Rapid Automated Mapping of Clouds on Titan With Instance Segmentation

• Published in: Journal of geophysical research. Machine learning and computation Vol. 2; no. 1
• Main Authors: Yahn, Zachary, Trent, Douglas M., Duncan, Ethan, Seignovert, Benoît, Santerre, John, Nixon, Conor A.
• Format: Journal Article
• Language: English
• Published: American Geophysical Union/Wiley 01.03.2025; Wiley
• Subjects: Astrophysics; clouds; deep learning; Earth and Planetary Astrophysics; instance segmentation; Sciences of the Universe; Titan
• ISSN: 2993-5210; 2993-5210
• DOI: 10.1029/2024JH000366

Despite widespread adoption of deep learning models to address a variety of computer vision tasks, planetary science has yet to see extensive utilization of such tools to address its unique problems. On Titan, the largest moon of Saturn, tracking seasonal trends and weather patterns of clouds provides crucial insights into one of the most complex climates in the Solar System, yet much of the available image data are still analyzed in a conventional way. In this work, we apply a Mask R‐CNN trained via transfer learning to perform instance segmentation of clouds in Titan images acquired by the Cassini spacecraft—a previously unexplored approach to a “big data” problem in planetary science. We demonstrate that an automated technique can provide quantitative measures for clouds, such as areas and centroids, that may otherwise be prohibitively time‐intensive to produce by human mapping. Furthermore, despite Titan‐specific challenges, our approach yields accuracy comparable to contemporary cloud identification studies on Earth and other worlds. We compare the efficiencies of human‐driven versus algorithmic approaches, showing that transfer learning provides speed‐ups that may open new horizons for data investigation for Titan. Moreover, we suggest that such approaches have broad potential for application to similar problems in planetary science where they are currently under‐utilized. Future planned missions to the planets and remote sensing initiatives for the Earth promise to provide a deluge of image data in the coming years that will benefit strongly from leveraging machine learning approaches to perform the analysis. Plain Language Summary Although deep learning models continue to be popular tools for addressing a variety of computer vision tasks, planetary science has yet to see their extensive utilization. One such application is cloud tracking in Titan's atmosphere, which has dynamic and complex weather phenomena. In this work, we train a deep computer vision model to identify these clouds from images of Titan's surface. We assess the accuracy of this model, and we also show that our approach is more efficient than human labelers. In addition to cloud detection, we also develop codes for automatically extracting data such as cloud areas and latitudes from our model's predictions. We also propose that similar methods have broad applicability to other planetary science problems, especially given several upcoming missions that promise an influx of high‐quality image data. Key Points We curate a subset of Cassini ISS Titan images and apply a Mask R‐CNN model, achieving high scores on cloud instance segmentation We develop codes for processing the predicted cloud instances to extract a final data set of cloud areas, centroids, and aspect ratios We demonstrate a transfer learning approach that indicates the feasibility of a study of the full Cassini ISS data and even other missions