MELTING OF COLUMBIA RIVER FLOOD BASALT SOURCE COMPONENTS BY DELAMINATION ABOVE THE YELLOWSTONE MANTLE-PLUME HEAD

The Columbia River Basalt Group (CRBG) differs from all other flood-basalt provinces in being dominated by high-Si, high-Fe basaltic andesites of the Grande Ronde Formation. Several workers have shown that these evolved lavas are not likely generated by partial melting of peridotite, but derive instead from a mafic eclogite or pyroxenite source. Such a source is one of several components evident in the CRBG isotopic data. A plume component, shared by all CRBG units, is least diluted in Imnaha Basalt, but is mixed with other components that first appear in the following order: depleted mantle (Steens Basalt), mafic crust (Grande Ronde Basalt) and Archean lithosphere (Wanapum and Saddle Mtns. Basalts). This sequence of source melting is consistent with a series of delamination events triggered by a northward-propagating mantle plume head, as predicted in recent numerical experiments of plume emplacement adjacent to a thick cratonic boundary. The model results of Burov et al. (in press) provide a rubric consistent not only with the CRBG record of chemical change through time, but also with the derivation of Grande Ronde Basalt from a mafic source in the lower crust. The model predicts that drip-like down-wellings of lithospheric mantle began near center of plume beneath SE Oregon, where melting of the depleted mantle produced MORB-like melts that mixed with plume melts to generate Steens basalt. After advancing into NE Oregon, a slab-like instability (also predicted by the model) developed at the front of the plume head, resulting in the descent of lithospheric mantle and lower crust. Here, enriched mantle from the plume rose into the lithospheric void where decompression generated OIB melts that erupted as Imnaha Basalt. Simultaneously, mafic lower crust of the delaminated slab descended into a thermal environment ~200^oC hotter than the solidus temperature of basalt at asthenospheric depths. Spontaneous melting of this mafic crust generated superheated high-Si, high-Fe lavas of the voluminous Grande Ronde Basalt. A mafic source for the Grande Ronde lavas places an important constraint on CRBG genesis. A model of plume-induced delamination satisfies this constraint while simultaneously providing a progression of melting events consistent with the stratigraphic record of chemical change through time.