Calcrete Field Trip - Herman Railcut
Directions - From the Cunningham Rd-Hwy 17 intersection at Othello, drive 4.5 miles north on Hwy 17. Turn L (west) onto Herman Rd if traffic is light. There is no left turn lane here. The safer choice for a string of vehicles is to turn R onto Herman Rd and make a U-turn in the Yancey Pallet parking lot, then cross Hwy 17 orthogonally. Drive a mile west on Herman Rd and park in small pullouts on both sides of road right next to RR tracks (6 vehicles) or proceed to the bottom of hill below tracks where a wide pullout can accommodate many vehicles. Hike the tracks to the north cut. For the south cut, make a gently rising traverse up the sagebrush slope, behind the stack of ties, toward a small tree in the distance. Stay above the wet ditch that parallels the tracks. The outcrop south of the road can get weedy and buggy in late summer. Better to go north after mid-July. Pray the mid-summer wind is not from the west (feed lots).
Outcrop Summary - The railcuts are located along the east bank, an outer bend, of Crab Creek Coulee just south of the Othello Channels constriction. To the northwest and across the valley is the butte-and-basin scabland of Drumheller Channels. To the east, the western Palouse Slope (loess hills). The railway runs along the flood-cut escarpment. The planar, flood-beveled surface that forms the bench above the outcrops extends east to Hwy 17 and dips gently south to Othello. Armoring calcretes have resisted stripping by repeated south-moving megafloods. We are marginal to high energy, erosive flows that rushed down the coulee. We are far removed from alluvial fans spilling from Saddle Mountains and Frenchman Hills uplifts. The slope of the Crab Creek valley is low and its stream power minimal. The boulder gravel found here is flood-laid. Savage Island lakebeds underlie the boulder gravel and accompanying water-worked pebbly silt diamict and nodular loess. A 1.5m-thick armor of calcrete caps the section.
In 2020, a 300m-long waterline trench was opened east of the Cunningham Rd-Hwy 17 intersection. In its walls, I found loose calcrete gravel with coarse, uncemented basaltic sand, south-dipping foresets, and large rip-ups of Savage Island, cemented loess, and calcrete. That late(?) glacial flood deposit underlies a large portion of Paradise Flats. See Coyan Rd write-up.
Druheller Channels maps - a.) Hand drawn map by Bretz (1928, Fig. 19), b.) Lidar hillshade from Washington Lidar Portal, c.) Aerial photo from Google Earth. Herman Railcut site is located just off the right edge of the map, just beyond the deeply scoured scabland surface. Ringold and some Pleistocene sediments, armored by calcrete, escaped complete stripping by Scabland floods.
Overprinted - Calcrete overprint on boulder gravel. Another 1.5m of calcrete, above, is out of the picture.
Post-Ringold unconformity - The cemented boulder gravel lies unconformably atop Ringold Savage Island lakebeds at the base of the exposure.
Two or three carbonate ledges - Pliocene Savage Island siltstone present at the base of the outcrop (not seen in photo above) is overlain by a boulder gravel capped by a calcrete ledge. Look for boulders of greenish sandstone and white mudstone in the gravel. Above the first ledge are lenses of silt-pebble diamict (water transported) and brown nodular loess (wind transported). Character of sediments between the two ledges differs between N and S cuts. A dense, blocky, fully plugged 1.5m-thick calcrete caps the outcrop and contains a few cobble-sized basalt clasts, but otherwise little hint of the parent material. It is almost certainly a composite (welded) paleosol.
Boulder gravel - South-dipping foresets and imbrication verify energetic southward transport down Crab Creek Coulee. Rip-ups of gray-green sandstone, a friable white-green mudstone, and hard white calcrete sized to 50cm occur with basalt. Sandstone is Vantage interbed scoured from Babcock Ridge to the north. Mudstone is from a nearby Ringold source. Lower-half carbonate rinds on most clasts is commonly used as an indication of modest age. However, given the thick overlying calcrete, that rule of thumb may need to be revisited. All indications is that this deposit represents a single ancient flood containing the same lithologies found in the expansion bar at the mouth of Frenchman Springs cataract.
