Seafloor bathymetry is not a paleoclimate proxy

Abstract:

Recent studies have proposed that the fabric of seafloor formed at mid-ocean ridges (MORs) records rapid (23-100 kyr) fluctuations in ridge magma supply caused by sea level changes modulating melt production in the underlying mantle. We assess the plausibility of this claim using models of crustal emplacement and faulting to derive the response function of seafloor bathymetry to a fluctuating ridge magmatic flux.

First, we show that crustal thickness variations inherited from melt supply modulation on Milankovitch time scales cannot generate any sizeable topography through isostatic or flexural compensation. Further, we expect this modulation to only generate limited (<< 600 m) crustal thickness variations, because the magma storage region beneath the ridge axis is sufficiently wide to damp rapid oscillations in melt flux.

Second, we consider the development of topography related to volcanic extrusion on the seafloor modulated by changes in crustal thickness. We estimate this process to generate at most 40 m of topography, which could be significantly reduced by cross-axis lava flow emplacement.

Finally, if seafloor bathymetry did record Milankovitch cycles, its dominant wavelengths should increase with increasing spreading rate. This is opposite to the observation of abyssal hill spacing decreasing from ~10 km down to ~2 km at slow to fast-spreading MORs. This observation is best explained by a model in which dominant seafloor wavelengths correspond to the spacing between successive normal faults that interact with dike injection at the ridge axis as they grow. Our models show that fault spacing is controlled by the time-average rate of magma injection and is insensitive to variations in injection rates on all Milankovitch frequencies.

Thus, we conclude that the "climate frequencies" found in some bathymetry datasets reflect only the inherent spacing of faults, set by the mechanical properties of the lithosphere and the average MOR magma supply.