Where is the Neogene Seismic Record for the Yakima Fold Belt?
"Considerable work awaits sedimentologists on and adjacent to the Columbia Plateau before we will be cognizant of all aspects of the tectonic, volcanic, and climatic influences on regional sedimentation during the Neogene." - Gary Smith and others (1989)
Where is the Neogene record of seismic shaking in eastern Washington?
The Yakima Fold Belt (YFB) in central Washington is a set of west trending, fault-cored folds located in the back-arc of the Cascadia Subduction Zone. The have been rising for at least the past 10 million years. The largest folds in the in the ~14,000 km2 province (i.e., Saddle Mountains anticline) have accommodated >1 km of shortening over this time (Staisch et al., 2017). Shortening on other structures appear roughly similar (Kelsey et al., 2017). Despite their topographic prominence and proximity to major infrastructure (hydroelectric dams, nuclear sites), we know little about the seismic history prior to 2 Ma.
Was rise of the Yakima Folds over the past 10 million years accompanied by large earthquakes?
The upper Ringold Formation at the White Bluffs contains small-scale soft deformation structures consistent with sedimentation. While examples of localized deformatiion occur at Saddle Mountains (Grolier and Bingham, 1978, p. 14, 32), no seismites (e.g., continuously deformed beds) have been found by me or reported by others to date. All known examples of substantially deformed RIngold occur within a few meters of mapped faults or can be directly attributed to distinct sedimentation events. The Ringold Fm is Miocene-Pliocene in age (9.5-3.4 Ma). The Neogene covers the Miocene and Pliocene (23.0-2.6 Ma).
Aeromagnetic mapping reveals a crustal connection between Eastern and Western Washington beneath the Cascade Mountains (Blakely et al., 2011). The Yakima Fold Belt is located at lower right, the Cascades at center, and Puget Sound at upper left.
Soft sediment deformation structures (clastic dikes, sand blows, liquefaction/fluid escape structures, continuous layers of contorted strata, etc.) near active mountain belts are commonly used to constrain the timing and magnitude of past seismic events. To date, widespread soft sediment deformation attributable to seismic shaking has not been identified in the Miocene-Pliocene Ringold Fm (8.0-3.4 Ma) (Russell, 1893; Merriam and Buwalda, 1917; Newcomb, 1958; Hays and Schuster, 1983; Lindsey, 1996; DOE, 2002; Williams et al., 2002; DOE, 2010; Staisch et al., 2017; Cooley, 2021). Late Neogene sediments with characteristics favorable to both form and preserve soft sediment deformation structures are abundant within Yakima Fold Belt province, though surprisingly few are known. Snipes Mountain near Granger, WA appears to be one of few places such evidence may exist, but no paleoseismic investigation of the Snipes Mountain area has been published to date.
"White Bluffs on the Columbia - The conspicuous white bluffs along the left bank of the Columbia, beginning 12 miles above Pasco and extending 30 miles up the river, are formed by the edges of soft strata of fine, thin-bedded sand and clay with layers of pure white volcanic dust, which were deposited as horizontal sheets in Lake John Day and have not since been disturbed." - I.S. Russell (1893, USGS Bulletin 108, p. 97)
Geodetic station data for the fold belt describe the current situation as "a slow strain environment". Reidel (1984) found deformation rates were different in the geologic past, with the period from 17 to about 10.5 Ma being a distinctly more active period of fold deformation than the period from 10.5 Ma to the present. The 10.5 date is a bit fuzzy. He argued 80% of the total strain occurred from 17–10.5 Ma. Staisch et al. (2018) reported deformation rate increased during the Pleistocene (<2.6 Ma). This finding contradicts previous interpretations and notions that "the entire fold belt has continued to develop in a pattern similar to that of the Pliocene and Miocene" (Reidel et al., 1994). Most believe there were two pulses of uplift, with most of structureal-topographic relief developed in the Pleistocene.
A 200m-long exposure of undeformed Miocene Ellensburg Fm along Crusher Canyon Rd at Selah, WA. The Yakima Ridge thrust is mapped ~2 miles to the south. Several other Yakima Fold Belt anticlines are located nearby. The sediments exposed here are highly susceptible to remobilization and liquefaction when shaken. In April 2022, I inspected this section (Canyon Cliffs subdivision) as well as other large cuts in the surrounding area. No clastic dikes or features consistent with seismically-induced liquefaction were observed. The elevation at the top of the outcrop is ~1380' (420m). Google Maps Street View photo.
Miocene Ellensburg Fm at Selah, WA. My photo.
Mapped surface ruptures in the fold belt are all young features - late Quaternary (Campbell and Bentley, 1981; West et al., 1996; Blakely et al., 2011). Data from trenched faults in Pasco Basin (Bennett et al., 2016), scarps visible in lidar images (Kelsey et al., 2017), and dates on young strath terraces in Yakima Canyon (Bender et al., 2015) also demonstrate late Quaternary fault movements, but provide little information on older (Neogene) movement other than general observations about which lava flows are folded. Uplift of the Cascade Range and rise of the Yakima Folds, with exception of the north-trending Hog Ranch-Naneum structure, occurred after Grande Ronde Basalt inundated a low-relief landscape and cooled around 15.6 Ma. Kelsey et al. (2017) places the onset of YFB tectonism at 12-7 Ma based on dating of a Quaternary strath terrace at the Manastash anticline, "the anticline started to contract a few to many millions of years after emplacement of the Grande Ronde Basalt". Structural relief, rupture lengths, and steady-state uplift rate estimates are provided in several reports.
