Geology in the Other Saddle Mountains

The Saddle Mountains form a coherent ridge that runs for more than 80km east to west across south-central Washington. The ridge is a north-vergent uplift with thrust faults on its northern flank. There is a prominent footwall bench along a portion of its north flank called Smyrna Bench, which is composed of sediments younger than the basalt bedrock (Pliocene Ringold Fm, Pleistocene flood deposits, and calcretes). Strata on the south flank of the range rolls gently into the broad syncline of Pasco Basin.

Cross sections through the Yakima Fold Belt in south-central Washington (top) and the Saddle Mountains anticline near Sentinel Gap (bottom). Thrust geometries are a bit old fashioned, but you get the picture. All layers belong to the Columbia River Basalt Group except for "RF" (RIngold Fm), seen at the very top of the fold. North and south are reversed in the two cross sections.

Geography of the Saddle Mountains. The Saddle Mountains are a coherent structural and topographic ridge cut by the Columbia River at Sentinel Gap. West of the river, the ridge is called Manastash Ridge. East of Smyrna Bench, the fold narrows and begins to plunge, dying out just south of Othello. Clearly, its all one mountain range. Important type sections for the Ringold Formation are located at White Bluffs and Taunton.

I was recently struck by statements about the geography of Saddle Mountains in an article by the U.S. Geological Survey (Staisch et al., 2017):

"The Saddle Mountains are ~115 km long, spanning from the eastern edge of Kittitas Valley to just south of Othello, Washington. The westward continuation of the Saddle Mountains, west of the Hog Ranch–Naneum anticline, is called the Boylston Mountains..."

"...outcrops on the Saddle Mountains differ from typical Ringold Formation strata, which is best exposed at the White Bluffs along the Columbia River and near the town of Taunton, Washington, southeast of the Saddle Mountains, respectively."

Saddle Mountains geography according to Staisch/USGS. Apparently, Taunton is not in the Saddle Mountains, but "southeast of Saddle Mountains". Hmm. The old Taunton power plant (Milwaukee Road Substation) and the railway it served are located on the north flank of the range more than 60m above valley floor and Crab Creek. White Bluffs is to the southeast, Taunton not so much.

115km length for the ridge was given in the article. That sounded too big, so I measured it in Google Earth. From Foster Creek near Badger Gap to the Sagehill Rd-Radar Rd intersection, I came up with 85km using a sinuous (generous) path length. Nitpicky? Maybe. But that's a 30km (25%) discrepancy that 7 co-authors and probably 3 reviewers missed. The entire structure may extend west through the Boylston Mts to eastern Kittitas Valley, but its segmented.

Errors happen, but the kinds of errors I was seeing in the first few pages were weird. I began to wonder if other things might be out of place.

I examined the 6 measured sections through Ringold strata shown in Figure 5. More than half of the total amount of section is depicted as "covered" (not exposed, not described). That's a lot. Generally, geologists don't report huge amounts of non-exposure. For comparison, a young Kevin Lindsey reported 24% covered in his 28 sections through the Ringold (>2750m of section). While I can understand the author's choice to do it the way they did (using top of Elephant Mountain basalt as base of section), it smacks of an overplayed hand.

Calcrete-cemented alluvial fan gravels cap the Ringold Fm at Saddle Mountains. Along the rim of Smyrna Bench, the distinctive Savage Island lakebeds-and-sands interval (uppermost Ringold) are well exposed. If the ridge poked out above Lake Ringold, then the lake's shoreline crossed the ridge somewhere. Deposits at Smyrna Bench suggest the shoreline was no farther east than Royal City (west of Wahatis Peak). According to Reidel's geologic map, Smyrna Bench lies in the footwall and no thrust exists along Crab Creek. The prominent slope break is erosional.

A few months ago I was working in the area and since I hadn't been to Saddle Mountains yet, I figured I'd head over there and check out one of the USGS's measured sections. The map shows only on road crossing Saddle Mountains, so I headed straight for it. Corfu Road bisects the range and is super rough on the north side, but smooth sailing on the south. Its the obvious place a geologist would go first. The famous Shell ARCO BN 1-9 well was drilled here and a graben scarp trench was cut nearby by West et al. (1996). I figured the USGS folks would have measured something here, too. Nope. Their nearest section was miles off. Hmm.

