On long scales of length and time, Earth’s crust and upper mantle flow like a stiff liquid. To understand how the rocks deform under geologic stresses, you need to know their viscosity—a property that depends on the rocks’ temperature, strain rate, and composition. Among those features, variations in composition, specifically trace amounts of water and magma, are the most difficult to determine but exert a strong influence on the rocks’ behavior (see the article by Marc Hirschmann and David Kohlstedt, Physics Today, March 2012, page 40). The hotter, wetter, or more molten a rock, the weaker it is. Fortuitously, the same factors that weaken a rock and lower its viscosity also make it more electrically conductive. Since the 1950s, researchers have been able to infer resistivity profiles as a function of depth in crustal and mantle rocks from variations in magnetic and electric fields they measure at Earth’s surface. The method, widely used for oil and gas exploration, is known as magnetotelluric (MT) imaging. Geologists Lijun Liu (University of Illinois at Urbana-Champaign) and Derrick Hasterok (University of Adelaide in Australia) have now derived an empirical conversion factor to determine viscosity variations from two-dimensional variations in electrical resistivity obtained from an MT survey across the western US—more specifically, the eastern Great Basin and the Colorado Plateau. The researchers calibrated the magnitudes of the viscosity variations with geodynamic flow models to produce a viscosity map, shown here. Spanning six orders of magnitude, the map predicts the region’s rough surface topography, crustal deformation, and mantle upwelling (upward arrows) more accurately than do standard geological models. The upwelling identifies spots of potential earthquakes or volcanism. (L. Liu, D. Hasterok, Science 353, 1515, 2016.)
