The Mendocino Triple Junction in California is where the North American, Pacific and Gorda plates come together. The North American plate is moving south while the Pacific plate is moving north. These two plates slip by each other with resultant earthquakes on the San Andreas Fault and subsidiary faults. While the Gorda plate is also moving north, it is simultaneously moving east, underneath the North American plate. This subduction of one plate beneath another creates the Cascade mountains, a volcanic range known for Mt. St. Helens and Mt. Rainier.
Recent seismic studies indicated that the crust at the Mendocino Triple Junction was nearly twice as thick as that north and south of the area. "Seismic images provide an instantaneous picture of the current crustal structure, but they do not confirm the mechanics of its creation," says Dr. Kevin P. Furlong, professor of geosciences at Penn State. "The possibility exists that the current crustal thickness pattern is simply inherited, with triple junction migration playing a minimal role in the deformation of the overlying North American plate."
Some researchers suggest that the thickened crustal area just happens to be at the triple junction point. They suggest that the thick crust and triple junction are coincidentally superimposed and that in the past, this was not so.
Furlong and Christopher A. Guzofski, who received a master's degree from Penn State and is now at Harvard, suggest in the latest issue of Geophysical Research Letters, that modeling of heat flow shows the thickening is linked to the passing of the triple junction.
At the point where the three plates come together, a complicated process dubbed the Mendocino crustal conveyer by the researchers, takes place. As the Pacific plate moves north, it drags the North American plate with it. However, the North American plate was moving south, so the crust bunches up on itself and thickens. The material for this thickening comes from south of the junction, an area that had previously thickened as the triple junction moved past.
As the Mendocino Triple Junction traveled north, it left what could be thought of as a giant worm trail. Near the junction, the thickened crust forms the first hump of the worm followed by a valley and a second smaller hump. South of these humps is the area where the worm passed, leaving the underlying rock forever altered by thickening and thinning.
"The coupling of the triple junction with the crustal thickening answers many questions about heat flow in the San Andreas fault area," says Furlong. "Chris found that the heat flow data fits the crustal conveyor model better than the traditional models."
According to the researchers, near the triple junction and north of it, there are heat flows that are quite low. In the past, this low heat flow was problematic because in the subduction zone, while lower heat flows were expected, the actual heat flows were far too low. However, if the additional depression of heat flow is caused by the thickening of the crust, then the temperature readings make sense.
There is a rapid rise in temperatures south of the Mendocino Triple Junction, which was unaccounted for in the past. With the crustal conveyor model, thinning south of the triple junction produces a rise in heat flow because with thinning, the hot areas are closer to the surface.
The numerical heat flow models found that the high heat flow in the coastal ranges, Bay Area and north is still caused by hot materials filling the slab window, the space left as the various plates move.
"The structure of the triple junction is more complicated than before, but now the heat flows observed make sense," says Furlong. "The numerical model also suggests that frictional heating is still very minor along the fault, leaving the San Andreas classified as a weak fault."
The researchers used observations of seismic activity collected by others, their model of the Mendocino crustal conveyor and a finite difference model that calculated the time and space varying temperature structure of the crust. "The static model does not represent what is going on geologically," says Furlong. "The area is actively deforming along the plate boundaries and using a static model to explain activity is dangerous. The crustal conveyor model more closely explains the geologic dynamic in the area."