top of page

Clastic Dikes: Sand Dikes Intruding Bedrock

Clastic Dikes in Basalt Bedrock

Sheeted dikes with sandy fills intrude downward into Miocene Columbia River Basalt in the Walla Walla Valley, Pasco Basin, Umatilla Basin, Lewiston Basin, and Yakima Valley of Washington, Oregon, and Idah. These same dikes intrude Neogene basin-fill deposits and Pleistocene sediments that predate Missoula Floods (pre-Late Wisconsin, >28 ka). The sheeted, per descendum dikes were not produced by liquefaction. Rather they are injectites (Cooley, 2020) formed by hydraulic fracture trigger by large overriding floods during the Ice Ages.

Pleistocene dikes in Miocene basalt. Sheeted dikes filled with unconsolidated silt-sand dike intrude several meters into basalt at several locations near Lewiston, ID. Note the outer walls lack silt skins. In basalt, the outer walls of dikes are missing. The impermeable rock prevents outer walls from forming because pore water exiting the dike (leakoff) cannot diffuse into rock like it does into sediment. Silt partitions between internal fill bands are present just like in other dikes. Once the first fill band is in place, the next-injected band can dewater into it, forming a skin at the contact between the two fills.

Weaver pit. Sheeted dikes occur with opal and gouge in basalt in southern Walla Walla Valley.

McCredie. A sheeted silt-sand dike descends from Touchet Beds through pillow basalt at Alder Ridge, WA.

Prosser. A sheeted sand dike sourced in Touchet Beds above intrudes basalt at Prosser.

Dikes in Neogene Basin-fill Sediments

Dikes intrude Miocene sedimentary interbeds (Ellensburg Fm) and the Pliocene Ringold Fm at Snipes Mountain, White Bluffs, Tule Rd, Foster Creek, Frenchman Hills, Alder Ridge, and other locations.

Emerald Road. Sheeted dikes filled with Touchet Bed sediment intrude oxidized Miocene sandstones (Ellensburg Fm) at Snipes Mountain near Granger, WA.

Tule Road. Clastic dikes descend from Touchet Beds into quartzite-bearing gravels (Snipes Mountain Conglomerate).

Foster Creek. Sheeted clastic dikes intrude fluvial sandstone of the Miocene Ellensburg Fm composed of gruss transported from the nearby Colville batholith. The surface above was swept by Ice Age floods and overridden by glacial ice of the Okanogan Lobe.

Leakoff halo. Evidence of how the dikes dewater can be seen clearly from time to time. A halo of lightly-cemented sand surrounds the dike, indicating the surrounds the dike was unconsolidated, dry, and ice-free with the dike intruded. "Leakoff halos" (a new term introduced here) do not form where dikes intrude impermeable bedrock, like basalt.

Dikes in the Clearwater Gravels

Near Lewiston, sheeted dikes sourced in Missoula flood rhythmites penetrate late Pliocene Clearwater Gravel (2-6 Ma), which overlie Lower Monumental Member basalt, contain non-basalt lithologies sourced to the east, and appear equivalent to the middle Ringold Fm (Taylor Flat Member) in the Pasco Basin (Webster et al., 2001).

Webster et al. (1982) observed,

Clastic dikes filled with sediment which can be followed upward into the Touchet Beds are present in some exposures of the Clearwater Gravel. Generally these are restricted to the upper gravel unit, but in the Northwest Aggregate quarry, east of the new Lewiston grade, they extend through the sand unit into the lower gravel unit.

Othberg et al. (2003) adds,

The clastic dikes are common features in the [megaflood] deposits and may follow coarser, sand and gravel facies. Where rhythmites mantle Clearwater Gravel and basalt, the clastic dikes cut deep into the gravel and follow joints and normal faults down into the basalt.

Dikes in Pleistocene Sediments Older than the Missoula Floods

Sheeted dikes with truncated tops also intrude Pleistocene sediments that pre-date Missoula flooding at Hwy 397, Hwy 124, Yakima Bluffs, Colville River, Rulo, Touchet Valley, Dodd Rd, and other locations. The dikes intrude Bonneville Flood deposits at Lewiston.

Arlington, OR. Late Wisconsin age sheeted sand dike intrudes an older soil surface with a polygonal crack pattern.

Ringold Road at White Bluffs. Cemented dikes in the upper Ringold Fm (lake beds) are truncated at their tops by a flood-deposited gravel (pre-Late Wisconsin) composed almost entirely of calcrete rip-ups. Right: Road crews known their geology.

