BIRD: LITHOSPHERE DYNAMICS AND CONTINENTAL DEFORMATION 381 Stress Field of North America the explanation for the Quaternary Wilshire arch under The publication of the World Stress Map has Los Angeles and Hollywood [Hummon et al, 1994] focussed thinking on the big picture, Zoback [1992] showed that to first order the world stress field is the Alaskan neotectonics result of ridge-push and continental collision resistance, with little evidence for strong basal tractions A compilation of 621 stress indicators in Alaska Richardson and Reding [1992] used thin-plate elastic shows a fan pattern of o, directions radiating from the syntaxis in the St. Elias range [Estabrook and Jacob, models to show that both shear and super-lithostatic normal tractions on the san andreas and caribbean 1991- In a uniform elastic material, such a stress pattern would suggest a stress singularity, or at least transforms are only 5 to 10x10 Pa, and that this value high values. However, a nonlinear thin-plate finite can be explained primarily by ridge-push effects. A collection of surface-wave moment tensors from 51 element model with faults [Bird, 1992b] shows that the western U.S. events confirms the o. directions of the stress magnitudes are low, and that the pattern is due to World Stress Map, and shows a convergence of fault with terrane collision of the yakutat block. thus tensional axes(o3) on the Mendocino triple junction Alaska is no exception to the general weakness of major As efforts are gradually applied to study paleostress faults. The same model also predicted rapid westward tra attention should be paid to the exhaustive study of consistent with the evidence of seismic slip vectors Bergerat et al. [1992 who used joints and faults to infer a 9-stage stress history for the Colorado Plateau [McCaffrey, 1992] the Jul be calcite twinning in limestones, which Craddock et al. [1993] Neotectonics of the new madrid seismic used to map the Paleozoic strain(and paleostress)field in the eastern U.S Li and Schweig [1993] questioned the view that the Mississippi embayment and New Madrid seismic zone California Neotectonics represents simple reactivations of the Precambrian Reelfoot rift. After observed S-waves that were Furlong [1993] presented an elegant synthesis of converted from P-waves to improve their velocity how the northward migration of the Mendocino triple model, chiu et al.[1993] relocated events and found a junction removed the Gorda plate from under North America, creating a "mantle San Andreas transform simple Califormia-type pattern with thrusting in a left which was to the northeast of the surficial margin; with step of a dextral strike-slip system. The important time, surficial fault activity jumped inboard, as from the questions remaining concern the overall rate of San Andreas to the Hayward-Calaveras fault system in deformation, and how(if) these faults connect to plate the San Francisco Bay area. The complicating effects of boundaries at each end. Liu et al. [1992] reoccupied a slightly convergent Pacific plate velocity since 7 Ma triangulation stations with GPS and found rapid strain are being recognized; in particular, seismic studies show accumulation (0.08 urad/a) with a N67 E shortening probable oceanic crust of the Pacific(?) plate direction(in agreement with stress data) and a relative underthrust along the California margin from Morro elocity of 5-7 mm/year across the network. Such a rate Bay north to San Francisco Page and Brocher, 1993] would obviously imply significant seismic hazard along strike from the 1811-1812 The maturity of California neotectonic studies is shown by the fact that Bird and Kong [1994] were able confirmation of the result by homogeneous geodetic to predict fault slip rates and geodesy to within 3 methods should be a high priority. If it is confirmed, we mm/year in a finite-element model; such models may may have a unique opportunity to study new faults in a serve as supplements to hard data in seismic hazard relatively simple midcontinental setting In the area of hazard studies, it is being realized that Huge Displacements of the Mantle few new(Pliocene-Quaternary) thrust faults in the Lithosphere Transverse Ranges actually break the surface, instead anticlines involving Holocene sediments should b Bird[1992a;,1994 assumed to overlie seismically dangerous blind thrusts formation of the Rocky Mountains and the Basin/ Shaw and Suppe [1994 used reflection sections and province by a single flat subduction event, whi balanced-section methods to infer three active thrust suggested had sheared away and displaced the entire under the Santa Barbara Channel, with slip rates that tectonic mantle lithosphere of the western U.S dd up to 3 mm/year (or half the geodetically According to this model, the only tectonic mantle determined rate of shortening ). A Wilshire fault with a lithosphere( defined as cold and strong) remaining in the slip rate of 1. 