The geological features of central Oregon are the result of plate tectonic activity that generate and deform the earth’s crust, and the geomorphological processes that modify its landscapes.  The spectacular Cascade Range is a volcanic arc generated by collision between the westward-moving North American plate and the eastward-advancing Farallon – Juan de Fuca plate.  The Farallon – Juan de Fuca plate is composed of oceanic crust and lithosphere which is subducting under the North American plate, and partially melting to form magma that is accumulating on the leading edge of the North American plate as a volcanic mountain range.  These mountains form a small part of a much longer subduction zone complex known as the “Ring of Fire” that wraps around the Pacific Ocean basin from South America to New Zeeland.  Much of Oregon is comprised of a succession of accreted terranes, micro-plates rafted along to the east and mashed onto the more ancient crystalline basement rocks of the North American craton.  Younger volcanic materials such as the Cascade volcanic arc and flood basalts of the Columbia River and Owyhee Plateaus are built upon these terranes.  These latter basaltic lava flows, as well as the bimodal mafic and felsic volcanics of eastern Oregon’s High Lava Plains may be related to the geologically recent impingement of a mantle plume beneath this region.  South-central Oregon also hosts significant block-faulted horst and graben structures related to Basin and Range crustal extension that developed in conjunction with the subduction of progressively younger, warmer, more buoyant oceanic plate, and concurrent impingement of the North American and Pacific plates along a northward expanding transform plate boundary.

These tectonically “constructed” landforms have concurrently undergone weathering and erosion by a host of geomorphic agents, chief among them, repeated glaciations that have scoured the western and High Cascades.  River valleys initially carved into the mountainous landscape were subsequently occupied by massive glaciers emanating from an extensive ice cap perched along much of the Cascade Crest to form numerous cirques, U-shaped valleys, and end moraine complexes.  These latter features of glaciation are nearly ubiquitous along both flanks of the High Cascades platform and record at least two major periods of glaciation.  Other geomorphic agents have been at work in central Oregon, including, of course large river systems like the McKenzie and the Deschutes, mass wasting, and lesser known phenomena like shoreline processes associated with the pluvial lakes of the Fort Rock-Christmas Valley area. 

The Regional Setting

The geology of central Oregon is extremely diverse.  Four distinct geological provinces merge in this region, including the High Cascade Volcanic Arc, the Blue-Ochoco Mountains, the High Lava Plains, and the northern Great Basin (Figure 1).  The geology of the Blue-Ochoco Mountains is only very briefly visited, while the geology of the other three provinces forms the principle focus of central Oregon Cascades section of my website.  These three areas contain features related to active tectonism, predominantly volcanic in nature, but also include faulting; and superimposed on these features are geomorphic and sedimentologic characteristics formed in response to climatic fluctuations of the geologically recent past.  The field trips presented in this section of my website are intended to highlight many of the more spectacular examples of these geological phenomena, self-guided field trips on which you will get to examine these geologic features firsthand, learn about the processes responsible for their formation, and explore their tectonic and climatic significance.

Figure 1.  The High Cascade Volcanic Arc, Blue-Ochoco Mountains, High Lava Plains, and Great Basin merge in central Oregon, providing considerable geological diversity over a relatively small region.

In general, the geology of the Cascades Volcanic Arc, High Lava Plains, and Great Basin provinces is closely related to interactions between the Pacific plate, Farallon-Juan de Fuca plate, and the North American plate during the last 35 million years (Priest, 1990; Taylor, 1990).  Figure 2 displays the current geometry of these plate interactions (the “modern” Farallon plate now represented by the Explorer – Gorda – Juan de Fuca plate remnants).  The High Cascades province represents the modern portion of the volcanic arc formed by subduction and partial melting of the Farallon (Juan de Fuca) plate under the North American plate at the Cascadia Subduction Zone.  The volcanic arc actually consists of two distinct parallel volcanic rock assemblages (Taylor, 1990) as shown in Figure 3: the older volcanic arc rocks of the Western Cascades which are dominantly composed of andesite, basaltic andesite, dacite, and associated pyroclastic material of intermediate composition; and the younger volcanic arc rocks of the High Cascades which are still dominantly composed of intermediate composition andesites and basaltic andesites, but with multiple isolated centers of more silicic dacite and rhyolite.

