ABSTRACT The Columbia River Gorge is the Columbia River’s long-held yet evolving passage through the volcanic arc of the Cascade Range. The globally unique setting of a continental-scale river bisecting an active volcanic arc at the leading edge of a major plate boundary creates a remarkable setting where dynamic volcanic and tectonic processes interact with diverse and energetic fluvial processes. This three-day field trip explores several elements of the gorge and its remarkable geologic history—cast here as a contest between regional tectonic and volcanic processes building and displacing landscapes, and the relentless power of the Columbia River striving to maintain a smooth passage to the sea. This field trip generally follows the GSA guide published in GSA Field Guide 55 (available at https://pubs.geoscienceworld.org/gsa ): O’Connor, J.E., Wells, R.E., Bennett, S.E.K., Cannon, C.M., Staisch, L.M., Anderson, J.L., Pivarunas, A.F., Gordon, G.W., Blakely, R.J., Stelten, M.E., and Evarts, R.C., 2021, Arc versus river—The geology of the Columbia River Gorge, in Booth, A.M., and Grunder, A.L., eds., From Terranes to Terrains: Geologic Field Guides on the Construction and Destruction of the Pacific Northwest: Geological Society of America Field Guide 62, p. 131–186, https://doi.org/10.1130/2021.0062(05) . DEDICATION Dedicated to Russell C. Evarts (7 April 1947–11 July 2017) and his contributions to Pacific Northwest geology. Russ Evarts devoted most of his 30-year career with the U.S. Geological Survey to geologic mapping of Oregon and Washington. His thorough geologic mapping of the near-vertical terrain of the western Columbia River Gorge underpins much of what is reported in this guide and continues to inspire our studies of the geology of the Pacific Northwest.

ABSTRACT The Columbia River Gorge is the Columbia River’s long-held yet evolving passage through the volcanic arc of the Cascade Range. The globally unique setting of a continental-scale river bisecting an active volcanic arc at the leading edge of a major plate boundary creates a remarkable setting where dynamic volcanic and tectonic processes interact with diverse and energetic fluvial processes. This three-day field trip explores several elements of the gorge and its remarkable geologic history—cast here as a contest between regional tectonic and volcanic processes building and displacing landscapes, and the relentless power of the Columbia River striving to maintain a smooth passage to the sea. DEDICATION Dedicated to Russell C. Evarts (7 April 1947–11 July 2017) and his contributions to Pacific Northwest geology. Russ Evarts devoted most of his 30-year career with the U.S. Geological Survey to geologic mapping of Oregon and Washington. His thorough geologic mapping of the near-vertical terrain of the western Columbia River Gorge underpins much of what is reported in this guide and continues to inspire our studies of the geology of the Pacific Northwest.

The middle Miocene Columbia River Basalt Group is the youngest and smallest continental flood basalt province on Earth, covering over 210,000 km 2 of mainly Oregon, Washington, and Idaho, with an estimated basalt volume of ~210,000 km 3 . A well-established regional stratigraphic framework built upon six formations contains numerous flows and groups of flows that can be readily distinguished by their physical and compositional characteristics, thus producing mappable units, the areal extent and volume of which can be calculated and correlated with their respective feeder dikes. The distinct physical features that help to define these units originated during their emplacement and solidification, as the result of variations in cooling rates, degassing, thermal contraction, and interaction with their paleoenvironment. Columbia River Basalt Group flows can be subdivided into two basic flow geometries. Sheet flows dominate the basalt pile, but the earliest flows comprising the Steens Basalt and some of the Saddle Mountains Basalt flows are compound flows with elongated bodies composed of numerous, local, discontinuous, and relatively thin lobes of basalt lava. The internal physical characteristics of the voluminous sheet flows are recognizable throughout their extent, thus allowing mechanistic models to be developed for their emplacement. The emplacement and distribution of individual Columbia River Basalt Group flows resulted from the interplay among the regional structure, contemporaneous deformation, eruption rate, preexisting topography, and the development of paleodrainage systems. These processes and their associated erosional and structural features also influenced the nature of late Neogene sedimentation during and after the Columbia River Basalt Group eruptions. In this paper, we describe and revise the stratigraphic framework of the province, provide current estimates on the areal extent and volume of the flows, and summarize their physical features and compositional characteristics.

