One-hundred and forty-seven new apatite (U-Th)/He (AHe) ages are presented from 32 sample locations along the flanks of the Rio Grande rift in New Mexico and Colorado. These data are combined with apatite fission-track (AFT) analyses of the same rocks and modeled together to create well-constrained cooling histories for Rio Grande rift flank shoulders. The data indicate rapid cooling due to extension from ca. 28 to 5 Ma in the Sawatch Range, ca. 28 Ma to Quaternary in the Sangre de Cristo Mountains, ca. 25 to 5 Ma in the Albuquerque Basin, and ca. 25 to 10 Ma in the southern Rio Grande rift in southern New Mexico. Rapid cooling of rift flanks followed the Oligocene ignimbrite flare-up, and the northern section of the Rio Grande rift in Colorado exhibits semicontinuous cooling since the Oligocene. Overall, however, rift flank cooling along the length of the rift was out of phase with high-volume magmatism and hence is inferred to have been driven mainly by exhumation due to faulting. Although each location preserves a unique cooling history, when combined with existing AHe data from the Gore Range in northern Colorado and the Sandia Mountains in New Mexico, together these data indicate that extension and exhumation of rift shoulders were synchronous along >850 km of the length of the Rio Grande rift from 25 to 10 Ma.
These time-space constraints provide an important new data set with which to develop geodynamic models for initiation and evolution of continental rifting. Models involving northward propagation of rifting and Colorado Plateau rotation are not favored as primary mechanisms driving extension. Instead, a geodynamic model is proposed that involves upper-mantle dynamics during multistage foundering and rollback of a segment of the Farallon plate near the Laramide hinge region that extended between the Wyoming and SE New Mexico high-velocity mantle domains. The first stage of flat slab removal accompanied ca. 40–20 Ma volcanism in the San Juan and Mogollon-Datil ignimbrite centers, which initiated asthenospheric upwelling and circulation. A second stage involved a ca. 30–25 Ma detachment of remaining fragments of the Farallon slab, intensifying asthenospheric upwelling and focusing it along a N-S trend beneath Colorado and New Mexico. By 25 Ma, the North American lithosphere had become weakened critically along this narrow zone, so that extension was accelerated, resulting in the observed 25–10 Ma cooling indicated by the thermochronologic data. This developed a central graben with increased fault-related high strain rates and resulted in maximum sediment accumulation in the Rio Grande rift. Our geodynamic model thus involves Oligocene removal of parts of the Farallon slab beneath the ignimbrite centers followed by a major Oligocene–Miocene slab break that instigated the discrete N-S Rio Grande rift through focused upper-mantle convection beneath the southern Rocky Mountain–Rio Grande rift region.