Testing the hypothesis that temperature modulates 410 and 660 discontinuity topography beneath the eastern United States

Abstract

The leading hypothesis to explain 410 and 660 km discontinuity topography and coincident velocity variations is the thermal hypothesis stated as: temperature variations are the primary modulator of discontinuity topography and seismic velocity variations. To test the thermal hypothesis, discontinuity topography maps are correlated with coincident P- and S-velocity variations for the eastern half of the United States sampled by IRIS-EarthScope USArray seismic data. The discontinuity topography maps were made via common-conversion point migration of P-wave receiver functions. The receiver functions were made using a multi-event and multi-station deconvolution method. Fundamental to our results is the choice of three-dimensional P- and S-velocity models, which are used as migration velocity models and for correlation analysis. Two three-dimensional velocity models are used in our analysis: the MITS-model of Golos et al. (2018) and the SL-model of Schmandt and Lin (2014). The Pearson correlation coefficient is used to estimate the degree of linearity between the discontinuity topography and coincident velocity variations. A bivariate regression of discontinuity topography versus coincident velocity variations (termed the mineral physics slope) is performed and compared to a range of slopes constrained by published velocity-temperature derivatives and Clapeyron slopes. Using spatially binning, the discontinuity topography and coincident velocity variations, spatial maps of the correlation coefficient and mineral physics slope are made. Most of the discontinuity sampling area has reasonable correlation values (≥0.4) and plausible mineral physics slope values. The veracity of the thermal hypothesis is assessed by integrating the probability density functions of the mineral physics slopes over a domain defined by the published range of 410 and 660 Clapeyron slopes. At the 410, the MITS-model and SL-model thermal hypothesis probabilities are 52% and 51%, respectively, and the seismic Clapeyron slope estimates are 2.7 and 1.3 MPa/K, respectively. At the 660, the MITS-model and SL-model thermal hypothesis probabilities are 54% and 75%, respectively and the seismic Clapeyron slope estimates are −1.1 and −1.7 MPa/K, respectively. These Clapeyron slopes estimates are in the middle of plausible Clapeyron slope ranges. Using these Clapeyron slopes, temperature maps show a ±300 K range at the 410 and a ±600 K range at the 660. For regions that are inconsistent with the thermal hypothesis, we suggest that the leading explanations are uncertainties in the velocity models used and secondarily, hydration effects.