The Columbia River flood basalt province, United States, is likely the most well-studied, radiometrically well-dated large igneous province on Earth. Compared with older, more-altered basalt in flood basalt provinces elsewhere, the Columbia River Basalt Group presents an opportunity for precise, accurate ages, and the opportunity to study relationships of volcanism with climatic excursions. We critically assess the available 40 Ar/ 39 Ar data for the Columbia River Basalt Group, along with K-Ar data, to establish an up-to-date picture of the timing of emplacement of the major formations that compose the lava stratigraphy. Combining robust Ar-Ar data with field constraints and paleomagnetic information leads to the following recommendations for the age of emplacement of the constituent formations: Steens Basalt, ca. 16.9 to ca. 16.6 Ma; Imnaha Basalt, ca. 16.7 to ca. 16 Ma; Grande Ronde Basalt, ca. 16 Ma to ca. 15.6 Ma; Wanapum Basalt, ca. 15.6 to ca. 15 Ma; and Saddle Mountains Basalt from ca. 15 Ma to ca. 6 Ma. The results underline the previously held observation that Columbia River Basalt activity was dominated by a brief, voluminous pulse of lava production during Grande Ronde Basalt emplacement. Under scrutiny, the data highlight areas of complexity and uncertainty in the timing of eruption phases, and demonstrate that even here in the youngest large igneous province, argon dating cannot resolve intervals and durations of eruptions.

The radiometric dating evidence for the timing and duration of volcanism for the Steens through Wanapum Basalt of the Columbia River Basalt Group is critically reviewed here. K-Ar dates generally underestimate the age of crystallization, though one important exception is detected, where excess argon led to dates that were too old. The 40 Ar/ 39 Ar results on whole-rock basalts from 1980 through 2010 are examined for statistical validity of plateau sections, as well as alteration state of the material dated. In most instances, listed ages are shown to be invalid. The 40 Ar/ 39 Ar total gas (fusion) ages are, in general, not accurate estimates of the time of formation of these rocks. The 40 Ar/ 39 Ar ages on plagioclase separates from basalts yield good estimates of the extrusion age of the lavas. New 40 Ar/ 39 Ar ages on whole-rock basalts are presented that are in good agreement with the plagioclase ages. Various forms of the geomagnetic polarity time scale for mid-Miocene time are examined, along with the ages of lavas and their magnetic polarity. The main sections of the Columbia River Basalt Group (Imnaha through Wanapum Basalt) were formed in ~0.5 m.y. between 16.3 and 15.8 Ma. Steens Basalt extrusion occurred about ~0.1 m.y. before the Imnaha Basalt and appears to have been a precursor to the more voluminous volcanism noted in the Columbia River Basalt Group.

The Steens Formation, or Steens Basalt, is formally recognized as the oldest lithostratigraphic unit of the Columbia River Basalt Group, with an estimated areal extent and volume of 53,000 km 2 and 31,800 km 3 , respectively. We integrate petrochemical, paleomagnetic, and 40 Ar- 39 Ar age data on 13 collected sections to help evaluate stratigraphic and petrogenetic relationships through the Steens succession. We estimate that the overall duration of Steens Basalt volcanism from lingering eruptions could be as much as 300,000 yr, centered at ca. 16.7 Ma, but that the far greater volume erupted in <50,000 yr at an effusion rate ~0.67 km 3 /yr. Lava flows of primitive, homogeneous tholeiite initially erupted over a wide expanse of eastern Oregon during a reversed polarity interval (R 0 ). Later eruptions became more focused at the presumed shield volcano at Steens Mountain, where dikes exploited a NNE-trending zone of crustal weakness related to the northeast extension of the mid-Cretaceous western Nevada shear zone. The Steens Mountain shield volcano generated increasingly more diverse flows of tholeiite, alkali basalt, and basaltic trachyandesite that erupted during a geomagnetic polarity transition culminating in upper flows of normal polarity (N 0 ). The Steens sequence is dominated by compound flows (~10–50 m thick) produced by the rapid eruption of thin (<2 m) pahoehoe flow lobes. Analysis of these stacked sequences in the Catlow Peak section reveals periodic recharge of the magma chamber and ubiquitous fractional crystallization of plagioclase and olivine in each compound flow, accommodated by plagioclase accumulation and selective crustal contamination. The overall flood basalt stratigraphy records a rapid and progressive change in eruption style, from the early, near-continuous eruptions of small-volume Steens Basalt flows to later, more episodic eruptions of large-volume, tabular flows comprising the Imnaha, Grande Ronde, and Picture Gorge Basalts.

