Pixel Level Decorrelation in Service of the \textit{Spitzer} Microlens Parallax Survey

TL;DR

This paper adapts Pixel Level Decorrelation (PLD) for Spitzer microlensing data, significantly improving photometric precision and potentially benefiting future space telescopes like the Nancy Grace Roman Space Telescope.

Contribution

The authors extend and adapt PLD for crowded-field microlensing observations, demonstrating enhanced photometric accuracy over existing pipelines.

Findings

Photometry precision improved by 1.5 to 6 times.

Method successfully applied to two microlensing events.

Potential applicability to future space telescopes.

Abstract

Microlens parallax measurements combining space-based and ground-based observatories can be used to study planetary demographics. In recent years, the Spitzer Space Telescope was used as a microlens parallax satellite. Meanwhile, \textit{Spitzer} IRAC has been employed to study short-period exoplanets and their atmospheres. As these investigations require exquisite photometry, they motivated the development of numerous self-calibration techniques now widely used in the exoplanet atmosphere community. Specifically, Pixel Level Decorrelation (PLD) was developed for starring-mode observations in uncrowded fields. We adapt and extend PLD to make it suitable for observations obtained as part of the \textit{Spitzer} Microlens Parallax Campaign. We apply our method to two previously published microlensing events, OGLE-2017-BLG-1140 and OGLE-2015-BLG-0448, and compare its performance to the state-of-the-art pipeline used to analyses \textit{Spitzer} microlensing observation. We find that our method yields photometry 1.5--6 times as precise as previously published. In addition to being useful for \textit{Spitzer}, a similar approach could improve microlensing photometry with the Nancy Grace Roman Space Telescope.