Basalt cobbles in upper calcrete - Firm basalt cobbles fully enclosed by thick, dense upper calcrete unit indicates the parent material was coarse grained, at least in part. Disintegrated basalt cobbles have been reported in a calcrete capping a reversed-polarity gravel at Old Maid Coulee (Baker, 1978) and a similar gravel at George (Bjornstad et al., 2001). Weathered basalt cobble are not conspicuous or abundant at Herman Railcut, but are present.
Origin of the boulder gravel - The boulder gravel cannot be an alluvial fan spilled north from Saddle Mountains or east from Frenchman Hills. Both ridges are too low and too distant. Likewise, the gravel is not that of ancestral Crab Creek, as its valley's low gradient indicates the stream was never competent enough to move such coarse sediment. All evidence indicates it was deposited by at least one "ancient" (pre-Wisconsin) Ice Age flood, possibly two.
Three ages of outburst flood gravels are recognized in the greater Channeled Scablands region (Van Alstine, 1982; Reidel and Fecht, 1994; Baker et al., 1991; Bjornstad et al., 2001; Pluhar et al., 2006; Cooley, 2022). The gravels at Herman Railcut, George, and White Bluffs Overlook are likely Qfs1 and time-correlative with certain calcrete-cemented alluvial fan gravels (non-flood) exposed along the Saddle Mountains crest.
Qfs3 - Touchet Bed silt-sand-gravel rhythmites and youngest "Hanford Fm" sands and gravels (sheets, bars). Deposits of jokulhlaups during Last Glacial. Contains extra-basinal, non-basaltic "exotic" clasts (granite, quartzite, schist, etc.). Widespread and well exposed throughout the Ice Age floodway in WA, OR, and ID. Contains Mt. St. Helens Set S ash. "Missoula flood deposits". Normal magnetic polarity.
Qfs2 - Cemented coarse and fine flood sediment with pedogenic carbonate to Stage II-IV, silcrete, and cambic soil horizons. May contain extra-basinal, non-basaltic "exotic" clasts. Not widely exposed. Estimated at 200,000 years old. Normal polarity.
Qfs1 - Flood sediments recognized by capping paleosols. May contain extra-basinal, non-basaltic "exotic" clasts. Estimated age >790,000 to ~1.8 million years based on reversed polarity signatures (older than Brunhes-Matuyama reversal). These so-called "pre-Wisconsin gravels" are not widely exposed at surface. Information is limited, mostly coming from drill cores from the Hanford Nuclear Site (Cold Creek bar and White Bluffs). Appears to pre-date oldest dated Palouse loess (~1.1 Ma). Obscure contact with underlying units including the Plio-Pleistocene Cold Creek unit, uppermost Ringold Fm, and associated paleosols.
George Gravel Pit - A similarly thick calcrete overlies a boulder gravel in the Quincy Basin to the north near George, WA, described by Baker (1973, p. 9),
The section at George (NW1/4 Sect 31, T19 N, R24E) shows a typical foreset flood gravel containing boulders as large as 3 feet in diameter. The uppermost 2 feet of the deposit is capped by horizontally laminated caliche...boulders of the Vantage sandstone...could only have been derived from outcrops along Babcock Ridge 6 miles to the [north]west.
Bjornstad et al. (2001, p. 707) summarizes information from previous reports including Webster et al. (1976, p. 11),
...[the gravel deposit] appears to be Early Pleistocene, primarily on the basis of the total depth of weathering (up to 4m), along with thick weathering rinds (up to 8cm) on basalt clasts.
In March 2020, I visited Baker's "George Gravel Pit" located along Road 1 NW (47.101035, 119.870250). A sign near the gate reads "George Drop Box". I observed south- and east-dipping foresets in a >3m-thick gravel deposit containing clasts of green opal, green-gray tuffaceous Vantage sandstone, and weathered basalt capped by >1m of dense, blocky calcrete.