Figure 2.  A diagram describing the current interactions between the Pacific, Farallon (Juan de Fuca), and North American plates.

Figure 3.  The Cascades volcanic arc; consisting of two parallel belts with distinctive volcanic rock assemblages, the Eocene to Pliocene Western Cascades and the Quaternary High Cascades provinces.

The Western Cascades formed during two episodes of volcanism from about 35 – 18 Ma and about 14 – 8 Ma, separated by an interval of relative uplift and erosion with little volcanic activity (Priest, 1990).  Convergence rates were high during these time intervals, indicated by the formation of a fairly low and broad volcanic arc containing copious amounts of volcanic rocks produced by high rates of volcanism (as much as triple that of today’s rate).  Convergence rates between the Farallon (Juan de Fuca) plate and the North American plate decreased significantly between about 9 and 7 Ma, accompanied by a steepening of the convergence angle and a shortening of the convergent margin as the Pacific plate – North American plate transform boundary migrated northward (Priest, 1990).  The combined result of these factors was the eastward migration and narrowing of the older volcanic arc and formation of the High Cascades volcanic arc.  Subsequent growth of the High Cascades from about 7 Ma to the present was accompanied by extension and thinning of the overlying North American plate and formation of a major N-S trending graben, subsequently buried by the eruptive products of the High Cascades volcanoes, a corresponding increase in mafic volcanism (although some silicic centers remained active), and further uplift of the Western Cascades (Priest, 1990).

The High Cascade volcanic arc includes nearly 20 major composite (strato-) volcanoes, exemplified by the Three Sisters in the central Oregon Cascades (Figure 4), among a total of over 4,000 separate volcanic vents including numerous stratovolcanoes, shield volcanoes, lava domes, and cinder cones.  Although volcanism associated with the Cascade Range began about 37 million years ago, most of the present-day High Cascades volcanoes are less than 2,000,000 years old, and the highest peaks of the modern volcanic arc are less than 100,000 years old. Twelve stratovolcanoes in the High Cascades province are over 10,000 ft (3,000 m) in elevation, three of which (the North, Middle, and South Sisters) form a significant focal point of this guidebook.  Newberry Volcano, a shield volcano offset to the east of the main Cascade Crest and by volume one of the largest of the Cascade Range volcanoes, formed of over 108 mi³ (450 km³) of material, presents another important field trip locus.

Figure 4.  The modern High Cascades volcanic arc as represented by the exemplary composite volcanoes: Broken Top Volcano (leftmost) and the Three Sisters Volcanoes (to the right – oldest to the north, youngest to the south).

The Basin and Range region of the western United States is a region of extensional block faulting that includes the Great Basin province and bordering areas such as the High Lava Plains province of central Oregon.  Rifting in this region began between 17 and 14 Ma, when northward migration of the Pacific plate – North American plate transform boundary shortened the length of the Cascadia Subduction Zone enough to allow partial coupling between the Pacific and North American plates (Carlson and Hart, 1987; Christiansen and McKee, 1986).  Increased coupling caused extension and normal faulting in the North American plate perpendicular to its continental margin.  Since about 14 Ma, tectonic activity in the Great Basin has become progressively more concentrated toward the margins, and much of the extension has been taken up by a wide zone of northwest trending strike-slip faults and subsidiary normal faults (collectively forming an oblique-slip fault zone) along the High Lava Plains in Oregon and Idaho (Christiansen and McKee, 1986).  Figure 5 provides a regional view of many of the known faults within the northern Basin and Range and Cascade Range of Oregon, including several which have been active during the Holocene.

Figure 5.  Normal and strike-slip faults of the northern Basin and Range and Cascade Range in central Oregon; major geologic provinces are indicated by dashed green lines, while significant volcanoes are indicated by red triangles.

