Columbia River Basalt Group,
Washington, Oregon, and Idaho

Columbia River Basalt Group

The Columbia River Basalt Group (CRBG) is the youngest and most studied flood basalt. The province underlain by the basalt is loosely termed the Columbia Plateau. Such an overall designation is a misnomer, however, for the basalt has been sharply folded and broadly warped, so that its top varies in elevation from slightly below sea level in the Pasco Basin to more than 2.5 kilometers above sea level in the Wallowa Mountains of northeast Oregon. ...

Figure 1: Pacific Northwest Volcanics showing the Cascade Range and Columbia Plateau, generalized subdivision of Cascade Range from Hammond (1979). Distribution of Columbia River Basalt Group from Tolan et.al. (in press). -- Modified from: Swanson, et.al., 1989, American Geophysical Union Field Trip Guidebook T106, p.2

Stratigraphy and Age

The group is formally divided into five formations, (Figure 15), which in turn are broken into formal and informal members (Swanson and others, 1979a; Camp, 1981; Beeson et.al., 1985; Reidel et.al., in press; Bailey, in press). ...

Figure 15: Stratigraphic subdivision of Columbia River Basalt Group. -- Modified from: Swanson, et.al., 1989, American Geophysical Union Field Trip Guidebook T106, p.21

Chemical Composition

-- (Not Online) --

Vent Systems

Linear vent systems occur only in the eastern half of the province, except for feeders of the Picture Gorge (the Monument dike swarm) near the southern limit of the province (Waters, 1961; Swanson et.al., 1975, 1979a; Tolan et.al., in press). Some vent systems are longer than 150 kilometers, and all trend within a few degrees of due north, mostly north-northwest. The systems are correlated with specific stratigraphic units chiefly by the presence of dikes of appropriate chemical composition, petrography, magnetic polarity, and stratigraphic position. Hundreds of dikes have been identified, although many probably represent en echelon segments of one vent system (Waters, 1961; Taubeneck, 1970). Most dikes are known from Chief Joseph dike swarm in the tri-state area of Washington, Oregon, and Idaho (Taubeneck, 1970). Distribution patterns for some flows of the Grande Ronde Basalt suggest that their feeder dikes are hidden beneath younger flows north of the Chief Joseph swarm (Reidel et.al., in press). The dikes typically are a few meters wide, but some are wider than 20 meters. Composite dikes, consisting of basalt of two or more different compositions and hence ages, have not been found, but multiple dikes, with internal jointing patterns suggestive of two or more pulses during the same intrusive event, are common. Remnants of spatter cones and other near-vent features, similar to those of modern Kilauea, occur locally (Swanson et.al., 1975). The degree of vesicularity and breakage of pyroclasts in these deposits resembles that of modern basaltic tephra, so it is unlikely that the magma was unusually rich in gas.

Volumes and Eruption Rates

Flood-basalt provinces by definition contain flows of huge volume, on the order of 5-10 cubic kilometers or more; this is the major distinction between flood-basalt and plains-basalt provinces (Greeley, 1977), such as the Snake River Plain and Iceland, in which flows are generally much less than one cubic kilometer. Recent work by Tolan et.al., (1987,in press) suggest that about 300 great flows were erupted on the Columbia Plateau, with an average volume per flow of about 580 cubic kilometers. Numerous low-volume flows are probably present near vents, but the great flows form most of the province. Flows with volumes more than 100 cubic kilometers occur in all the formations, but most such flows were erupted during Grande Ronde time, when at least 110 major flows with volumes of 90 cubic kilometers to more than 5000 cubic kilometers were produced (Reidel et.al., in press; Tolan et.al., 1987, in press). Such a huge eruption, if from a shallow crustal source, would probably have resulted in recognizable subsidence along the trace of its fissure system. No such evidence has been found, so the inference on geologic grounds alone is that the eruptions tapped a deep reservoir. Viewed regionally, the overall saucer-like subsidence of the province might reflect withdrawal of magma, but its amplitude and diameter would demand a deep source. Petrologic evidence (Helz, 1978; Hooper, 1984; Wright et.al., in press) supports a deep storage reservoir, perhaps at the base of the lithosphere.

