Abstract Numerous space agencies are planning sustainable human presence at the lunar south pole. Understanding the nature of the lunar crust, particularly its three-dimensional density structure, is important, for example, for the search for water ice predicted to be present. The paucity of in situ measurements requires that we exploit the knowledge obtained through orbital measurements. Two data sets, gravity and topography, in combination with local geology, can provide constraints on the density of the crust. We analyzed the radial component of the gravity field to determine the density of the crust, which we relate to depth using a point-mass spherical harmonic approach. Our investigation provides estimates of density and density gradient at 20 sites in the south polar region: the original 13 Artemis sites, 5 permanently shadowed regions (PSRs), the south pole, and a site on the floor of the Amundsen crater. We find that the mean density in the top 2–10 km for the six PSRs (2475 ± 108 kg m −3 ) is comparable to that (2483 ± 147 kg m −3 ) for the 14 non-PSRs. The weighted mean gradient for the five PSRs and the Amundsen floor is 3 times greater than that for the 14 non-PSRs in the 2–4 km depth range; the average density increases with depth at 25 kg m −3 km −1 . To investigate whether a relationship exists between inferred areal coverage of ice and our density results, we compared densities with results for reflectance in several common PSR localities. We find lower densities and higher gradients for areas characterized by greater areal coverage of ice.
Goossens et al. (Fri,) studied this question.