Will Levandowski

Origins of topography in the western U.S.: Mapping crustal and upper mantle density variations using a uniform seismic velocity model

To investigate the physical basis for support of topography in the western U.S., we construct a subcontinent scale, 3-D density model using ~1000 estimated crustal thicknesses and S velocity profiles to 150 km depth at each of 947 seismic stations. Crustal temperature and composition are considered, but we assume that mantle velocity variations are thermal in origin. From these densities, we calculate crustal and mantle topographic contributions. Typical 2σ uncertainty of topography is ~500 m, and elevations in 84% of the region are reproduced within error. Remaining deviations from observed elevations are attributed to melt, variations in crustal quartz content, and dynamic topography; compositional variations in the mantle, while plausible, are not necessary to reproduce topography. Support for western U.S. topography is heterogeneous, with each province having a unique combination of mechanisms. Topography due to mantle buoyancy is nearly constant (within ~250 m) across the Cordillera; relief there (>2 km) results from variations in crustal chemistry and thickness. Cold mantle provides ~1.5 km of ballast to the thick crust of the Great Plains and Wyoming craton. Crustal temperature variations and dynamic pressures have smaller magnitude and/or more localized impacts. Positive gravitational potential energy (GPE) anomalies (~2 × 1012N/m) calculated from our model promote extension in the northern Basin and Range and near the Sierra Nevada. Negative GPE anomalies (−3 × 1012N/m) along the western North American margin and Yakima fold and thrust belt add compressive stresses. Stresses derived from lithospheric density variations may strongly modulate tectonic stresses in the western U.S. continental interior.

Seismological estimates of means of isostatic support of the Sierra Nevada

The modern topography of the Sierra Nevada (California, USA) has been attributed to rapid uplift following foundering of negatively buoyant lithosphere into the asthenosphere since ca. 10 Ma. Uplift now manifests as ∼2 km mean topographic relief between the crest of the southern Sierra Nevada and the western foothills and 1–2 km between the Sierran crest and adjacent Basin and Range. In this study, we use seismic P-wave velocity structures derived from teleseismic tomography to estimate the lithospheric density structure in the region and thus infer the current sources of topographic support. We exploit the different derivatives of crustal density with temperature and wave speed to attempt to identify a single solution for crustal density and temperature that satisfies flexural isostasy and the P-wave tomography. This solution yields both temperature variations compatible with observed heat flow and Bouguer gravity anomalies concordant with observations. We find that the topographic gradient between the crest and the eastern Great Valley is due to both crustal and mantle sources. Despite a greater thickness, the foothills crust is less buoyant than that beneath the range crest, accounting for ∼1 km of the topographic difference. High densities are due principally to composition. High-velocity upper mantle (∼50–100 km depth) is also observed beneath the foothills but not the range crest, and this contrast explains an additional 1 km of topographic difference. Miocene or more recent removal of such upper mantle material from the Sierran crest, as inferred from xenoliths, would have triggered rapid uplift of ∼1 km. Our findings are consistent with the removal of negatively buoyant material from beneath the Sierra Nevada since the Miocene.

Linking Sierra Nevada, California, uplift to subsidence of the Tulare basin using a seismically derived density model

Seismic tomography has previously imaged the high-velocity “Isabella anomaly” southwest of the Sierra Nevada beneath the Tulare basin, a region of ~1 km of anomalous Pliocene subsidence. Additionally, it has been proposed that the eastern Sierra has risen 1–2 km since the Miocene in response to removal of dense lithospheric material. The Isabella anomaly has been variably interpreted as either this lithospheric material or a neutrally buoyant stalled fragment of the Farallon slab. To discriminate between these two, we estimate upper mantle density variations from seismic velocities and show that the estimated mass anomaly accords with 60 km of cold lithospheric material removed from beneath the southern Sierra, sufficient for 1.3 km of range uplift. A flexural model of the surface response to mantle loads predicts 1.3–1.7 km of anomalous subsidence of the Tulare basin, several hundred meters more than is observed. Nevertheless, beam-formed receiver functions show up to 10 km of crustal thickening beneath the basin, which we attribute to viscous response of the crust to mantle loading. This anomalous crustal thickness, the post-Miocene subsidence of the Tulare basin, and the uplift of the Sierra can all be explained by redistribution of cold continental mantle lithosphere; therefore, the Isabella anomaly is more plausibly such continental material than a stalled Farallon slab fragment.

An updated stress map of the continental United States reveals heterogeneous intraplate stress

Knowledge of the state of stress in Earth’s crust is key to understanding the forces and processes responsible for earthquakes. Historically, low rates of natural seismicity in the central and eastern United States have complicated efforts to understand intraplate stress, but recent improvements in seismic networks and the spread of human-induced seismicity have greatly improved data coverage. Here, we compile a nationwide stress map based on formal inversions of focal mechanisms that challenges the idea that deformation in continental interiors is driven primarily by broad, uniform stress fields derived from distant plate boundaries. Despite plate-boundary compression, extension dominates roughly half of the continent, and second-order forces related to lithospheric structure appear to control extension directions. We also show that the states of stress in several active eastern United States seismic zones differ significantly from those of surrounding areas and that these anomalies cannot be explained by transient processes, suggesting that earthquakes are focused by persistent, locally derived sources of stress. Such spatially variable intraplate stress appears to justify the current, spatially variable estimates of seismic hazard. Future work to quantify sources of stress, stressing-rate magnitudes and their relationship with strain and earthquake rates could allow prospective mapping of intraplate hazard.Crustal stress in the interior of the United States is spatially variable and largely controlled by local forces, rather than those transmitted from tectonic plate boundaries, according to a map of the continental stress field.