Model calculations (Shaw and Swanson, 1970), based on evidence that little cooling occurred during flowage of hundreds of kilometers, suggest that eruption and emplacement spanned only a few days for these huge outpourings. The flows evidently moved 5-15 kilometers per hour or faster (Shaw and Swanson, 1970) and advanced as sheet floods; no lava tubes have been found. The evidence for little cooling during transport is that flows quenched to glass when they entered water after traveling several hundred kilometers; the crystal content of the glass is no higher than that of chilled margins of feeder dikes. The effective viscosity of the lava was not unusually low; in fact, calculations of the viscosity based on the chemical composition for a range of reasonable water contents indicate that the lava was somewhat more viscous than modern Kilauea lava. According to Shaw and Swanson (1970), the high rate or eruption (about 1 cubic km/day/linear kilometer of fissure or higher, 3-4 orders of magnitude faster than rates of Hawaiian and Iceland eruptions (Swanson et.al., 1975)) combined with the huge volume of available magma enabled the flows to travel so far, a relation that Walker (1973) found on empirical grounds. Regional mapping (Swanson et.al., 1979b, 1980, 1981) indicates that the flows ponded against opposed slopes, including locally recognizable natural levees, and formed low-aspect-ratio (0.0002-0.0001) lava lakes generally 30-40 meters thick and 200-400 kilometers in diameter. The lakes cooled to ambient temperatures within a few years to a few tens of years (Long and Wood, 1986).

Average Magma Supply Rate

About 5.5% of the group was erupted during the first 0.5 m.y. of volcanism to form the Imnaha Basalt (Tolan et.al., in press). Activity peaked during 1.5-m.y. Grande Ronde time, when 87% of the group was erupted, 85% as the Grande Ronde Basalt and 2% as the Picture Gorge Basalt and other units. Activity waned thereafter, accounting for 6% of the group during Wanapum time, which lasted about 1 m.y., and 1.5% during prolonged Saddle Mountains time from about 14 to 6 Ma.

During peak activity the average interval between major eruptions was about 13,500 years, the average volume per major flow was about 1350 cubic kilometers, and the average magma supply rate was 0.1 cubic kilometer per year. The average supply rate is identical to that calculated for historical time at Kilauea (Swanson, 1972; Dzurisin et.al., 1984). On this basis, a larger heat source for Grande Ronde Basalt than for modern Kilauea is unnecessary. An important distinction between the two provinces is that magma "leaks" from Kilauea almost continuously, whereas magma for the CRBG, if produced at the Kilauea rate, must have been stored for thousands of years before eruption in order to account for the huge volumes of single flows. (Baksi (1988) believes the maximum rate of supply, in late Grande Ronde time, was 2-3 times that at Kilauea; this interpretation is based on his revision of the geomagnetic time scale and seems overly dependent on his model for the time scale.)

The difference in eruption style between the CRBG and Kilauea may result from the 40-kilometer-thick continental crust (Catchings and Mooney, 1988; Michaelson and Weaver, 1986) above the melting zone for the CRBG compared to the 12-kilometer-thick oceanic crust above the melting zone for Kilauea. The continental crust may have provided a density trap keeping magma from rising until the trap was broken tectonically. Petrologic evidence suggests that the basalt was derived from a complex, multisource zone in the mantle with little crystal fractionation or crustal contamination during rise to the surface (Hooper, 1982, 1984; Wright et.al., in press; Carlson, 1984; Carlson and Hart, in press; Hart and Carlson, 1987). This interpretation implies that the rise was rapid, consistent with the idea that fracturing, not a rising diapir, caused the magma to move to the surface.