Mercury's Tectonic and Geodynamic History: 1. Contractional Tectonic Landform Analysis and Tectonic Strain Using Machine Learning

Abstract Mercury's tectonic record is dominated by shortening landforms, including lobate scarps, high‐relief ridges and wrinkle ridges. Previous analyses of these structures have used displacement–length ratios to constrain the planet's global contraction to a range of either no more than 2 km or up to 7 km. This important discrepancy has strong implications for our understanding of Mercury's geodynamic history and results from different interpretations of the tectonic record. Studies in favor of less contraction do not include wrinkle ridges and other small‐scale landforms as part of the record of Mercury's contraction, while prior works arguing for large contraction are subject to greater uncertainty regarding the strain associated with individual landforms. Here we use existing mapping of tectonic landforms together with machine learning to evaluate ridge height for a more robust strain analysis. Tectonic strain is converted to global contraction considering fault orientation. Removing small secondary landforms that are parallel and in close vicinity to a primary longer landform leads to a global contraction of about 6.3 km, which is substantially higher than when neglecting wrinkle ridges (∼1.2 km). Considerable global contraction is also potentially accommodated by non‐tectonic means, indicating that the tectonic record underestimates the magnitude of contraction Mercury experienced. Tectonic strain exhibits prominent lateral variation, with some regions experiencing near‐zero strain, while others recorded substantial deformation. Our analyses reveal a period of rapid contraction from 4.1 to 3.9 Ga at 0.02–0.04 km/Myr, followed by much lower rates of contraction, which could mark the onset of Mercury's inner core nucleation.