Shear wave velocity, seismic attenuation, and thermal structure of the continental upper mantle

Seismic velocity and attenuation anomalies in the mantle are commonly interpreted in terms of temperature variations on the basis of laboratory studies of elastic and anelastic properties of rocks. In order to evaluate the relative contributions of thermal and non- thermal effects on anomalies of attenuation of seismic shear waves, Q(s)(-1) and seismic velocity, V-s, we compare global maps of the thermal structure of the continental upper mantle with global Q(s)(-1) and V-s maps as determined from Rayleigh waves at periods between 40 and 150 s. We limit the comparison to three continental mantle depths (50, 100 and 150 km), where model resolution is relatively high. The available data set does not indicate that, at a global scale, seismic anomalies in the upper mantle are controlled solely by temperature variations. Continental maps have correlation coefficients of < 0.56 between V-s and T and of < 0.47 between Q(s) and T at any depth. Such low correlation coefficients can partially be attributed to modelling artefacts; however, they also suggest that not all of the V-s and Q(s) anomalies in the continental upper mantle can be explained by T variations. Global maps show that, by the sign of the anomaly, V-s and Q(s) usually inversely correlate with lithospheric temperatures: most cratonic regions show high V-s and Q(s) and low T, while most active regions have seismic and thermal anomalies of the opposite sign. The strongest inverse correlation is found at a depth of 100 km, where the attenuation model is best resolved. Significantly, at this depth, the contours of near- zero Q(s) anomalies approximately correspond to the 1000degrees C isotherm, in agreement with laboratory measurements that show a pronounced increase in seismic attenuation in upper mantle rocks at 1000-1100degreesC. East- west profiles of V-s, Q(s) and T where continental data coverage is best (50degreesN latitude for North America and 60degreesN latitude for Eurasia) further demonstrate that temperature plays a dominant, but non- unique, role in determining the value of lithospheric V-s and Q(s). At 100 km depth, where the resolution of seismic models is the highest, we compare observed seismic V-s and Q(s) with theoretical V-S(T) and Q(S)(T) values, respectively, that are calculated solely from temperature anomalies and constrained by experimental data on temperature dependencies of velocity and attenuation. This comparison shows that temperature variations alone are sufficient to explain seismic V-s and Q(s) in ca 50 per cent of continental regions. We hypothesize that compositional anomalies resulting from Fe depletion can explain the misfit between seismic and theoretical V-s in cratonic lithosphere. In regions of active tectonics, temperature effects alone cannot explain seismic V-s and Q(s) in the lithosphere. It is likely that partial melts and/ or fluids may affect seismic parameters in these regions. This study demonstrates that lithospheric temperature plays the dominant role in controlling V-s and Q(s) anomalies, but other physical parameters, such as compositional variations, fluids, partial melting and scattering, may also play a significant role in determining V-s and Q(s) variations in the continental mantle.