We examined Grande Ronde Basalt lava flows from surface sections and boreholes throughout Washington, Oregon, and Idaho to determine chemical and physical properties that would allow the recognition and mapping of these flows on a regional scale. We estimate there are ~100 flows covering nearly 170,000 km 2 , with a total volume of ~150,400 km 3 , that were erupted over four polarity intervals (reverse 1, normal 1, reverse 2, and normal 2) in ~0.42 m.y. These flows are the largest known on Earth, with individual volumes ranging from ~100 km 3 to greater than 10,000 km 3 . Although all known Grande Ronde Basalt flows erupted in the eastern part of the Columbia River flood basalt province, the thickest and most complete sections (>3 km) occur in the central Columbia Basin. From the center of the basin, the number of flows decreases outward, resulting in a nearly complete stratigraphy in the interior and an abbreviated and variable stratigraphy along the margins. The areal extent of many flows suggests that the Chief Joseph dike swarm greatly expanded after Imnaha Basalt time, and now many dikes are buried beneath younger flows in the eastern part of the province. The Grande Ronde Basalt has a relatively uniform lithology with only a few distinctive flows. However, when compositions are combined with paleomagnetic polarity, lithology, and stratigraphic position, the Grande Ronde Basalt can be subdivided into at least 25 mappable units. Grande Ronde Basalt flows are siliceous, with typically SiO 2 >54 wt%, MgO contents ranging from ~2.5 to 6.5 wt%, and TiO 2 ranging from 1.6 to 2.8 wt%, with an enrichment in iron and incompatible elements relative to mid-ocean-ridge basalt. Although most Grande Ronde Basalt flows have homogeneous compositions, some are heterogeneous. Dikes that fed the heterogeneous flows show that the first composition erupted was not typical of the flow, but as the eruption progressed, the compositions gradually evolved to the bulk composition of flow. The average effusion rate was ~0.3 km 3 /yr, with basalt volume peaking during the R2 polarity with the eruption of the Wapshilla Ridge Member. Eruption and emplacement rates for the flows are controversial, but available data collected from the field suggest that many of the flows could have been emplaced in a few years to perhaps a decade.

The Frenchman Springs Member is the most voluminous member of the Wanapum Basalt, Columbia River Basalt Group. We have revised the distribution maps and estimates of the areas and volumes for the units of the Frenchman Springs based on mapping and fieldwork that have been conducted since the previous compilation in 1989. The revised estimates indicate that many of the Frenchman Springs flows are significantly larger than previously believed. The total area underlain by Frenchman Springs flows is now estimated to be ~72,595 km 2 ; the revised volume is estimated to be ~7628 km 3 . Based on the new data, the Basalts of Palouse Falls, Ginkgo, Sand Hollow, and Sentinel Gap are significantly more voluminous than previously indicated, while the Basalt of Silver Falls is significantly smaller. We have also reduced the number of units within the Frenchman Springs Member to five. The Basalt of Lyons Ferry has been combined with the Basalt of Sentinel Gap based on the similarity of their geochemistries and paleomagnetism and the lack of an identified dike/vent system. Geochemical variations in Frenchman Springs Member flows (as well as those in the Roza and Priest Rapids Members, Wanapum Basalt) are consistent with being produced by open magma system processes.

The lack of field evidence for significant crustal stretching during the main phase of flood basalt eruption rules out a passive-mantle origin for the Columbia River Basalt Group. The short duration (1.1 m.y.) of the main-phase lavas, their large volume (~195,500 km 3 ), and their high eruption rate (~0.178 km 3 /yr) in an area of minimal extension (<<1%) are better explained by an active-mantle upwelling origin. Such a scenario could be satisfied by the arrival of a lower-mantle plume, or by the ascent of upper mantle through a slab gap or around the truncated edge of the Farallon plate. These competing ideas have been based on the interpretation of seismic data beneath the Yellowstone region that appear to resolve either a sheet of mantle upwelling extending downward into the mantle transition zone, or an irregular corridor of upwelling extending through the transition zone and into the lower mantle to ~900–1000 km depth. A lower-mantle origin is supported by the seismically resolved upward deflection in the ~660 km discontinuity at the base of the transition zone, which in turn is consistent with a lower-mantle chemical signature reflected in high 3 He/ 4 He ratios found in derivative basalts from both the Snake River Plain and the Columbia River flood basalt province. Mid-Miocene arrival of the Yellowstone plume was centered beneath the Oregon-Nevada border region, where there are direct spatial and temporal connections among the initiation of flood basalt volcanism at ca. 16.7 Ma, the initiation of rhyolitic volcanism at the western end of the Snake River Plain hotspot track at ca. 16.5 Ma, and the initiation of a major period of uplift and crustal extension in the northern Basin and Range between ca. 17 and 16 Ma. Rapid uplift and the propagation of volcanism away from this region are consistent with models of plume emplacement. The inability of seismic studies to image a mantle-plume fabric of radiating flow is due to its destruction by corner flow, which is evident in the contemporary east-west flow fabric derived from shear-wave splitting measurements.