If the fold belt's 12 fault-cored folds are capable of producing earthquakes with magnitudes >M 6.8 as researchers have claimed, most recently USGS (Staisch et al., 2018), there should be ample sedimentary evidence of strong shaking preserved in >100m of Ringold Fm section exposed at the White Bluffs. The Ringold section represents >4 million years of near-continuous sedimentary filling of the greater Pasco Basin syncline. Ringold deposition was coeval with YFB uplift.
Other large, well-known exposures of thick sedimentary sections, including around Saddle Mountains, Kittitas Valley, Quincy Basin, and Yakima Valley should likewise show evidence of earthquake disturbance at recurrence intervals reported for YFB faults. Abundant borehole information from the Hanford Site (>7500 wells) should, too. Sedimentary interbeds in the Columbia River Basalt (Ellensburg Fm/Latah Fm), the Thorp Gravels (4.9 to 2.9 Ma), and Ellensburg-equivalent basin fill deposits in the Dalles-Umatilla syncline (Newcomb, 1966; Madin and McClaughry, 2019) should contain similar evidence (repeatedly deformed strata).
Thick breccia zones (>100m wide) are associated with frontal thrusts of YFB ridges. Columbia River Basalt appears to accommodate fault slip by crumbling. How seismogenic is crumbling basalt? How is energy attenuated in a wide breccia zone versus a narrow rupture?
Coarse, basaltic "fanglomerates" (boulder gravels) that mantle the flanks of YFB ridges. They are proximal alluvial fan deposits shed into adjacent synclinal valleys during uplift. If uplift was continuous since the Miocene, the coarse gravels should be abundant near folds throughout not only the Pleistocene-Holocene section, but the entire Pliocene and latter part of the Miocene section as well (i.e., in strata as old as 10 Ma). When do the angular, syn-tectonic gravels first arrive? That is, when do the folds emerge from the broad Columbia plain to form topographic ridges? Can the paleosurface on the base of the gravels be reconstructed?
The sedimentary record preserved in synclinal basins of central Washington should reflect >10 million years of seismogenic uplift. The uplift is easy to spot. Is evidence of strong shaking?
Shaded relief map of the top of the Ringold Fm, a sedimentary unit dozens of meters thick covering ~11,500 square kilometers and parts of 3 counties. Map by Kennedy/Jenks Consultants for Franklin Conservation District (Triangle Associates, 2003, p. 10).
Patterns of fault displacement come in 5 flavors: Episodic-quiescent, Episodic-active, Decelerating, Constant, Accelerating. Which style describes Yakima Fold Belt over its lifetime? Have uplift rates and seismic recurrence increased or decreased over time? How do Pleistocene-Holocene recurrence data compare to the Miocene-Pliocene record? Figure by McAlpin and Nelson (Figure 1.4, p. 10) in McAlpin (2009).
A well-established relationship exists between earthquake magnitude and the distribution of liquefaction (Ambraseys, 1988; Galli, 2000). The Galli study, which compiled a ton of data from a bunch of studies around the world, tells us that an M 6.5 to 7.0 quake will produce surface ruptures and liquefaction structures out to about 100km. The yellow circle is 100km-diameter in size and approximates the predicted damage halo for a Yakima Fold Belt quake. The circle encompasses portions of the eastern slope of the Cascades, Pasco Basin, Umatilla Basin, southern Okanogan, and western Palouse Slope.
Data compiled from earthquake-affected regions around the world reveal a fairly robust relationship between magnitude and liquefaction (figure redrawn from Galli, 2000). Based on the various curves, the damage halo for an M 6.5-7.0 quake would extend outward 25-125 km from the epicenter. A reasonable search area for liquefaction features in Eastern Washington could reasonably be constructed from a set of 75 km-radius circles placed on the centroids of all mapped YFB faults.
If the faults that core the Yakima Folds are capable of generating M 6.5 to 7.0 magnitude earthquakes, then "secondary sedimentary evidence" of strong shaking should be present somewhere in the sedimentary record of central Washington (Ringold Fm, Ellensburg Fm, Pleistocene flood deposits, modern alluvium). The Ringold Fm covers about 11,500 km2. Figure modified from McAlpin and Nelson (Figure 1.6, p. 16).
A simplified cross section through a portion of the Yakima Fold Belt centered on the Pasco Basin (Reidel et al., 1989). Seismic data and wellbore information indicate a basal decollement exists at ~10km depth (Campbell and Bentley, 1981; Miner, 2002).
Modern cross sections through the Saddle Mountains anticline (Crane and Klimczak, 2020) redrawn by me.