Between 17 and 14 Ma, the predominant volcanism in the Great Basin was basaltic, but somewhat more intermediate and relatively small in volume in the central Great Basin and progressively less intermediate and more basaltic and voluminous grading northward into the High Lava Plains (Christiansen and McKee, 1986).  Decompression melting within the underlying upper mantle caused the generation of basaltic magma which intruded into, and through, the overlying lithosphere and crust during extension.  Generation of magma and crustal thinning resulted in an increased regional heat flow and reduced crustal rigidity which caused regional uplift by thermal expansion and the production of rhyolitic magmas by localized partial melting of the lower crust.  Since approximately 14 Ma, basaltic and bimodal basaltic-rhyolitic suites have characterized volcanism throughout the Great Basin, but Christiansen and McKee (1986) suggest that eruptions have occurred in successively narrower zones near its margins (Figure 6).  This may be a direct response to the increasing concentration of extension and normal faulting toward these same marginal areas.


Figure 6.  Two types of spatial and temporal trends associated with Teritary and Quaternary High Lava Plains volcanism.  Flood-basalt volcanism from 16.6 to 15.0 Ma began from a source area in southeastern Oregon and spread rapidly to the north, northeast, and south (A); while post-15 Ma, more silicic, bi-modal volcanism occurred along two age-progressive trends to the northeast along southern Idaho’s Snake River Plain, and to the west across the High Lava Plains of eastern Oregon (B).

The most striking feature of the High Lava Plains is the west-northwest trending en echelon oblique-slip faults of the Brothers Fault Zone (Figure 5) which delimit the northern edge of the well-developed N-S trending Basin and Range extensional structures at the northwestern margin of the Great Basin (Lawrence, 1976).  Some of the larger north- and northeast-trending normal faults of the Basin and Range change to a northwest trend and appear to blend into the oblique-slip faults of the Brothers Fault Zone.  Many normal faults can be easily observed on the landscape, and Figure 7 provides a topographic map view of a small segment normal faults associated with the Brothers Fault Zone at Horse Ridge, near Brothers, OR (Figure 5).  Eruptive centers for both basaltic and rhyolitic volcanic rocks are concentrated along this zone of faults (Jordan et al., 2004).  Middle Miocene basalt flows containing distinct phenocrysts of plagioclase and minor andesite flows, commonly known as the Steens Basalts, form much of the older rocks of the High Lava Plains (Walker, 1981).  These flows are overlain by tuffaceous sedimentary rocks and widespread sheets of ash-flow tuff dated between 9 and 6.5 Ma (Walker, 1981), the most significant of which was the Rattlesnake Tuff (Stimac and Weldon, 1992; Streck et al., 1999).  The Harney Basin forms a structural depression at the eastern end of the High Lava Plains overlying the original magma chambers that provided the eruptive source for these tephra units.  Younger late Miocene to Pleistocene basalt flows bury these older rocks at the western end of the High Lava Plains.  Silicic volcanic rocks represented by rhyolite domes and related ash flows in the High Lava Plains display a systematic progression in age (Walker, 1974; Jordan et al., 2004), becoming younger toward the west, with the most recent period of rhyolitic volcanism occurring at Newberry volcano during the Holocene. 

Figure 7.  A topographic map view of a small segment of normal faults associated with the Brothers Fault Zone at Horse Ridge, near Brothers, OR.

