Large numbers of earthquakes occur in subduction zones that are marked by dipping, narrow high seismic velocity slabs. The existence of these fast velocity slabs can cause serious earthquake mislocation problems that can bias estimates of seismic travel time residuals. This can affect the recovery of subducting slabs in tomography as well as introduce significant artifacts into lower mantle structure in tomography models. In order to better account for known subducting slabs, we performed a new P and S wave joint tomography inversion incorporating a three-dimensional thermal model of subducting slabs in the starting model. In addition, velocity and source locations were inverted for simultaneously. Our new P and S models feature higher-amplitude subducting slabs compared with previous global tomography results. The S to P heterogeneity ratio based on the new tomography model indicates that thermal elastic effects alone cannot explain all the heterogeneities in the lower mantle. Much of the observed abnormal S to P heterogeneity ratio can be explained by anelastic effects, the spin transition, and phase transitions of bridgmanite to post-perovskite in the lower mantle.Plain Language Summary Seismic tomography uses seismic travel time data to image deep earth velocity structure. However, it has been shown that the existence of subducting slabs can significantly bias the imaging result. In order to reduce this effect, we produced a new P and S wave tomography model that included theoretical three-dimensional subducting slab structures in the starting model. The new model has higher-amplitude subducting slabs compared with other models. Based on the new model, we conclude that it is difficult to explain the P and S velocity anomalies found in the deep mantle by temperature variations alone without invoking complex mineral phase transitions, large anelastic effects, or chemical variations. Key Points:• We developed new P and S tomography models incorporating 3-D subducting slabs • Mislocation effects caused by subducting slabs were reduced by inverting for velocity and source location simultaneously • The spin transition and the post-perovskite phase transition may explain lower mantle heterogeneitiesSupporting Information:• Supporting Information S1

Seismic tomography has revealed the existence of large-scale velocity heterogeneities in the mantle. The interpretation of seismic velocity anomalies in terms of temperature and chemical composition is nonunique. We use geodynamic observations including gravity, plate motions, dynamic topography, and excess ellipticity of the core-mantle boundary combined with seismic observations to investigate the thermo-chemical structure of the mantle through joint inversions. An outstanding issue, however, is the physical connection between mantle density anomalies and the surface geodynamic observations, which requires knowledge of the mantle viscosity structure. Here we perform joint inversions assuming different viscosity profiles and examine the dependence of the results on the viscosity. We first assume that mantle heterogeneity is due to thermal variations, which places a constraint on the relation between seismic velocity and density, and we subsequently relax the constraint to allow for potential nonthermal effects. In all of our joint inversions, a nonthermal origin of density anomalies is required to explain the geodynamic data, though the amount varies with the assumed viscosity structure. A common observation is a high-density chemical signal in the center of the large low-shear-velocity provinces at the base of the mantle resulting in a near neutral or slightly dense overall buoyancy there. Using the derived density models and their corresponding viscosity profiles, we also calculate instantaneous mantle flow fields. The predicted flow fields derived from joint inversions are generally similar but are quite different from flow fields using density models derived from a posteriori scaling of pure seismic tomography models.Plain Language Summary The origin and evolution of Earth's mantle have been long-standing fundamental questions in geosciences. We use both global seismological and geodynamical (gravity, topography, plate motions, and excess ellipticity of the core-mantle boundary) data sets to investigate whether lateral changes in mantle structure can be explained solely by temperature variations or whether the mantle must also have significant chemical variations. Our results indicate the presence of chemically distinct mantle anomalies. In particular, we find two large regions at the base of the mantle that appear to be chemically distinct with hot mantle upwellings surrounding them. We also derived several models of 3D density variations in the mantle assuming different viscosity profiles. These models were then used to predict the present-day mantle convective flow. We show that the mantle viscosity structure does not have a strong influence on the pattern of large-scale mantle flow. We find, however, that 3D density models derived by simple (1D) a posteriori scaling of tomography models obtained only from seismic data yield predictions of significantly different mantle flow.

