COSO VOLCANISM IN SPACE AND TIME
Volcanism in both Coso and Long Valley commenced ~3.5 Ma ago. The fields lie at opposite ends of an elliptical area of low-volume, high-K2O lavas that were erupted 3.4-3.6 Ma ago; this weird event has been explained by melting of shallow, metasomatized lithosphere shortly after lithospheric delamination. Volcanism swept into this region, which was free of Cenozoic magmatism (i.e., no hint of an ancestral arc), from the north, east, and south, defying explanation by what is known of late Cenozoic plate interactions. The westward sweep does, however, correlate with a general westward stepping of extension. Pliocene lavas at Coso would be at home in an arc, spanning 48-70 wt% SiO2 with abundant andesite and high field strength element (HFSE) depletions, whereas younger Pleistocene lavas are more bimodal, with 45-60 and 74-78 wt% SiO2 and significantly higher HFSE contents. The switch from arc-like, low-HFSE to bimodal, high-HFSE volcanism occurred ~1.5 Ma ago, well over 10 Ma after subduction shut off at this latitude.
Magmatism at Coso has much more mantle-like Sr and Nd isotopic ratios than Long Valley, and these ratios have shifted to more primitive values over time. Highly evolved late Pleistocene rhyolites have eNd values of +1 to +5, and eNd decreases (i.e., becomes more crust-like) with distance from the geothermal field. These geochemical characteristics indicate that localization of the geothermal field at a releasing stepover has drawn up relatively primitive mantle that was rapidly processed into highly evolved rhyolite, dumping usable energy into the crust in the process.