The volcanic filling of the lunar maria, the formation of mascons, the tectonic features of the mare regions, and the thermal evolution of the lunar lithosphere are all interlinked. To explore these processes for the major mascon maria (Serenitatis, Humorum, Imbrium, Orientale, Crisium, Nectaris, Smythii, and Grimaldi), we determine the characteristics and geometry of tectonic features in each mascon mare region, the distribution of major mare units and the flooding and subsidence history for each basin, the temporal relations between tectonic features and geologic units, the effective lithospheric thickness as a function of time in each mascon region, the implications of the derived thicknesses for lunar thermal structure, the relationship between local and global sources of stress in controlling lunar tectonic history, and the consequent limitations this relationship places on lunar thermal evolution.

All mascon basins display some associated tectonic features: linear rilles, which are graben-like features resulting from horizontal crustal tension, and/or mare ridges, which result from horizontal compression and buckling of near-surface material. Crisium, Smythii, and Nectaris have ridges but no associated rilles. Mare ridges occur almost exclusively in mare basalt deposits. Where developed, rilles tend to be concentric to the basin and to occur outside the ridge systems, usually in the adjacent highlands. Volumes, sequences, and timing of basalt emplacement are determined from basin geometry and by stratigraphic reconstruction from remote sensing data and age determinations. Amounts of basin subsidence are determined from patterns of geologic units and present mare surface topography. Total volumes of basalt vary as a function of basin size, amount of subsidence, and stages of flooding (for example, Orientale, little flooding, and Imbrium, extensive flooding). The vast majority of filling for Serenitatis, Crisium, Nectaris, Imbrium, Humorum, and, possibly, Orientale apparently occurred early (3.8–3.6 b.y.). Lesser amounts were added to Serenitatis, Crisium, Imbrium, and Humorum at 3.6–3.2 b.y. Minor amounts were subsequently added to Serenitatis, Imbrium, and Crisium. The earliest basalts subsided shortly after their emplacement, indicating that the mare basalt load was superisostatic at that time. Linear rille formation appears to be restricted to the period prior to about 3.6 ± 0.2 b.y., although subsidence continued well past that time. Mare ridges occur throughout mare regions. Both subsidence and mare ridge formation must have continued until after emplacement of even the youngest mare basalt units.

Detailed models of the flexural response of the lunar lithosphere to the mare basalt loads at the times of emplacement of major mare units permit evaluation of the effective thickness T of the elastic lithosphere. The distribution of well-developed graben concentric to each mare basin center is matched by a spatially variable thickness of the elastic lithosphere during the time of rille formation 3.6–3.8 b.y. ago: T ≲ 25 km for Grimaldi; T = 40–50 km for Serenitatis, Orientale, and Humorum; T ≃ 50–75 km for Imbrium; and T > 75 km for Nectaris, Smythii, and Crisium. The distribution of mare ridges and the topographic relief of present mare surfaces are matched by a greater lithospheric thickness (T ∼ 100 km) at the time of emplacement of the youngest mare units for most mare basins, and the required spatial variation in lithospheric thickness at ∼3.0 b.y. may be less than at ∼3.6 b.y. The growth of the lunar lithosphere beneath each mare basin is a natural consequence of the cooling of the outer portions of the moon. The global synchronism for the cessation of linear rille formation can be explained as being due to the superposition onto the local stress of a global thermal stress that shifted from extensional to compressional as the moon changed from net expansion to net contraction at or before 3.6 ± 0.2 b.y. ago.

A pronounced spatial variation in effective lithosphere thickness during the time of early mare volcanism is clearly indicated. This variation is not simply the product of the average lunar thermal evolution, nor is it explainable solely on the basis of differences in mare fill age or in basin size or basin age. These variations likely represented large-scale inhomogeneities in the thermal structure of the lunar crust and uppermost mantle, arising from lateral variations in crustal heat sources, crustal heat transport, or sublithospheric heat flow. These inhomogeneities apparently lessened in magnitude with time as thermal variations were smoothed out by crustal heat conduction.