Detection of a peculiar noise type in the TESS "fast" light curves

TL;DR

This paper identifies a high-frequency blue noise in TESS light curves likely caused by satellite jitter, and discusses methods for mitigating this noise through decorrelation techniques, highlighting the challenges of simple correction methods.

Contribution

The paper detects a new high-frequency noise component in TESS data and evaluates decorrelation techniques for noise mitigation, emphasizing the limitations of simple polynomial corrections.

Findings

Blue noise correlates with satellite pointing instability.

Decorrelating against subpixel position reduces noise.

Simple polynomial corrections are often insufficient.

Abstract

We present the detection of a peculiar high-frequency noise component in the 20 second cadence SAP (Simple Aperture Photometry) light curve of TESS (Transiting Exoplanets Survey Satellite). This effect (labeled as blue noise) may be attributed to the pointing instability (also known as satellite jiiter) of the satellite. We present a common technique used in the mitigation of the jitter, by decorrelating against the subpixel position of the photo-center of the point spread function of the star. We also show that a simple linear or polynomial technique may not yield satisfactory corrections, as the behavior or attitude of the noise properties may change considerably throughout the light curve.