Brief mentions of the George Gravel Pit and a few others with similarly old, soil-capped gravels are made in Bretz et al. (1956), Richmond et al. (1965), Webster et al. (1976), Nummedal (1978), Grolier and Bingham (1978), and an IAFI online field guide by Karl Lillquist (2017). Waitt (2021) in GSA Special Paper 548 describes the George site and a nearby one containing two flood gravels and calcrete that he calls Burke (after a nearby rail stop). Baker's nearby "Winchester Wasteway" site site is a canal exposure off Road 6NW. The canal is now called Frenchman Hills Wasteway.
Do carbonate fronts advance up or down in soils? - The world literature on pedogenic carbonate is clear on the direction of water movement in soils. Soil water, which transports dissolved carbonate, moves in various directions (Zamanian et al., 2016). Water percolation from the surface moves dissolved ions downward. Evapotranspiration moves only water upward. Throughflow moves both water and ions subhorizontally with bedding. Capillary rise moves water and minor amount of dissolved carbonate in any direction. The movement of carbonate in the soil profile is therefore restricted. Dissolved ions move almost entirely downward and sideways with soil water, not upward with evapotranspiration. For surface-derived Ca2+ ions and other dissolved inorganic carbon species, its a one-way trip.
Where upward growth (thickening) of Bk horizons occurs, the addition of new sediment (new parent material) to the surface is required. In the windy Palouse, for example, surface burial by silt raises the ground surface. With burial, the Bk horizon shifts upward and thickens. The A-horizon may or may not be preserved in the buried soil and the lower B-horizon becomes inactive (too deep).
At Herman Railcut and George, secondary carbonate clearly overprints the top of the boulder gravel, evidence that the CaCO3 front moved downward and horizontally. Horizontal stringers and vertical seams of caliche in finer grained diamict attest to horizontal and vertical flow. Little evidence of fossil organic matter (A-horizon) is present at Herman Railcut or elsewhere in the Othello area. Fossil B-horizons dominate and are recognized by their suite of characteristic features (cemented root casts of various diameters, cicada burrows, horizontal stringers, vertical seams).
Seams and stringers - Horizontal stringers and vertical seams of CaCO3 in diamict and nodular loess. Cementation shown here is both diffuse (mostly not visible) and concentrated along fractures (bright white). Cementation here is at the higher end of Stage I, but nowhere near Stage III.
Brown nodular silt-sand - Lenses of the brown, sandy, nodular sediment occurs with diamict between the two carbonate ledges. Diamict implies water-transported loess that forms few to no bedforms, but entrains small clasts that appear to float in the silt matrix. The sediment is mostly quartz and feldspar grains - the typical composition and size fraction in eolian deposits of the Pasco Basin (mostly silt with some fine sand), but a closer look will reveal coarser "floating" grains of angular basalt (sand to pebble sizes). Here, loess deposits reflect a more proximal position than most Palouse loess. The presence of the fine sand and "sand sheet" deposits should be expected (Sweeney et al., 2017). Are the basaltic grains eolian (saltation)? Colluvial (fallen from local outcrops)? Products of local overland flow during storms (non-channelizd sheet wash)? Moved within the soil profile by burrowing cicada?
Spring on Ringold - A spring emerges from the top of the low-permeability Savage Island at the far south end of the south railcut. The silty-clayey Ringold restricts downward infiltration and promotes lateral movement of irrigation water supplied by center pivots in the field above. Recall that the regional water table was artificially raised with construction of Grand Coulee Dam and the Columbia Basin Irrigation Project canal network. No tufa or travertine is precipitating at this spring as is common in parts of the Rockies and Desert Southwest where limestone bedrock is abundant. Columbia Basin's basalt bedrock appears to contribute very little calcium to soils or groundwater despite its Ca-bearing minerals.
Refs: Book N, p. 83, Book M, p. 127-129, Book O, p. 28, Baker (1973, 1978), Webster et al. (1976), Nummedal and Baker (1978), Bjornstad et al. (2001, p. 707)
Links to Other Stops:
Lind Coulee Fault at O'Sullivan Reservoir https://www.skyecooley.com/single-post/lind-coulee-fault-at-o-sullivan-reservoir