Mapping "Brain Coral" Regions on Mars Using Deep Learning

One of the main objectives of the Mars Exploration Program is to search for evidence of past or current life on the planet. To achieve this, Mars exploration has been focusing on regions on Mars that may have liquid water. A set of critical areas may have seen cycles of ice thawing in the relatively recent past in response to periodic changes in the obliquity of Mars. In our study, we employ the use of convolutional neural networks to identify specific areas of the Martian surface that have a terrain feature referred to as "Brain Coral". This terrain type is known for its resemblance in morphology to "sorted stone circles" found on Earth, which may indicate that it formed due to cycles of freezing and thawing. Brain coral is a term used to describe the texture of a specific type of Martian terrain, which is characterized by stippled, labyrinthine ridges and are found primarily in mid-latitudes. They are often associated with lineated valley fill (LVF) and concentric crater fill (CCF) (e.g., [1] ; [2] ; [3] ). In arctic regions on Earth, "sorted stone circles" form as a result of rock-bearing and water-rich soil layers undergoing cycles of expansion and contraction in response to periodic freezing and thawing of the soil ( [4] ; [5] ; [6] ). These cycles result in rock and soil separation which in turn leads to pattern formation (e.g. [7] ; [8] ; [9] ). However, even though brain coral on Mars and sorted stone circles on Earth have similar appearances, it does not necessarily mean that the same processes formed them. There are multiple young geomorphic features on Mars that suggest thawing occurred in the recent past and may still be ongoing, but there is ongoing controversy about it ( [10] , [11] , [12] , [13] , [14] ). As such, it is crucial to compare existing hypotheses and conduct a thorough examination of this terrain to better understand its formation.