REGIONAL SYNTHESIS OF STEENS BASALT, COLUMBIA RIVER BASALT GROUP

The Steens Basalt of southeastern Oregon is the oldest stratigraphic unit of the Columbia River Basalt Group. We estimate its areal extent and volume at ~53,000 km² and ~31,800 km³, respectively. Whereas individual lava flows on the Columbia Plateau are typically 15-50 m thick, with volumes of 100 to >1000 km³, the Steens lavas are much thinner, between 1-5 m, with significantly smaller volumes. The thin nature of these flows is misleading, however, because most occur as stacked pahoehoe flow lobes that erupted in rapid succession to form compound flows that vary in thickness from 10 to ~50 m. Each compound flow forms a distinct chemical group, typically having well-defined chemical trends marked by progressive compositional variations in the constituent flow lobes.

The regional distribution and compositional variation of Steens Basalt has been aided by XRF analyses on 393 samples collected from 13 stratigraphic sections across the Oregon Plateau. The lowermost flows (lower Steens) are tholeiitic basalts with remarkable trace-element homogeneity. These extend across the entire breadth of the Oregon Plateau. In the Plateau’s southern and central parts, the overlying flows (upper Steens) show increasing compositional scatter, varying from transitional basalt to alkalic basalt to trachyandestic lavas toward the top of the section which increase in number and thickness to the south. In the northern and eastern parts of the Plateau, upper Steens Basalt flows show much less chemical variation and are mostly tholeiitic, but they are still typically more evolved than the underlying lower Steens tholeiites.

The chemical recognition of compound flows in the Catlow Peak section is used in deciphering the temporal evolution of the eruption and associated magma-chamber processes. The chemical data is consistent with open-system processes of crystal fractionation combined with the periodic replenishment and tapping of crustal magma chambers. Most compound flows show clear evidence of plagioclase accumulation associated with magma recharge. The chemical data are also consistent with selective contamination of the lavas, perhaps the result of an early stage of crustal assimilation that preceded a major crustal melting event at ~16.5 Ma to produce the Tuff of Oregon Canyon and other rhyolites from the McDermitt caldera.