Peering into lunar permanently shadowed regions with deep learning

• Published in: Nature communications 2021-09, Vol.12 (1), p.5607-5607, Article 5607
• Main Authors: Bickel, V. T., Moseley, B., Lopez-Francos, I., Shirley, M.
• Format: Article
• Language: English
• Subjects: 639/33/445/215; 639/705/117; 639/705/794; 704/445/125; 704/445/215; Algorithms; Deep learning; Geomorphology; Humanities and Social Sciences; Ice; Lunar craters; Machine learning; Mission planning; Moon; multidisciplinary; Post-production processing; Science; Science (multidisciplinary)
• ISSN: 2041-1723; 2041-1723
• DOI: 10.1038/s41467-021-25882-z

The lunar permanently shadowed regions (PSRs) are expected to host large quantities of water-ice, which are key for sustainable exploration of the Moon and beyond. In the near future, NASA and other entities plan to send rovers and humans to characterize water-ice within PSRs. However, there exists only limited information about the small-scale geomorphology and distribution of ice within PSRs because the orbital imagery captured to date lacks sufficient resolution and/or signal. In this paper, we develop and validate a new method of post-processing LRO NAC images of PSRs. We show that our method is able to reveal previously unseen geomorphological features such as boulders and craters down to 3 meters in size, whilst not finding evidence for surface frost or near-surface ice. Our post-processed images significantly facilitate the exploration of PSRs by reducing the uncertainty of target selection and traverse/mission planning. Some regions on the Moon are permanently covered in shadow and are therefore extremely difficult to see into. We develop a deep learning driven algorithm which enhances images of these regions, allowing us to see inside them with high resolution for the first time.