Subsidence of “normal” seafloor: Observations do indicate “flattening”

Seafloor topography is a key observational constraint upon the evolution of the oceanic lithosphere. Specifically, plots of oceanic depth (z) versus crustal age (t) for “normal” seafloor are well explained by depth‐age predictions of thermal contraction models such as the half‐space and cooling plate model. Old seafloor (t > ∼70 Ma) shallower than that predicted by half‐space cooling (i.e., z ∝ √t), or “flattening,” is a key but debated discriminator between the two models. Korenaga and Korenaga (2008) in a recent paper find normal seafloor depths of all ages to be consistent with a z ∝ √t model, thus supporting a cooling half‐space model for all ages of seafloor. Upon reevaluation, however, the mean depths of their “normal” seafloor flatten at ages >70 Ma, e.g., by 723.2 ± 0.5 m (1 standard error) for t > 110 Ma. This observed inconsistency with the z ∝ √t model is statistically significant (>99.9%) and remains robust (>94%) even if the number of effective independent depth observations is argued to be low (e.g., n = 10). So, if any statistically significant conclusion can be drawn from the observed depths of rare old normal seafloor, it is that old seafloor flattens, which is incompatible with the cooling half‐space model applying to all ages of seafloor but does not preclude a cooling‐plate style approximation to lithospheric evolution.