Abstract At the surface, strike-slip fault stepovers, including abrupt fault bends, are typically regions of complex, often disconnected faults. This complexity has traditionally led geologists studying the hazard of active faults to consider such stepovers as important fault segment boundaries, and to give lower weight to earthquake scenarios that involve rupture through the stepover zone. However, recent geological and geophysical studies of several stepover zones along the San Andreas fault system in California have revealed that the complex nature of the fault zone at the surface masks a much simpler and direct connection at depths associated with large earthquakes (greater than 5 km). In turn, the simplicity of the connection suggests that a stepover zone would provide less of an impediment to through-going rupture in a large earthquake, so that the role of stepovers as segment boundaries has probably been overemphasized. However, counter-examples of fault complexity at depth associated with surface stepovers are known, so the role of stepovers in fault rupture behaviour must be carefully established in each case.

Nearly twenty flows of the Columbia River Basalt Group (CRBG) can be paleomagnetically and chemically correlated westward as far as 500 km from the Columbia Plateau in Washington, through the Columbia Gorge, to the Coast Range of Oregon and Washington. In the Coast Range near Cathlamet, Washington, the CRBG flow stratigraphy includes 10 flows of Grande Ronde Basalt (1 low-MgO R 2 flow, 6 low-MgO N 2 flows, 3 high-MgO N 2 flows), 2 flows of Wanapum Basalt (both flows of Sand Hollow from the Frenchman Springs Member), and the Pomona Member of the Saddle Mountains Basalt. Elsewhere in the Coast Range, additional Grande Ronde Basalt flows, including flows of Winterwater or Umtanum, and additional Wanapum flows, including the flows of Ginkgo, have been reported. Thus at least 18 to 20 CRBG flows reached the coast region. Several of these distal flows have distinctive chemical and magnetic characteristics that are shared by nearby isolated intrusions in Coast Range sedimentary rocks, thus strongly supporting recent suggestions that these intrusions are invasive bodies fed by CRBG flows. Magnetization directions from several flows indicate 16 to 30° of clockwise rotation of the coast with respect to the plateau since middle Miocene time.

A new isostatic residual gravity map of the conterminous United States presents continent-wide gravity data in a form that can be readily used, with geologic information and other geophysical data, in studies of the composition and structure of the continental crust. This map was produced from the gridded gravity data on which the recently released Gravity Anomaly Map of the United States is based. About 1 million onland and 0.8 million offshore gravity observations interpolated to a 4- by 4-km grid serve as the basis for both maps. The Airy-Heiskanen model of isostatic compensation of topography applied to topographic and bathymetric data averaged over 5- by 5-min compartments was used to remove, to first order, the large, long-wavelength Bouguer gravity anomalies caused by deep density distributions that support topographic loads. The parameters used in the Airy-Heiskanen model were topographic density, 2.67 g/cm 3 ; sea-level crustal thickness, 30 km; and density contrast across the base of the model crust, 0.35 g/cm 3 . Many of the conspicuous short-wavelength anomalies (widths less than several hundred kilometers) on the isostatic residual gravity map correlate with mapped or near-surface geologic features, and primarily reflect shallow-density distributions rather than any departures from isostatic equilibrium. In general, gravity highs occur over (1) mafic igneous bodies emplaced in rift or magmatic arc settings or as isolated intrusions controlled by structures; (2) accreted slices of mafic oceanic, island-arc, or transitional crust; and (3) uplifted crystalline basement. Gravity lows are found over (1) thick bodies of felsic intrusive or extrusive rocks; (2) sedimentary deposits in extensional, convergent, or transform settings; and (3) depressed crystalline basement. Anomalies with widths as much as 1,000 km or more also appear to reflect crustal properties in many cases—several broad gravity highs are associated with crust having a high average seismic wave velocity, and comparable broad gravity lows occur over areas of low average seismic velocity. Alternative ways of viewing the isostatic residual gravity data provide additional information about density distributions in the crust. The first-vertical derivative map accentuates gravity anomalies over shallow, abrupt density changes at the expense of those resulting from deeper or more gradual density transitions. The maximum horizontal gradient map contains information about the locations of pronounced density boundaries. Two-dimensional spectral analysis of the gravity data provides a quantitative means for identifying dominant fabrics in the gravity field and for distinguishing various terranes from each other. Neither Bouguer nor isostatic residual gravity anomalies are particularly well suited for practical modeling of deep structure in conjunction with deep seismic information. However, a scheme in which the entire Earth outside the area of interest is approximated by laterally homogeneous layers and isostatically compensated topography, and in which the area of interest is modeled using the seismic constraints applied in a two-and-one-half-dimensional geometry, holds promise for exploiting useful features of both the Bouguer and isostatic residual gravity anomalies.