top of page

Sanpoil-style Syn-sedimentary Deformation

Undeformed flood package. A single, meter-thick Missoula flood bed formed by when a rush of water, released out of western Montana, entered Glacial Lake Columbia, deepening the lake and driving a pulse of sediment-laden water across the lake floor. Energy decreases upward through the bed. Bouma terms apply here. This example is not from the Sanpoil, but from a nearby shoreline bluff along the Columbia River.

Mud squirts. Mobilized lacustrine muds are dragged upward into an overriding sand bed deposited during a Missoula flood. The "T-shaped mud squirts" formed as a result of soupy (saturated) lake bottom muck being disturbed by sand carried at the base of the flow, essentially plowing its way along the lake bottom. Mud squirt formation involves rapid deposition and loading during flooding and viscous drag by the bottom-hugging current. The added water column is doing the loading, not really the sediment. The sticky, clayey, ductile material doesn't just erode or disaggregate like granular material might. It hangs together in laths that quickly deform ductily in response to the flow. Mud squirts are not injection structures (not clastic dikes). They are fundamentally different than the sheeted sand-silt-gravel injection dikes found in the Touchet Beds to the south. Rather, they are slender, three dimensional bulbs that appear as ascending flame structures in 2D and often alternate with sagging load casts of sand. Close observation will confirm minimal amount of crosscutting is taking place. Twists and turns taken by the flame structures largely conform with swirls in the sand that surrounds them. Termination occurs at their t-shaped tips. T-shaped portions are often parallel with bedding in the sand.

The Sanpoil Valley is an arm of Lake Roosevelt located upstream of Grand Coulee. Shoreline bluffs expose Glacial Lake Columbia varves, Missoula flood deposits, and younger outwash gravels between Keller, WA and the Keller Ferry dock (free ferry across the Columbia run by WSDOT).

In 1882, Captain T.W. Symons wrote of the Sanpoil,

Three miles below we passed the mouth of the Sans Poil River. This comes in from the north, rising the mountains nearly due west of Kettle Falls, and flows through a region in which there is much good farming land. This word has been variously spelled, but the above I believe is correct, as it seems to be a French name applied to the Indians living along its banks, on account either of the scarcity or shortness of their hair and beard, or from the fact that they were very poor and had no furs to sell to the traders. Old Pierre told me this latter was the origin of the word.

And about the inital encounter of Grand Coulee a few mile downstream,

After about eleven miles are passed we come to the mouth of the Grand Coulee, which, however, would not have been noticed if old Pierre had not told us, as it presents the same appearance as the rest of the left bank, the Coulee bottom being high above the river.

Observational test. A few questions can be asked in the field in order to distinguish flame structure from clastic dikes:

1.) Are the structures dragged up from intact strata below (can you trace the connection back to the source)?

2.) Do the structures crosscut layering in the sand (injected)?

3.) What is the nature of any internal stratification (banding)? If the banding used to be bedding, then it formed prior to deformation (flame structure). If banding is a product of the deformation, then its likely a clastic dike.

4.) What is the nature of the contact from which the structures originate?

Both flame-like and dike-like behavior. In most of the structures, internal banding used to be bedding in a varved interval deformed by the sand (flame structure behavior). But portions of many structures crosscut bedding (dike-like behavior). There is clearly both soft sediment deformation and injection going on, but the dominant process at play is rapid sedimentation (sedimentary loading), not crack initiation and propagation (hydrofracture). The deformation is entirely caused by flood-loading (density current loading).

Sedimentary structures in the Sanpoil Valley are a.) classic loading structures (flame structures, load casts) found in any textbook on sedimentary structures, and b.) not so well understood structures that involve elevated pore pressure and fluidized sand injection along bedding (sand injectites). The formation of both can be attributed to the energy imparted to the lake bottom by density currents generated by outburst floods. Soft sediment structures are commonly found along Banks Lake, Hawk-Indian Creek valley, near China Creek, and elsewhere in the Columbia River Valley where Glacial Lake Columbia beds were inundated by Ice Age floods.

Muddy laths. Muddy laths ripped up by and incorporated into a flood sand bed. Not fractures filled by sediment. No hydraulic fracture here. Just rapid sedimentation and viscous drag. Banding preserved in the clayey rip-ups is bedding.

Turbidite is not out of fashion in Sanpoil Valley. Articles written a few years ago interpreted certain Missoula flood deposits as "turbidites". Rhythmites exposed in the Sanpoil Valley and upper Columbia River Valley do indeed exhibit a Bouma-like architecture. Floods beds entered the "stilling basin" of Glacial Lake Columbia and exhibit different sedimentary characteristics from flood deposits farther south in ephemeral Lake Lewis. Authors don't use "turbidite" anymore. The term has fallen out of favor. But the field geology is hard to deny. When you're up here in the Sanpoil Valley, staring at this stuff with your own eye balls, "turbidity current" and "density current" just work.

Looks like lateral injection. Coarse sand layers with abrupt contacts top and bottom - that is, sand beds that lack a sedimentary transition with the varves below might be layer-parallel sand injectites (sills). The term "injectites" is borrowed from petroleum geologists (watch on YouTube "Webinar: Injectites Virtual Field Trip" or read articles by Hurst, Huuse, Cartwright, Cobain, Monnier, and others). Injected sands in the Sanpoil Valley form relatively thin sills and low-angle dikes that invade thicker varved intervals. Look at how the upper and lower contacts of the sand bed are identical - just clean lines. No scoured, wavy bottom surface, no gradational top. That's not a sedimentation pattern that makes sense with waxing and waning flood energy and resumption of lake bottom deposition.

My tests for sand injectite vs. flood-laid sand bed include:

1.) What is the nature of the top and bottom contacts of coarse sand beds? Do sedimentary patterns favor typical sedimentary transitions consistent with changes in flow energy or abrupt, intrusion?

2.) Are suspected sand beds internally stratified or massive?

3.) Do sets of thin sands in varves merge into thicker beds along strike?

4.) Are there angular relationships between sand beds and varved intervals?

Beach hiking at low reservoir periods. Hike the west shoreline of the Sanpoil Arm during low water periods (Winter-Spring) to access a few good exposures of lake and flood beds, some deformed.

Rotational slide blocks. Active rotational landslides along the west shoreline of Sanpoil Arm. The scarps provide excellent views of the local stratigraphy described by Atwater in the 1980s. Some of his outcrops are gone, but new ones can be found nearby.


Here's a printable 8.5" x 11" PDF of the map for the lower Sanpoil Valley. I made it. PDF is the same map as shown below.

Sanpoil Area Map Hwy 21
Download PDF • 2.15MB


Last 50 Posts
All Posts by Month
    bottom of page