Thin fluvial sands in the upper Ringold (Savage Island Member) are nearly always deformed, iron-stained, and sometimes firmly cemented. You can even find some pretty cool boulder-sized concretions that preserve the soft sediment deformation.

A conspicuous, 1.5m-thick, gray, crossbedded sandstone occurs below Savage Island lake beds and above the highest red conglomerate at Smyrna Bench. Possibly ashy. Doesn't look like a sheet flow deposit to me. Sluggish fluvial instead (sand bedded channel).

The post-Ringold alluvial fan-silt Diamict-calcrete stratigraphy (so-called "Plio-Pleistocene unit") at Smyrna Bench is more complicated than published articles, gray lit summaries (Hanford reports), and map unit descriptions would suggest. Its character is remarkably similar nearly everywhere.

A few of my notes on the upper Ringold and post-Ringold "fanglomerates". Coarse alluvial fan gravels occur as thin beds (channel fills, <1m thick) or thick, amalgamated debris flow deposits (channels and levees, 5-15m thick). These coarse grained, proximal, fluid flows (not gravity topples) were shed off the Saddle Mountains anticline during uplift. Dips on the fans at Smyrna Bench (footwall) are remarkably shallow (<5 deg). Silt diamicts (reddish layers in the photo) accompany the fan gravels and often exceed their volume. They contain abundant evidence of soil formation (bioturbation, horizonation, oxidation, cementation). Fan gravels and silt diamict seem to be something of a paired system. Because of the immediate proximity of Smyrna Bench to the fold, I interpret the silt diamicts there as proximal, reworked loess that accumulated on the fan surface and that was periodically swept entrained by distributary channels.

Windblown dust was abundant during fold development and fan deposition. Dust was probably generated from large areas of newly-exposed Ringold sediments following drainage of the last Pliocene lake from Pasco Basin at ~3 Ma (base level dropped >100m). The diamicts are dominantly silt, but contain sparse pebble- to gravel-sized clasts and rip-ups, most commonly calcrete and laminated siltstones. Because they are dominantly silt, they display faint bedforms, yet the presence of larger clasts indicates the flow regime was capable of transporting material much larger than silt. However, silt might have been all that was available to the shallow, surface-sweeping fan channels. The contacts between the fanglomerates and diamicts are sharp, not gradational. Silt diamicts are found elsewhere in the region, commonly along the margins of the Missoula floodway, where ancient floods cut shallow channels cut through virgin loess hills (Spencer and Jaffee, 2002; Bader et al., 2016). Abundant silt in Columbia Basin is commonly linked to glacial outburst floods, but there exists a time gap a the end of the Pliocene when a considerable volume of non-flood related loess could have been generated. The gap exists between base level fall (~3 Ma) and the formation of large proglacial lakes along the Cordilleran Ice Sheet margin (~1.5 Ma).

Sheet flood sand beds are not common in the fan complex, which is surprising. They should be here. The diamicts may be serving as the sheet flood facies. Standard facies models for alluvial fans (i.e., Miall) may not work perfectly.

The diamicts are not far-transported and do not appear to be the distal, down-fan facies of the gravels. Nor do they appear part of the paleo-Crab Creek valley's floodplain ("sidestream deposits"). But if they are, then energetic fans spilled directly off of a steep, bedrock fold front onto a wide, flat, sluggish, loess-dominated floodplain. That explanation seems weird since the fan gravels are essentially flat-lying right up to the thrust. The landscape is too compressed. Basic fan geometry and repose angles dictate more space is required between colluvial hollows (sediment source) and the fan toe/floodplain (sediment sink). Saddle Mountains anticline is not wide enough (5-10km) for deep canyons to have formed and collected a trough of low-angle fan deposits. And the ridge isn't deeply dissected. Shallow dips on fans seem to require a relatively long run-out distance.

Is it possible they aren't sourced at Saddle Mountains? If the gravels at Smyrna Bench are not sourced from the Saddle Mountains, then things get a lot more interesting: Where's the topographic high?

Amalgamated fanglomerates with calcic soil profiles occur at the same stratigraphic position as fan gravel-silt diamict-calcrete deposits in the previous photo. Lateral changes in fan gravel thickness along the Saddle Mountains front likely reflect larger and smaller fans (e.g., larger and smaller drainages) within a more or less continuous fan apron.

