Imnaha Basalt, Columbia River Basalt Group

The Columbia River volcanic episode began with the eruptionof the coarsely porphyritic Imnaha Basalt between 17.0 and 16.5m.y. B.P. Lava poured from NNW trending vertical fissures andlocal vents north and south of the Seven Devils-Wallowa Mountainsdivide, covering a deeply dissected topographic surface of morethan 30, 000 km², with an estimated volume of 6000 km³. A minimumof 26 flows or flow units are represented in 14 or 15 members.These include 11 chemical types and are exposed in sectionsranging to 577 m in thickness. All flows have normal polaritywith the exception of the youngest and oldest whose polarityis either reversed or transitional. The petrologic and majorelement chemical features of the Imnaha Basalt have much incommon with those of the Picture Gorge Basalt exposed in theJohn Day Basin of north-central Oregon, but the latter is younger,equivalent in age to part of the Grande Ronde Basalt formation. Using major and trace elements, the flows of Imnaha Basalt areclearly distinguished from those of all other formations ofthe Columbia River Basalt Group. Imnaha Basalt has lower SiO₂,K₂O, Ba and Rb than does Grande Ronde Basalt and differs frommost Wanapum and Saddle Mountains Basalt flows in its lowerTiO₂ and P₂O₅ contents. The 11 Imnaha chemical types fall into two subgroups, the AmericanBar (AB) and Rock Creek (RC) subgroups, which differ in thecoarseness of their groundmass, the abundance of olivine, theirphenocryst assemblages, their SiO₂ contents, CaO/Al₂O₃ ratio,and in their Sc, V, Sr, and Ni contents. Flows of the two subgroupsinterdigitate, but AB flows are predominant at the base of thesequence and RC flows at the top. One flow is a hybrid of thetwo magma types. Neither subgroup displays a significant variation in SiO₂ content,but each does show systematic variation in K₂O, P₂O₅, TiO₂,Ba, Zr, Rb, and the REE, all of which vary inversely with MgO/(MgO+ FeO + Fe₂O₃). AB flows show a systematic increase in contentsof the incompatible elements upward in the succession, whileRC flows show a less obvious decrease upwards. Modelling of major and trace elements indicates that the chemicalvariations within each subgroup can be explained by simultaneouscrystal fractionation and assimilation of lower crustal material,in which the mass assimilated is only slightly less than thatlost by crystal fractionation; the mass fractionated varyingup to 50 per cent and the mass assimilated up to 42 per centof the original magma mass. These processes alone cannot explainthe relationships between the two Imnaha subgroups, nor thatbetween either subgroup and the overlying aphyric Grande Rondebasalt. The value of more complex quantitative models, in whichrecharge by more primitive magma, a variable composition forthe lower crustal contaminant, and the partial melting of aheterogeneous source, is limited by lack of data. Some suchprocess, or combination of processes in addition to a combinationof crystal fractionation and lower crustal assimilation, wouldseem to be required to account for the diversity in the earliestColumbia River basalts.