More recent research suggests that the bimodal volcanism of Oregon’s High Lava Plains is associated with the impingement of a mantle plume on the underlying lithospheric plate beginning about 16 million years ago (Kistler & Peterman,1978; Hart et al., 1984; Carlson and Hart, 1987; Draper, 1991; Leeman et al., 1992; Ferns et al., 1993a and b; Ernst et al., 2001 and 2003; Christiansen et al., 2002; Glen and Ponce, 2002; Hooper et al., 2002; Camp et al., 2003; Camp and Ross, 2004; Jordan et al., 2004; and Coble and Mahood, 2008).  Two types of spatial and temporal trends associated with this volcanism have been recognized in the Columbia Plateau-Oregon Plateau-Snake River Plain-Northern Nevada Rift magmatic system (Camp and Ross, 2004; Jordan, 2004).  Initial eruptions began in southeastern Oregon and spread rapidly from the source area along three primary trends to the north, northeast, and south associated with the main eruptive phase of flood-basalt volcanism that occurred from 16.6 to 15.0 Ma that generated the Columbia River Basalts and Steen’s Basalts (Figure 6).  Eruption of more silicic magmas in northwestern Nevada and southeastern Oregon between 16.5 and 15.5 Ma (Coble and Mahood, 2008) initially overlapped basaltic volcanism, but then spread along two age-progressive trends (Camp and Ross, 2004): one to the northeast along southern Idaho’s Snake River Plain, and the other to the west across the High Lava Plains of eastern Oregon (Figure 6).  The post-15 Ma volcanism on the High Lava Plains was characterized by distinctly bimodal volcanism: widespread but sporadic eruptions of low-volume, high-alumina, olivine-rich tholeiitic basalts (Camp and Ross, 2004); and the eruption of rhyolitic lavas in more than 60 domes and dome complexes, three major ash flow tuffs (Prater Creek, Devine Canyon, and Rattlesnake), and several minor tuffs (Jordan et al., 2004).  As shown in Figure 6, late Tertiary and Quaternary eruptions were generally time-transgressive to the east and west, although the trend is moredistinctive with silicic volcanism than it is with mafic volcanism  (Hart et al., 1984; Ferns et al., 1993a; Christiansen et al., 2002; Hooper et al., 2002; and Camp et al., 2003).

Camp and Ross (2004) believe that these spatial and temporal trends in volcanism represent a two-stage spreading model for the Yellowstone mantle plume (Figure 8). The model suggests that impingement of a mantle plume head in northwestern Nevada and southeastern Oregon resulted in rapid spreading of mafic volcanic activity in all directions, but preferentially to the northwest beneath a relatively thin, weak lithosphere comprised of accreted oceanic terranes (under eastern Oregon).  Spreading to the south, beneath thicker transitional to continental lithosphere produced shallow intrusions and a much smaller volume of lava along the southward propagating Northern Nevada Rift System.  Shearing of the plume head against the westward-advancing, thick cratonic margin (under Idaho) after 15 Ma allowed the hotter plume tail to generate a hotspot track through the overriding craton to the east.  Once sheared and decapitated from its feeding plume tail, the plume head continued to spread beneath the lithosphere, advancing westward at a much slower rate by asthenospheric drag and by counterflow above the descending Farallon – Juan de Fuca plate, thus generating a westward migration of bimodal basaltic and rhyolitic magmas at a rate consistent with mantle convection.


Figure 8.  A two-stage spreading model for the Yellowstone mantle plume (modified from Camp and Ross, 2004) which explains the trends in bi-modal volcanism observed in southeastern Oregon relevant to this guidebook.

