Automatic classification of primary and secondary craters near Copernicus crater with updated insights on its impact trajectory and age

Abstract

Impact cratering is the dominant geological process shaping the Moon's surface. Primary craters form from direct asteroid or comet impacts, while secondary craters are created by debris ejected from these primary impacts. Accurately identifying secondary craters within the primary crater population is essential for understanding planetary processes and surface ages. However, manually distinguishing these secondary craters can be time-consuming and challenging. In this work, a statistical analysis of 5,460 primary and secondary craters reveals significant differences in their spectral characteristics. These differences are postulated to originate from distinct degrees of modification to the target materials and weathering processes. Employing a deep learning model, the research specifically targets the Copernicus crater region to automate the identification of secondary craters. The model classified ∼285,000 secondary and ∼39,000 primary craters with diameters from 200 m to 5 km. Secondary craters make up 89% of the total at 200–280 m, decreasing to around 65% at 4,520–5,000 m. The azimuthal distribution of identified secondary craters suggests an oblique impact from southeast to northwest that formed the Copernicus crater. The model age, based on craters superposed on the ejecta, estimates the Copernicus crater to be ∼755 Ma, overlaying a 3.69 Ga surface. The estimated ages align with previous research. The method is best suited for geologically homogeneous, airless surfaces, and is limited when older primary craters are buried by later ejecta or when ancient craters exhibit similar spectral features due to degradation.