Rulo. Several generations of sheeted dikes that show age (oxidized, clay-rich, cemented) cut loess paleosols and other sediments at the Rulo site in northern Walla Walla Valley. Dikes are truncated at their tops against unconformities. Elevation is below 366m. See Bader et al., (2016).

Plucker. Sheeted and lightly cemented dikes (pre-Late Wisconsin) are truncated by Touchet Beds at this roadcut in the Touchet Valley.

Dike-like Structures

Sediment-filled joints in basalt, like those at Reith Road, Threemile Canyon, Sagebrush Flat, and Frenchman Hills, are not "Touchet-type" clastic dikes. Likewise, rubbly gouge zones in basalt, like those at Finley Quarry described by numerous authors (Jones and Deacon, 1966; Jahns, 1967; Brown, 1968; Sherrod et al., 2016; Coppersmith et al., 2014), are not dikes.

Bingham et al. (1970) explains,

The large "clastic dikes" in the Finley Quarry reported by Jones and Deacon (1966, plate 13, photograph 5), Jahns (1967, fig. 11) and R.E. Brown (1968, p. 26) are not Touchet-age clastic dikes, but instead are older sedimentary fillings of joints and fissures that opened up as the blocks tipped slightly northward, probably toward an unsupported fault scarp.

Finley Quarry. Dike-like structures are actually rubble-filled tensile cracks. Parts of the basalt cliff toppled north. The cracks are incipient topples. For more info on Finley Quarry, see my other post:

Houghton Road. A vertical gouge- and sediment-filled fracture in basalt at a county quarry in the Yakima Valley north of Sunnyside (Rattlesnake Hills).

Sediment-filled joints. Silt-filled joints in Columbia River Basalt at Threemile Canyon, Umatilla Basin. The open cracks accumulated silt, both windblown and washed-in. A light calcium carbonate overprint apparent on some.

Phreatic fills. Fracture fills formed when wet sediment was overridden by molten basalt. Hwy 397 west of Finley, WA.

Melted diatomite? Numerous Miocene-age dikes filled with remobilized diatomite cut a Miocene basalt pillow-palagonite complex at exposures on the south flank of Frenchman Hills. For more info on the geology of Frenchman Hills sites, see my other post:

Sagebrush Flat. Muddy sediment fills a fracture in Grande Ronde basalt formed when the lava flow invaded a wetland. Beezley Hills west of Soap Lake, WA.

Gouge at the County Pit. Gouge zone in basalt exposed in an active quarry at intersection of Rd A SE and Rd 10.5, Frenchman Hills.

Hanson's wedges. Rubble-filled wedge in Glacial Lake Missoula varve beds in western Montana. Larry Smith of Montana Tech also has written on these wedges. Michelle Hanson photo.

Dikes Hosted Entirely Within the Ringold Fm

Dozens of thin, cemented, unsheeted to crudely-sheeted dikes occur in the Pliocene Ringold Fm at numerous locations both above and below the level of Ice Age flooding. Unsheeted dikes in Ringold at White Bluffs, Smyrna Bench, Watt Lane, and Othello are sourced entirely within the Ringold and are commonly associated with a hard, white claystone that occurs with paleosols and exhibits blocky parting.

Smyrna Bench. Small, unsheeted dikes hosted entirely within the Ringold Fm are occasionally found. These dikes descend a short distance from the light-colored, sandy layer into the reddish coarse sand-gravel unit below.

Columbia River Road. White dikes in the Ringold convey irrigation water downward from the top of the bluff, several tens of meters above. The white color is a mineral overprint; they are cemented with CaCO3. This dike might actually begin in flood deposits.

Hard white claystone. Hundreds of small, single-fill dikes intrude this meter-thick hard, white claystone near Othello. This visually conspicuous, dike-bearing unit resides within a paleosol interval in the Savage Island Member (uppermost Ringold) and is commonly spotted in the northern Pasco Basin.

Clastic Dikes Beyond the Ice Age Floodway

Dikes in Miocene and Pliocene strata outside the Ice Age floodway are few in number and differ fundamentally in size, fills, and cementation from sheeted dikes within it. Most appear to be incidental structures formed prior to the Pleistocene due to local factors. A particularly conspicuous unsheeted mudstone dike cuts mixed Ellensburg sediments at the Ellensburg Rodeo Grounds. The dike is sourced entirely within the Ellensburg Fm and appears to be the only one in the large, well-visited exposure near Central Washington University. It is intensely fractured, a result of its relative brittleness.

