Hypothesis-testing proposed control of strain weakening by crustal quartz abundance

Lowry and Pérez-Gussinyé (Nature, 2011) observed a surprising correlation of low crustal seismic velocity ratio, V_{P/V_S}, with both high lithospheric temperature and Cordilleran deformation. We hypothesize that the relationship is most plausibly explained by a robust dynamical feedback in which ductile strain first localizes in relatively weak, quartz-rich crust, and then initiates processes that promote advective warming, hydration, and further weakening. Strain-weakening by such a feedback mechanism not only would explain Wilson-cycle stationarity and spatial distributions of deformation, but it also would lend insight into the timing and distribution of thermal uplift. Here we test the hypothesis by examining heat transfer processes in the Cordillera. Preliminary modeling of western US thermal structure, assuming crustal heat production that linearly relates surface heat flow to surface radiogenic heating, suggests that nearly half of all thermal transfer in Cordilleran lithosphere is advective. We will test the effects of uncertainty in crustal heat-production by examining alternative thermal models with upper mantle temperatures derived from P_n and Rayleigh phase velocity measurements, as well as heat production modeled for compositions inferred from crustal V_{P/V_S}. Assuming the estimate proves robust, advective heat transfer in the Cordillera is roughly an order-of-magnitude higher than can be explained by McKenzie-style stretching at geodetically-observed strain rates. This implies a very significant role for mass transfer via melts and other fluids. Spatial frequency distributions of NAVDAT volcanism in the western US suggest a strong bias toward extrusion in quartz-rich crust, opposite the relationship expected if spatial expressions of volcanism were primarily controlled by melt/crustal relative buoyancy. High ³He/⁴He ratios indicative of a mantle contribution similarly suggest a bias toward regions of quartz-rich crust. Both of these observations are consistent with the strain-weakening mechanism proposed by Lowry & Pérez-Gussinyé (2011). Intriguingly, preliminary analyses of the relationships of melt geochemistry to V_{P/V_S} and crustal thickness indicate that low melt fractions occur preferentially under thicker crust, but California and the highly-extended Basin-Range province express a somewhat different relationship than the Colorado Plateau-Rio Grande Rift region, and both behave more simply than the Pacific northwest where subduction and plume processes appear to dominate over lithospheric strain. The full meaning of these and other relationships is as yet uncertain. We will discuss several possible implications.