5-3.2 mm/year has also been proposed western U.S. should be a =40-km layer which has formed by cooling since mid-Tertiary timesBIRD: LITHOSPHERE DYNAMICS AND CONTINENTAL DEFORMATION 381 Stress Field of North America The publication of the World Stress Map has focussed thinking on the big picture; Zoback [1992] showed that to first order the world stress field is the result of ridge-push and continental collision resistance, with little evidence for strong basal tractions. Richardson and Reding [1992] used thin-plate elastic models to show that both shear and super-lithostatic normal tractions on the San Andreas and Caribbean transforms are only 5 to 10×106 Pa, and that this value can be explained primarily by ridge-push effects. A collection of surface-wave moment tensors from 51 western U.S. events confirms the σ1 directions of the World Stress Map, and shows a convergence of tensional axes (σ3 ) on the Mendocino triple junction [Patton and Zandt, 1991]. As efforts are gradually applied to study paleostress, attention should be paid to the exhaustive study of Bergerat et al. [1992], who used joints and faults to infer a 9-stage stress history for the Colorado Plateau since the Jurassic. Another source of data can be calcite twinning in limestones, which Craddock et al. [1993] used to map the Paleozoic strain (and paleostress) field in the eastern U.S. California Neotectonics Furlong [1993] presented an elegant synthesis of how the northward migration of the Mendocino triple junction removed the Gorda plate from under North America, creating a "mantle San Andreas transform" which was to the northeast of the surficial margin; with time, surficial fault activity jumped inboard, as from the San Andreas to the Hayward-Calaveras fault system in the San Francisco Bay area. The complicating effects of a slightly convergent Pacific plate velocity since 7 Ma are being recognized; in particular, seismic studies show probable oceanic crust of the Pacific(?) plate underthrust along the California margin from Morro Bay north to San Francisco [Page and Brocher, 1993]. The maturity of California neotectonic studies is shown by the fact that Bird and Kong [1994] were able to predict fault slip rates and geodesy to within 3 mm/year in a finite-element model; such models may serve as supplements to hard data in seismic hazard estimation. In the area of hazard studies, it is being realized that few new (Pliocene-Quaternary) thrust faults in the Transverse Ranges actually break the surface; instead, anticlines involving Holocene sediments should be assumed to overlie seismically dangerous blind thrusts. Shaw and Suppe [1994] used reflection sections and balanced-section methods to infer three active thrusts under the Santa Barbara Channel, with slip rates that add up to 3 mm/year (or half the geodetically￾determined rate of shortening). A Wilshire fault with a slip rate of 1.5-3.2 mm/year has also been proposed as the explanation for the Quaternary Wilshire arch under Los Angeles and Hollywood [Hummon et al., 1994]. Alaskan Neotectonics A compilation of 621 stress indicators in Alaska shows a fan pattern of σ1 directions radiating from the syntaxis in the St. Elias range [Estabrook and Jacob, 1991]. In a uniform elastic material, such a stress pattern would suggest a stress singularity, or at least high values. However, a nonlinear thin-plate finite￾element model with faults [Bird, 1992b] shows that the stress magnitudes are low, and that the pattern is due to the juxtaposition of transpression on the Fairweather fault with terrane collision of the Yakutat block. Thus, Alaska is no exception to the general weakness of major faults. The same model also predicted rapid westward transport of the west Aleutian forearc, which is consistent with the evidence of seismic slip vectors [McCaffrey, 1992]. Neotectonics of the New Madrid Seismic Zone Li and Schweig [1993] questioned the view that the Mississippi embayment and New Madrid seismic zone represents simple reactivations of the Precambrian Reelfoot rift. After observed S-waves that were converted from P-waves to improve their velocity model, Chiu et al. [1993] relocated events and found a simple California-type pattern with thrusting in a left step of a dextral strike-slip system. The important questions remaining concern the overall rate of deformation, and how (if) these faults connect to plate boundaries at each end. Liu et al. [1992] reoccupied triangulation stations with GPS and found rapid strain accumulation (0.08 µrad/a) with a N67°E shortening direction (in agreement with stress data) and a relative velocity of 5-7 mm/year across the network. Such a rate would obviously imply significant seismic hazard along strike from the 1811-1812 aftershock zone, so confirmation of the result by homogeneous geodetic methods should be a high priority. If it is confirmed, we may have a unique opportunity to study new faults in a relatively simple midcontinental setting. Huge Displacements of the Mantle Lithosphere? Bird [1992a; 1994] modeled the simultaneous formation of the Rocky Mountains and the Basin/Range province by a single flat subduction event, which he suggested had sheared away and displaced the entire tectonic mantle lithosphere of the western U.S.. According to this model, the only tectonic mantle lithosphere (defined as cold and strong) remaining in the western U.S. should be a ≈40-km layer which has formed by cooling since mid-Tertiary times