The choice of the reference electrode scheme is an important step in event-related potential (ERP) analysis. In order to explore the optimal electroencephalogram reference electrode scheme for the ERP signal related to facial recognition, we investigated the influence of average reference (AR), mean mastoid reference (MM), and Reference Electrode Standardization Technique (REST) on the N170 component via statistical analysis, statistical parametric scalp mappings (SPSM) and source analysis. The statistical results showed that the choice of reference electrode scheme has little effect on N170 latency ( p > 0.05), but has an significant impact on N170 amplitude ( p < 0.05). ANOVA results show that, for the three references scheme, there was statistically significant difference between N170 latency and amplitude induced by the unfamiliar face and that induced by the scrambled face ( p < 0.05). Specifically, the SPSM results show an anterior and a temporo-occipital distribution for AR and REST, whereas just anterior distribution shown for MM. However, the circumstantial evidence provided by source analysis is more consistent with SPSM of AR and REST, compared with that of MM. These results indicate that the experimental results under the AR and REST references are more objective and appropriate. Thus, it is more appropriate to use AR and REST reference scheme settings in future facial recognition experiments.

We present a method for constructing the average waveform shape (hereafter called “empirical wavelet”) of seismic shear waves on an event‐by‐event basis for the purpose of constructing a high‐quality travel time data set with information about waveform quality and shape. A global data set was assembled from 360 earthquakes between 1994 and 2017. The empirical wavelet approach permits documentation of the degree of similarity of every observed wave with the empirical wavelet. We adapt the empirical wavelet to all pulse widths, thus identifying broadened (e.g., attenuated) pulses. Several measures of goodness of fit of the empirical wavelet to each record are documented, as well as signal‐to‐noise ratios, permitting users of the data set to employ flexible weighting schemes. We demonstrate the approach on transversely polarized SH waves and build a global travel time data set for the waves S, SS, SSS, Sdiff, ScS, and ScSScS. Onset arrival times of the waves were determined through a correlation scheme with best‐fitting empirical wavelets. Over 250,000 travel times were picked, from over 1.4 million records, all of which were human‐checked for accuracy via a Portable Document Format (PDF) catalog file making system. Many events were specifically selected to bolster southern hemisphere coverage. Coverage maps show that, while the northern hemisphere is more densely sampled, the southern hemisphere coverage is robust. The travel time data set, empirical wavelets, and all measurement metrics are publicly available and well suited for global tomography, as well as forward modeling experiments.

We construct geographically localized bin stacks of waveforms, called virtual stations, to enhance signal‐to‐noise ratios (SNRs) for travel time and waveform measurements of multibounce S and ScS phases (S up to S6 and ScS up to ScS5), as well as direct S, ScS, and Sdiff, on tangential component data. Major arc S and ScS multibounce waves were also measured. Virtual station data are referenced to empirical wavelets constructed from direct S waves for each event. The virtual station approach is useful for low SNR data, bolstering wave path coverage in the southern hemisphere. Goodness of fit measurements between the adapted empirical wavelet and virtual station waveforms are documented, as well as SNRs, allowing for objective definition of travel time measurement quality. From a data set of 360 earthquakes and 8,407 seismographic stations, nearly 4 million records were utilized to construct 248,657 virtual station stacked seismograms, which were compared to best‐fitting empirical wavelets. After human inspection of virtual station results, 8,871 travel time measurements were retained from 19 different minor and major arc seismic wave types. Higher multibounce data improve sampling of the southern hemisphere. From 188,003 single seismograms, 3,331 multibounce wave measurements were also made. Comparisons of single seismogram and virtual station stack measurements show a consistent bias: Virtual stack onset times are systematically early due to a broadening effect from stacking records with arrival time differences, which we correct for. The travel time and waveform measurements are publicly available.