Testing formation processes of brain coral terrain on Mars

Role: Collaborator
Team: Eldar Noe Dobrea (PI; PSI), Alphan Altinok (JPL), Kyle Pearson (JPL), Stephen Wood (PSI)
Funding: NASA Mars Data Analysis Program

A primary goal of the Mars Exploration Program is to search for evidence of past or present life. Martian environments with periodic thawing of near-surface ice are considered high-priority targets for this search. We are exploring the hypothesis that Martian brain coral terrain, named for its visual similarity to oceanic brain coral, forms through repeated freeze-thaw cycles and the sorting of surface rocks.

To map the global distribution of brain coral terrain, we used convolutional neural networks on the full MRO/HiRISE dataset, identifying 430 images that contain this landform. The terrain is predominantly found within a specific range of latitudes and elevations, with the largest concentration near 40°N at the dichotomy boundary in the northern hemisphere, mainly forming on slopes of less than a few degrees.

By analyzing the large set of HiRISE images with detected brain terrain, we calculated crater size-frequency distributions. The data revealed fits to the production function at approximately 3 million years (Ma) for craters smaller than 100 meters and around 25 Ma for larger craters. These results suggest partial resurfacing until about 3 Ma, after which smaller craters began to accumulate again according to the production function.

Additionally, we developed an automated algorithm to measure the frequency and size distribution of boulders on Mars using HiRISE image data. The algorithm, trained on about 4,000 hand-labeled samples, uses a random forest model to detect and segment rocks in the images. We integrated this rock detection algorithm with segmentation masks, including those outlining the brain coral terrain. This integration enables comparative studies of rock distributions inside and outside these regions. Preliminary findings indicate that rocks are more concentrated on ridges and their surrounding escarpments.

Pearson et al. (2024) Morgan et al. (2024) Tenth International Conference on Mars Noe Dobrea et al. (2024) Tenth International Conference on Mars