A comparison of lithospheric thickness models

Abstract

The outermost layer of the solid Earth consists of relatively rigid plates whose horizontal motions are well described by the rules of plate tectonics. Yet, the thickness of these plates is poorly constrained, with different methods giving widely discrepant results. Here a recently developed procedure to derive lithospheric thickness from seismic tomography with a simple thermal model is discussed. Thickness is calibrated such that the average as a function of seafloor age matches the theoretical curve for half-space cooling. Using several recent tomography models, predicted thickness agrees quite well with what is expected from half-space cooling in many oceanic areas younger than ≈110Myr. Thickness increases less strongly with age for older oceanic lithosphere, and is quite variable on continents, with thick lithosphere up to ≈250km inferred for many cratons. Results are highly correlated for recent shear-wave tomography models. Also, comparison to previous approaches based on tomography shows that results remain mostly similar in pattern, although somewhat more variable in the mean value and amount of variation. Global correlations with and between lithosphere thicknesses inferred from receiver functions or heat flow are much lower. However, results inferred from tomography and elastic thickness are correlated highly, giving additional confidence in these patterns of thickness variations, and implying that tomographically inferred thickness may correlate with depth-integrated strength. Thermal scaling from seismic velocities to temperatures yields radial profiles that agree with half-space cooling over large parts of their depth range, in particular for averaged profiles for given lithosphere thickness ranges. However, strong deviations from half-space cooling profiles are found in thick continental lithosphere above depth ≈150km, most likely due to compositional differences.