Moon and background removal algorithm for all-sky imager

TL;DR

This paper presents a novel algorithm for removing moon contamination from all-sky images, enabling clearer observation of auroral activities by effectively handling saturated and glow pixels caused by the moon.

Contribution

The authors developed a generalized moon removal algorithm for all-sky imagers that preserves auroral features while eliminating moon-related artifacts, improving data quality for auroral studies.

Findings

Effective removal of moon surface saturation and glow pixels.

Preservation of authentic auroral structures in cleaned images.

Algorithm does not address cloud contamination.

Abstract

All-sky imagers (ASIs) are used to record auroral activities from the ground but are often contaminated by the moon. Here, we studied the THEMIS ASIs data and developed an algorithm to eliminate the moon which can be generalized to other types of ASIs. With our algorithm, the ASI pixels within the moon's surface are typically saturated and thus removed by the algorithm. The ASI pixels within the moon's glow are close to but not saturated and thus can be calibrated by the algorithm to recover auroral structures within the glow. For pixels far away from the moon or when there is no moon, the algorithm preserves typical auroral forms, from the transient features of auroral streamers and pulsating aurora to more stable features of pre-onset arcs. Note that the algorithm does not treat clouds, which is a known limitation.