The density of craters on a planetary surface directly relates to the age of the surface. As the surface ages, however, craters can be erased by subsequent large impacts via direct overprinting, known as geometric crater obliteration. Such counts become increasingly limited as surfaces become more heavily cratered. Techniques to infer the statistics of the regions obliterated by craters were developed in the past decade. Such techniques, however, have only been used for regional studies. Herein, we present a study of the global density of lunar impact craters ≥20 km in diameter using both traditional crater-counting and buffered nonsparseness correction (BNSC) crater-counting techniques. By comparing the measurements, we quantify the influence of geometric crater obliteration on the visible lunar crater record. Our results reveal that geometric crater obliteration erased up to three-fifths of craters ≥20 km in diameter that formed on the most ancient lunar terrains, whereas younger surfaces, like the Procellarum KREEP Terrane, show little to no evidence of such crater obliteration. The differences in derived crater densities highlight ancient surfaces in which the effects of geometric crater obliteration must be considered to characterize their cratering histories. Furthermore, our results identify the most heavily cratered area on the Moon, a region of the lunar highlands between Smythii basin and the South Pole–Aitken (SPA) basin (Smythii–SPA–Highlands); the number of impacts revealed by the BNSC technique for this region is consistent with estimates derived from the abundance of highly siderophile elements and from modeling crustal porosity.
