Average physical structure of cratonic lithosphere, from thermodynamic inversion of global surface-wave data

Abstract

Seismic and mantle-xenolith data both show that cratons have the thickest, coldest lithosphere of the Earth’s tectonic environments. Yet, the quantitative depth distributions of temperature, seismic velocity and mass density (herein referred to as “density”) in cratonic lithosphere are uncertain, even on average across all cratons. Seismic surface-wave data offer abundant information on the thermal structure of the lithosphere at present, but seismic-velocity profiles in tomographic and other seismic models are highly non-unique at the relevant depth scale lengths of tens of kilometres. Here, we relate surface-wave measurements averaged over all cratons globally to the average physical properties of cratonic lithosphere using the recently developed methods of seismic thermography. The thermodynamic inversion of the Rayleigh and Love wave phase-velocity curves yields a model of the average structure of cratonic lithosphere, including the profiles of temperature, S- and P-wave seismic velocities, density and radial seismic anisotropy. Average depleted peridotite composition of the cratonic lithosphere was taken from the literature. Assuming 1290 °C as the temperature at which convection commences, which defines the bottom of the mechanical lithosphere, the best-fitting average depth of the cratonic lithosphere-asthenosphere boundary (LAB) is 228 km, with an uncertainty range of 211–242 km, as estimated using the model-space projection approach. The model fits the observed phase velocities very closely (misfits < 0.055%), while also matching the observed topography and surface heat flow. Assuming a lower LAB temperature results in a shallower LAB and a broader transition from the conductive lithospheric geotherm to the mantle adiabat, with a similar fit to the data. Our craton-average lithospheric model offers a useful reference for geophysical studies and for the joint analysis of geochemical and geophysical data. It confirms that cratonic lithosphere is, on average, isopycnic: cratonic and non-cratonic upper-mantle density profiles are very similar. A large proportion of published pressure–temperature measurements from mantle xenoliths is close to our craton-average lithospheric geotherm. Many of the measurements from below 150 km depth, however, show temperatures significantly higher than cratonic average, which offers useful evidence on the evolution of cratons and generation of kimberlites.