The Giant Accreting Protoplanet Survey (GAPlanetS): Optimization Techniques for Robust Detections of Protoplanets

TL;DR

This paper introduces a systematic grid search optimization method for high-contrast imaging data analysis, improving the detection sensitivity of protoplanets by optimizing PSF subtraction parameters.

Contribution

It presents a new grid search technique for optimizing PSF subtraction parameters in high-contrast imaging, enhancing detection reliability of young exoplanets.

Findings

Optimization improved companion SNR by up to 1.2σ.

Multi-metric approach yields more robust detections.

Method enhances detection sensitivity near the 5σ threshold.

Abstract

High-contrast imaging has afforded astronomers the opportunity to study light directly emitted by adolescent (tens of Myr) and ``proto" ($<$10Myr) planets still undergoing formation. Direct detection of these planets is enabled by empirical Point Spread Function (PSF) modeling and removal algorithms. The computational intensity of such algorithms, and their multiplicity of tunable input parameters, has led to the prevalence of ad-hoc optimization approaches to high-contrast imaging results. In this work, we present a new, systematic approach to optimization vetted using data of the high-contrast stellar companion HD 142527 B from the Magellan Adaptive Optics (MagAO) Giant Accreting Protoplanet Survey (GAPlanetS). More specifically, we present a grid search technique designed to explore three influential parameters of the PSF-subtraction algorithm pyKLIP -- annuli, movement, and KL modes. We consider multiple metrics for post-processed image quality in order to optimally recover at H$\alpha$ (656nm) synthetic planets injected into contemporaneous continuum (643nm) images. These metrics include: peak (single-pixel) SNR, average (multi-pixel average) SNR, 5$\sigma$ contrast, and false-positive fraction. We apply continuum-optimized KLIP reduction parameters to six H$\alpha$ direct detections of the low-mass stellar companion HD142527 B, and recover the companion at a range of separations. Relative to a single-informed, non-optimized set of KLIP parameters applied to all datasets uniformly, our multi-metric grid search optimization led to improvements in companion SNR of up to 1.2$\sigma$, with an average improvement of 0.6$\sigma$. Since many direct imaging detections lie close to the canonical 5$\sigma$ threshold, even such modest improvements may result in higher yields in future imaging surveys.