Craig's Hill dike. An unsheeted mudstone dike cuts Ellensburg Fm at the Ellensburg Rodeo Grounds.

Craig's Hill. The clastic dike at Craig's Hill, Ellensburg sketched by Mabry (2000, p. 24). Site stratigraphy is more complicated than shown here. For more info on the Craig's Hill site, check out my other post:

Sedimentary Dikes Intruding Bedrock in Other Regions

Sedimentary dikes that intrude older bedrock are not exclusive to megaflood regions. Reports of lithified sandstone dikes intruding fractured bedrock are numerous (Walton and O'Sullivan, 1950; Birman, 1952; Harms, 1965; Stuart and Furnes, 1976; Roth, 1977; Jolly and Lonergan, 2002).

Sheeted Dikes Analogous to Touchet-type Dikes

Likewise, vertically sheeted, per descendum dikes nearly identical to "Touchet-type" dikes are reported in non-flood settings where rapid sedimentation, overloading, and hydrofracture occurred. Sheeted dikes are found in subglacial deposits (Sharp, 1968; Mangerud and Skreden, 1972; Broster, 1991; Larson and Mangerud, 1992; LeHeron and Etienne, 2005; Sagredo et al., 2011; Crossen, 2014; Talbot and Von Brunn, 1989; Phillips et al., 2013, Ravier et al., 2015), lahar deposits (Herriott, 2014), and beneath debris flows (example from Black Dragon Canyon, UT show below). There appear to be clear ties between sheeted dikes, hydraulic fracture and downward injection (i.e., injectites). None appear related to liquefaction.

Alaska Range. Sheeted dikes in lahar deposits at Mount Spurr, Alaska (Herriott, 2014). The arrows are his. They point to silt partitions between fill bands. Photo used by permission.

Sweden. Sheeted dike in glacial sediment in Sweden. Photo from Larsen and Mangerud (1992).

Wrangell-St. Elias National Park. Sheeted dike in glacial sediment. Bering Glacier, Alaska. Photo by Crossen (2014).

Poland. Sheeted dike in tensile fractures that opened repeatedly in bedrock. The authors attribute this dike to extension at the bedrock at the crest of a fold. The sheeting bears a striking resemblance to Touchet-type dikes and indicates repeated incremental opening. Haluszczak (2007) photo.

Puget Sound. Sheeted dike in mixed glaciomarine sediments at Whidbey Island, WA (Puget Lobe). My photo taken 2011.

San Rafael Swell. Dikes descend from base of an overriding debris flow at Black Dragon Canyon, UT. My photo taken c. 2009.

British Columbia. Injection dike (subglacial) descends into sandy sediment below Broster and Clague (1987).

Iceland. Sheeted dikes in glacial deposits by Ravier et al. (2015, Fig. 5c,d).

East side of the Sierras. Clastic dikes descend from gravel bed into till at Big Pumice Cut, CA (Sharp, 1968).

Iceland. Dikes sourced in a sandy diamicton (subglacial till) intruded downwards beneath a glacier during ice advance. Ice loaded the sediment resulting in dikes filled with laminated sediment where laminae parallel dike margins (Le Heron and Etienne (2005).

Hawk Creek. A light-colored sand injectite intrudes varve beds and flood sands in the Glacial Lake Columbia. Note the stair-stepping base if this sill-like injection. Some injectites are rubble-filled with chaotically-oriented stratified chunks (rip-ups). Many birfurcate. Injection episodes coincide with repeated catastrophic flooding and loading, the likely trigger. I logged twenty four sand beds (flood deposits) with a deformed zone beneath each at this location in 2017. I am the first to recognize and describe these structures in the Upper Columbia. Clear relationships between flood loading, sheeted dikes, and sand injectites exist in dozens of sites like this in the megaflood region. Flood injectites are a thing. Hawk Creek also goes by Olson Canyon and Indian Canyon. The locals are friendly. My photo taken 2017. To learn more about injectites at Hawk Creek, see my post:

Identical Dikes, Same Trigger

Dikes that intrude Miocene and Pliocene bedrock and sediments are identical in size and appearance to those in unconsolidated sediments. All share the same geologic trigger and propagation mechanism. All exposures examined by the author, dike-bearing units were exposed at the surface and overridden by glacial outburst floods. No sheeted dikes in bedrock, observed or reported, originate within buried sediments or at significant depth below the surface.

Touchet-type dikes. A typical sheeted, silt-sand-filled clastic dike in flood rhythmites, the Touchet Beds.

Your comments are always welcome:


Last 50 Posts
All Posts by Month
    bottom of page