edinburgh, Royal Observatory, Blackford Hill, Edinburgh, EH9 3HJ, UK
e-mail: pengyu.liu@ed.ac.uk
³ Centre for Exoplanet Science, University of Edinburgh, Edinburgh, UK
⁴ Anton Pannekoek Institute for Astronomy, University of Amsterdam, Science Park 904, 1098 XH Amsterdam, The Netherlands
⁵ Max-Planck-Institut für Astronomie, Königstuhl 17, 69117 Heidelberg, Germany
⁶ Astrophysics, University of Oxford, Denys Wilkinson Building, Keble Road, Oxford OX1 3RH, UK
⁷ Steward Observatory, University of Arizona, 933 N. Cherry Ave., Tucson, AZ 85721, USA
⁸ Academia Sinica, Institute of Astronomy and Astrophysics, 11F Astronomy-Mathematics Building, NTU/AS campus, No. 1, Section 4, Roosevelt Rd., Taipei 10617, Taiwan

Received: 31 October 2022
Accepted: 24 April 2023

Abstract

Context. The vector Apodizing Phase Plate (vAPP) is a pupil plane coronagraph that suppresses starlight by forming a dark hole in its point spread function (PSF). The unconventional and non-axisymmetrical PSF arising from the phase modification applied by this coronagraph presents a special challenge to post-processing techniques.

Aims. We aim to implement a recently developed post-processing algorithm, temporal reference analysis of planets (TRAP) on vAPP coronagraphic data. The property of TRAP that uses non-local training pixels, combined with the unconventional PSF of vAPP, allows for more flexibility than previous spatial algorithms in selecting reference pixels to model systematic noise.

Methods. Datasets from two types of vAPPs are analysed: a double grating-vAPP (dgvAPP360) that produces a single symmetric PSF and a grating-vAPP (gvAPP180) that produces two D-shaped PSFs. We explore how to choose reference pixels to build temporal systematic noise models in TRAP for them. We then compare the performance of TRAP with previously implemented algorithms that produced the best signal-to-noise ratio (S/N) in companion detections in these datasets.

Results. We find that the systematic noise between the two D-shaped PSFs is not as temporally associated as expected. Conversely, there is still a significant number of systematic noise sources that are shared by the dark hole and the bright side in the same PSF. We should choose reference pixels from the same PSF when reducing the dgvAPP360 dataset or the gvAPP180 dataset with TRAP. In these datasets, TRAP achieves results consistent with previous best detections, with an improved S/N for the gvAPP180 dataset.