The Tops of Clastic Dikes

The tops of most clastic dikes in the megaflood region are truncated by erosion (erosive contacts between flood beds), but not all. Sometimes you get lucky. In certain places dikes connect to their source beds, in others a few fill bands (sheets) connect smoothly with bedding. The following are a few examples of both truncated-top dikes and connected-top dikes. All photos are mine.

Bedforms become infill. The quarter is located at a bedding contact between two flood rhythmites. A fracture opened in the top of the lower bed and filled with sediment. The sand was not vented upward from some layer below. See how the rippled sand grades smoothly upward into the overlying massive silty-sand? This is how it works.


A dike that originates in coarse sand at the base of the uppermost rhythmite descends for several meters through six underlying rhythmites, branching and tapering along the way. The fracture opened in the substrate when floodwater inundated the landscape. Touchet Beds are unusually thick here in Burlingame Canyon at Gardena, Walla Walla Valley.

An unsheeted, single-fill dike intrudes downward into coarse, laminated sands at an angle oblique to the cut face. It is filled by material characteristic of the larger deposit; dike fills look like the stuff they intrude. The dike's top is truncated by coarser gravel deposited by a subsequent flood. Some amount of erosion (vertical loss) has taken place along that contact, but no a huge amount. We cannot see how the dike connected to its source bed above due to the erosion, but clearly the dike has no connection to a source bed below. The base of the structure is well exposed. This is a high energy flood channel/eddy bar setting with southward flow indicators. Smith Coulee, Channeled Scabland.

A more subtle variant. A sheeted dike, containing 10 sheets, intrudes the Touchet Bed and is filled by silty, sandy Touchet Bed sediment. Look at the left side of the dike. Follow the three left-most sheets up from the bottom of the photo to where the make a sharp leftward bend and merge with bedding. This relationship is subtle, but crucially important to understanding how these dike formed. It is the top of a clastic dike, the place where sheeting begins and a the place a dike begins to descend. This relationship applies to nearly all Pleistocene-age clastic dikes in the megaflood region of Eastern Washington. Slackwater setting. Pataha Creek Valley.

A sheeted dike, composed of 11 sheets, intrudes Touchet Bed rhythmites and is filled by Touchet Bed sediment. The three sheets along the right margin are from an older dike. They link to low-angle bands to the left of the vertical dike. Two episodes of diking here, one crosscutting the other. If you only observed the portion dike in the lower half of the photo, you would have one impression (one dike w/ 10 fill bands). If you observed the upper half, two dikes, one with 7bands the other with 3. Slackwater setting. Pine Creek, Walla Walla Valley.

A large sheeted dike intrudes Touchet Beds. Look at the left margin just left of the hoe. The left-most sheet is about 15cm wide. It begins as subhorizontal bedding (bedforms), then abruptly descends, becoming a dike fill band that joins the other bands (a younger set) and cuts across the light-colored rhythmite immediately below. Slackwater setting. Lowden, Walla Walla Valley.

A dike descends from its silty source bed. Backflood, spillover, and slackwater setting. Tucannon River Valley.

A dike descends from its gravelly source bed, the gravelly base of a rhythmite. Tucannon River Valley.

A sand dike descends from its source bed. We clearly see that the coarse base of the dark-colored sandy unit (basal portion of a rhythmite) is incised into light-colored silt beneath (slackwater top of the previous rhythmite). The top of the dike is not truncated by the sandy unit, but is connected to it. Tucannon River Valley.


Light-colored slackwater upper portions (U) are delineated from coarser grained lower portions (L) in four megaflood rhythmites (A,B,C,D). The dike which cuts across A and B can be interpreted in two ways. Either it formed prior to deposition of rhythmite C and its top is truncated by C, or it formed during deposition of rhythmite C and is continuous with the lower portion of C. Tucannon River Valley.

Convoluted bedding gives way to a descending dike. This dike didn't form at the beginning of flooding (from base of coarse grained lower portion), but a bit later, during slackwater period (upper portion). Are those dish structures in the lower bed? Slackwater setting in "late" rhythmites. Confluence of Walla Walla River and Columbia River at Wallula. You can see Wallula Gap from here.

Light-colored Touchet Beds unconformably overly Neogene fluvial gravels of the Snipes Mountain Conglomerate. The light-colored dike, descends through the gravel and is filled with Touchet Bed sediment. There are no light colored sediments beneath the gravel deposit, only basalt bedrock. Tule Road, Yakima River Valley.

Several sheeted dikes descend from light-colored Touchet Bed sediment that unconformably overlies oxidized fluvial sandstone of the Snipe Mountain Conglomerate (Ellensburg Fm). Sediment contained in the dikes is sourced in the overlying Touchet Beds, not in strata beneath the sandstone. Megafloods overrode most of fault-bounded Snipes Mountain. Several outcrops along Emerald Road are flood beds composed almost entirely of reworked Snipes Mountain. Emerald Road near Granger, Yakima Valley.

Several generations of clastic dikes are truncated by erosional surfaces (a,b,c,d) in a complex deposit consisting of loess units, reworked loess (diamict) of pre-late Wisconsin age, pedogenic carbonate paleosols, Missoula flood deposits, and other stuff. Note the dikes never crosscut younger horizons, only older ones. Rulo Site, Walla Walla Valley. Redrawn from Bader et al. (2016).

Three generations of clastic dikes were documented in three flood rhythmites (R1,R2, R3) at Moxee Mammoth Site by Lillquist et al. (2005). Redrawn from his figure.

A sheeted dike with two branches (lower left) cuts silty, oxidized sediment and has its top truncated by an erosional surface, above which lies a stack of calcic paleosols developed in loess (mostly). Elevation of the site is 285m, well below the level of Lake Lewis (~370m). Rulo Site, northern Walla Walla Valley. Pencil for scale.

Try your hand at interpreting the crosscutting and injection relationships.

The top of this dike fades out below the contact with the overlying rhythmite due to bioturbation and soil processes. The prominent cobble is in place.

Sags in bedding crosscut lower beds at Walla Walla River bluffs near Touchet, WA.

A dike descends through several rhythmites at Wallula. The top is truncated by a bedding contact between two Touchet Beds. Lower Walla Walla Valley at Wallula.

Dikes descend from and are truncated by dark-colored flood channel gravels (up-valley backflood or down-valley spillover?) cutting silty, light-colored Touchet Beds at Smith Hollow Rd, Tucannon Valley.

Sheeted silt-sand dike intrudes Columbia River Basalt at Lewiston, ID. The fill is Touchet Bed material. There are no interbeds in the basalts that could have contributed sediment like this.



Sheeted silt-sand dike intrudes a joint in Columbia River Basalt at Weaver Pit, Walla Walla Valley, WA. The joint only has to open a little bit each time to form a dike like this, apertures of ~1 cm. Note lack of shearing; joint opening is Mode I.


Dike begins to form in sandy sediment and descend.


Most of the time, this is what you see. The top of the dike just isn't exposed or is hard to access safely...


...But don't forget about the bottom. This dike descends from silty-sandy Touchet Beds and pinches out in boulder gravels deposited by the Bonneville Flood. This is the same dike shown in the photo above. Lewiston, ID.

Last 50 Posts
All Posts by Month