The Yellowstone plume investigation keeps broadening. In the image below the heavy black line is the Pacific/Farallon ocean spreading ridge at ~30 million years ago when it first encountered North America. The current position of North America is outlined in blue. The path of the North American Continent as defined by the stable corners of the State of Colorado are shown in 10 million year increments by colored arrows on the right. Here the Farallon plate is reconstructed to the east of the ridge in the same colors as to the west.. The Farallon Islands are not on the Farallon Plate.. They are actually the granitic north end of the Salinian/Mojave batholith, sheared off and currently moving north west of the San Andreas Fault, which was just starting 30 Mya. The stripes in the ocean floor representing 5 My increments are based on magnetism reversing with the earth’s magnetic field as the ocean floor lava cooled. These “isochrons” are symmetrical on opposite sides of the ocean spreading ridges all over the planet. Here the isochrons are “restored” by copying and pasting their sisters on the opposite side of the ridge. Copying and pasting works because fracture zone offsets are conserved across ocean spreading ridges. They offset the same way on both sides of the ridge.
This exercise shows some important features. About 80 Mya North America was moving NW, but spun and changed direction about 90 degrees to the SW. This change aimed the continent almost directly at the Mendocino Fracture Zone, a huge offset in the isochrons that can be traced for 150 My across the Pacific Ocean floor. This change in direction marked the end of the Andes style arc volcanism that produced the various granitic batholiths like the Sierra and Baja and a transition to more “flat slab” subduction that caused the Absaroka volcanics at Yellowstone further inland and generally the Laramide uplift of the Rocky Mountains. The more direct opposition in direction of the Farallon and North American plates was likely a factor along with the increasing proximity of the rise producing warmer and more buoyant ocean floor. Laramide time was pretty much over at 30 Ma and this snapshot and the San Andreas phase that would move the Mendocino Fracture north to its present location off Mendocino had begun.
It can be seen that the copied and pasted restored isochrons overlap considerably south of the Mendocino Fracture Zone, likely indicating faster spreading than to the north of it. If they had been split and nested correctly the “Mendocino Projection” would be even greater. The furthest restored isochron was 60 Mya. But seismic tomography shows nothing approaching a coherent slab. Even the best tomographic images through the Cascadia subduction zone today and corroborated by independent seismic loci along the Benioff zone show a slab that is broken. The section line can be seen above the section running through the flatter southern part of the zone from the coast to Utah.
Further south the tomographic sections devolve into jumbled blue blobs. Below are various model interpretations of a section between Southern California and Iowa.
This is the realm of the San Andreas Fault which is inland of the coast ranges here, about a quarter if the way across California in the white ticks. In many of the models slab detachment allowing the mantle to upwell appears to mean letting the shallow melt from the former East Pacific Rise upwell.
geosciencebigpicture 
