Synoptic View of Lithospheric S-Wave Velocity Structure in the Southern United States: A Comparison of 3D Seismic Tomographic Models

The southern U.S. continental margin records a history spanning ca. 1.2 Ga, including two Wilson cycles. However, due to a thick sediment cover, the paucity of significant local seismicity, and, until recently, sparse instrumentation, details of this passive margin’s tectonomagmatic evolution remain disputed. This paper compares recent S-wave tomography and crustal thickness models based on USArray data to help establish a framework for geodynamic interpretation. Large-scale patterns of crustal velocity anomalies, corresponding to major regional features such as the Ouachita orogenic front and the Precambrian margin, are generally consistent between the models. The spatial extent of smaller-scale tectonic features, such as the Sabine Uplift and Wiggins block, remains poorly resolved. An inverse relationship between crustal thickness and Bouguer gravity across the continental margin is observed. This model comparison highlights the need for additional P-wave tomography studies and targeted, higher density station deployments to better constrain tectonic features. INTRODUCTION The southern U.S. margin (Fig. 1) ranges from the stable Laurentia craton beneath Oklahoma to a stretched and thinned passive margin to oceanic lithosphere in the deep Gulf of Mexico, recording within it a geologic history that includes two complete Wilson cycles (Thomas, 2006). Due to its extensive hydrocarbon reserves, the southern U.S. has been the focus of intensive seismic exploration. However, until recently, studies of its deep structure trailed those of other U.S. continental margins. The result is that the tectonomagmatic evolution of the southern U.S. margin remains poorly understood. The primary contributing factors to this status quo are (1) the presence of a thick sediment cover that obscures crustal structure through most of the region, (2) the paucity of significant local seismicity, and, until recently, (3) sparse seismic instrumentation in the region. Earthscope’s USArray temporarily densified the set of broadband seismographs available for studies of the region’s lithosphere (http://www.usarray.org/ researchers/obs/transportable). Approximately 435 stations occupied a total of 1830 locations in the continental U.S., for two years each, at a nominal spacing of 70 km. In USArray’s wake, there has been a surge in the number of continental-scale tomographic studies presenting snapshots of the compressional and shear wave velocities of the region’s crust and upper mantle. Although the volume of seismic data available for studies of the region has increased dramatically and sampling of the subsurface has improved as well, the presence of a thick layer of sediments and relatively low levels of seismicity (with the exception of Oklahoma) continue to challenge efforts to image the lithosphere. The collection of models for the southern U.S. region represents the state-of-theart of seismic tomography: a broad range of approaches, the inclusion of various types of data, and different choices of solution schemes. These seismic velocity models can be used to study the mineralogical, compositional, and thermal state of the current crust and upper mantle, and thereby provide critical constraints on geodynamic models, as well as serving as a foundation to launch further investigations. They also showcase the various techniques and innovations of seismic tomography. But, first, robust tectonic features must be identified. Well-constrained features should appear consistently across models. Differences between models could be due to (1) types of data incorporated, such as body wave arrival times, surface wave dispersion, receiver functions, or combinations of two or more data types; (2) measurement techniques employed; (3) the theoretical basis of the forward calculation, such as ray theory versus finite difference versus finite frequency; (4) the initial model and parameterization used; (5) regularization choices (“damping” and “smoothing” schemes and parameter values); and (6) inversion methods, such as gradient-based local minimization versus global optimization techniques. The purpose of this study is to provide a systematic analysis of similarities and differences between recent shear wave tomographic models with respect to the lithospheric structure of the southern U.S. continental margin. Similar comparisons have been conducted for the western U.S. by Becker (2012) and Pavlis et al. (2012).