A spectacular record of late Quaternary climatic fluctuations is superimposed on the volcanic and tectonic features of central Oregon expressed as unique sedimentary deposits and geomorphic landforms.  The High Cascades and portions of the Western Cascades were inundated by broad sheets of coalescent valley glaciers as recently as about 18,000 years ago (Crandell, 1965) (Figure 9).  Evidence for multiple periods of glaciation has been reported for the upper watershed of the Metolius River (Scott, 1977 and Bevis et al., 2011), the headwaters of the Deschutes River and the upper Tumalo Creek drainage (Scott et al., 1990 and Bevis et al, 2013), the upper McKenzie River watershed (Bevis et al., 2008), and the Whychus Creek watershed (Bevis et al., 2011); each drainage is highlighted in Figure 9.  Figure 10 presents the limit of ice extent in the McKenzie River (Figure 10a), Metolius River (Figure 10b), and the Whychus Creek (Figure 10c) drainages mapped by this author (Bevis et al., 2008, 2011, and 2013).  Figure 11 displays reconstructions of the former glacial systems occupying these drainages during the LGM between about 24,000-18,000 years ago: the McKenzie River watershed (Figure 11a), the Metolius River watershed (Figure 11b), and the Whychus Creek watershed (Figure 11c).  The McKenzie and Metolius River drainage basins preserve deposits associated with two glaciations: the younger is correlated with oxygen isotope stage 2, the late Pleistocene last glacial maximum (LGM), named the Cabot Creek glaciation by Scott (1977); the older is correlated with oxygen isotope stage 6, a late middle Pleistocene glacial episode, named the Jack Creek glaciation of Scott (1977).  The upper Deschutes River and the Tumalo Creek and Whychus Creek drainages appear to only contain deposits of the younger glaciation, although this is likely a result of older glacial deposits being overwhelmed and buried by a more extensive ice mass during the younger glacial period.  An interesting interplay between volcanism and glaciation may be the culprit here.  South Sister Volcano began erupting less than 100,000 years ago, post-Jack Creek glaciation, and its presence added a significant amount of high-elevation real-estate as a source area for growth of subsequent glaciers during the late Pleistocene LGM (Cabot Creek glaciation).  This author believes that South Sister’s development allowed a more extensive glacial system to occupy the upper Deschutes River valley and the drainages of Tumalo Creek and Whychus Creek during the Cabot Creek glaciation, a system that advanced over the earlier formed, less extensive Jack Creek deposits (Bevis et al., 2011 and 2013).

Figure 9.  The extent of alpine glaciation in the Oregon Cascades during the late Pleistocene glacial maximum (modified from Crandell, 1965).

Figure 10.  Map showing the extent of alpine glaciation in the central Oregon Cascades during the late Pleistocene LGM (Cabot Creek) and late middle Pleistocene pre-LGM (Jack Creek) glaciations of the upper McKenzie River (A), Metolius River (B), and Whychus Creek (C) watersheds.

Figure 11.  Maps presenting reconstructions of the alpine glacial systems associated with the late Pleistocene LGM (Cabot Creek glaciation) that occupied the upper McKenzie River (A), Metolius River (B), and Whychus Creek (C) watersheds.

Several of the taller peaks in or bordering the northern Great Basin of south-central Oregon also exhibit signs of glacial activity, including Newberry Volcano, described on this website (Osborn and Bevis, 2001).  Many of the structural basins in south-central Oregon contained large “pluvial” lakes during these glacial episodes (Figure 12), associated with cooler temperature (and more importantly, less evaporation) and possibly higher precipitation. Distinct shoreline features and lacustrine sediments from several lake levels are recognized in the Alvord, Chewaucan, and Fort Rock basins (visited in this guidebook), among others (Allison, 1979 and 1982; Freidel and McDowell, 1992; Carter et al., 2006; and Jensen, 2006).  Some of these basins may have contained lakes for much of the Quaternary.  They may contain detailed sedimentologic records of past climatic fluctuations and Cascadia volcanic eruptions such as near the Ana River in the Chewaucan basin (Weldon et. al., 1992).  The Fork Rock basin contains Cascadia volcanic eruptions such as near the Ana River in the Chewaucan basin (Weldon et. al., 1992).  The Fork Rock basin contains evidence for a long period of occupation by water indicated by the presence of numerous maars of Quaternary age (Peterson and Groh, 1961, 1963; Heiken, 1971).  Maars are explosive eruption craters and rings of tuffaceous pyroclastic material often found associated with lacustrine deposits, suggesting violent volcanic eruptions through saturated sediments.  Several unique geomorphic processes are currently active in the basins of south-central Oregon, such as the playa-dune complexes in the Alkali basin and mass wasting in the Summer Lake basin (Weldon et. al., 1992).

Figure 12.  A map depicting the late Pleistocene “pluvial” lake basins of south-central Oregon and adjacent areas of California and Nevada; present-day lakes and population centers are shown for reference.