Automated Stratigraphy and Chronology Map of Lunar Mare Basalts

For crater classification, five features were selected from both spatial clustering and morphological aspects, including eccentricity, depth-to-diameter ratio, Voronoi area, chain coefficient, and cluster coefficient, to construct a Random Forest classification model. This enabled automated classification of primary and secondary craters with diameters larger than 600 m in the mare region, attaining an overall accuracy of 92.81%.

CSFD dating of lunar mare basalts was performed using the basalt units derived from automatic delineation and the filtered primary craters. The ages obtained in this study yield a mean absolute difference of 0.25 Ga compared with existing researches, and are generally consistent with the radioisotopic ages of samples returned by the Apollo, Luna, and Chang’e missions.</p><p dir="ltr">Citation:</p><p dir="ltr">Zhang, H., Zhang, J., Zhang, X., Jin, L., Zhou, Z., Wang, S., Wang, Y., Chen, J., Qiao, L., Ling, Z., Yue, Z., (2026), Automated Stratigraphic Mapping and Chronology of Lunar Mare Basalts.

IEEE journal, submitted.</p>