This photo in Grolier and Bingham (1978) of a small fold along the Saddle Mountains crest caught my eye years ago. I always meant to check it out, so I did. Photo below was taken near fold hinge.

The formation names have been revised since 1978, but the fold is still there(!). Elephant Mountain Basalt flow is intercalated with west-derived Ellensburg Fm sediments of varying thickness. Little to no Ringold in the photo. The "calcrete cap" (actually a fanglomerate-calcrete-loess diamict unit) rests unconformably atop the older folded units and itself folded to near vertical just off the photo to the right. I can't spell Pleistocene, apparently.

Another view of the stratigraphy along the ridge crest. View to the east. The calcrete-bearing unit is thick here and appears relatively conformable with the underlying units, but its not. It rests directly on the weathered surface of the Elephant Mountain Basalt. Check out the spectacular baked contact at the base of the flow.

Calcrete unit is folded to near vertical near the fold hinge. I'm guessing there is a small fault or two here, but the exposure isn't great. Shadow at left is the steeply-dipping Elephant Mountain Basalt. View is due north into Crab Creek Valley.

Ellensburg-affinity cobbles litter the surface along the ridge in the vicinity of the fold hinge. The folded calcrete pedocomplex, which overlies the cobbles, crops out behind. View to the east.

Wahatis Peak makes for a nice detour if its not too hazy. Summit best gained by DR-Z 400 or Tenere 700 (with heated grips and optional center stand).

Spectacular spheroidal weathering is common in certain basalt flows at Saddle Mountains. Check out the Wahatis Peak summit area and the gullies that drain Smyrna Bench (Hunzinger and Elephant Mountain flows).

I parked near the saddle crest and hiked north to the first outcrop I could see. Found really good exposures of the upper Ringold and cemented Plio-Pleistocene fan gravels along the rim of Smyrna Bench. Spectacular soft sediment deformation in thin fluvial sands within the laminated silts below the capping calcretes. And there was our old friend the 10.5 Ma Elephant Mountain basalt down there, prominent in the floors of gullies. Why no USGS section here?. Hmm.

A hasty sketch of an outcrop at Smyrna Bench made in fading light. I've since returned to do a proper job. That day, I was focused on documenting soft sediment deformation in the uppermost Ringold. I'd heard that little of it had been found out here, so I was surprised it was in such abundance (8 deformed horizons in a 3m-high exposure). To my eye, the upper Ringold at Smyrna Bench appears nearly identical to exposures I have described to the east, west, and south. I must be missing the "differing outcrops" mentioned in the article.

A quick field map to jog my memory and help hasten my return. Hike follow the summit graben scarp for a few hundred meters. You won't meet anyone but the odd coyote on your hike.

There is something odd about the USGS/Staisch article. I recall feeling strange about it at first reading, but not because of its overreach with respect to the Saddle Mountains' seismic hazard. On the bright side, the work covers a lot of ground - a truly collaborative, interdisciplinary effort. Most of the authors have been at this for a long time, especially the intellectual leads Harvey Kelsey and Richard Blakely. The digitized update of Reidel's map is great, though a few helpful references were left off (Wahatis Peak, Corfu Road, Crab Creek, the railroad, the road system). The new dates on the Ringold will certainly endure.

However, the field work portion of the project as expressed through the measured sections (actual geology) leaves a lot to be desired. The sections lack certain details that someone familiar with the area would know to include. They lack a familiarity with alluvial fan systems and calcic paleosols. Tectonism and landscape stability constitute the important story preserved in the Ringold. Robust editorial review also remains a question. Its odd that Steve Reidel, the local expert, is not a co-author. All authors are from away (Seattle, California, Denver, etc.).

In the end, the article succeeds in moving the ball forward on Yakima Fold Belt development. I recognize the substantial contributions by the lead author, but fail to see a novel theory behind any of it. And I just cannot shake the feeling that a bunch of senior geologists conspired to prop up a new hire.

The lonely red tank awaits your visit. Bring your drill rig.


Staisch et al., 2017, Miocene–Pleistocene deformation of the Saddle Mountains: Implications for seismic hazard in central Washington USA, Geological Society of America Bulletin

Last 50 Posts
All Posts by Month