top of page

Fault Scarp or Ranch Road? Review of USGS paleoseismic trench near Wallula, WA

Trenching at Wallula, WA

In 2020, the U.S. Geological Survey excavated a trench across a suspected fault scarp near Wallula, WA in an effort to better understand earthquake hazards in south-central Washington (Brocher and Sherrod, 2018; Angster et al., 2020). In this post, I review the land use history of the 'Starthistle Trench' (formerly called 'SUK Trench') site recorded in several decades of aerial photographs.

The site is located along the trace of the Olympic-Wallowa Linement and locally the Wallula Fault Zone. The site itself is located in open rangeland - geomorphically stable, sparsely vegetated, undeveloped, and never glaciated. Some windblown dust is accumulating today, but no other sediment. No oral history or historical accounts of a fault rupturing the ground here are known. The combination of the Horse Heaven Hills ridge (s-wave amplification?) and the Columbia River trough (s-wave deamplification?) provide nearby topographic relief.

Location of Starthistle/SUK Trench. The trench is located along the Wallula Fault Zone east of the Columbia River off Hwy 12 west of Walla Walla, WA.

Namesake. Graffiti on an old shed visible from Hwy 12 just east of Wallula Jct provided the name for the SUK Trench, thankfully renamed Starthistle Trench.

Trench site. The Starthistle Trench site on a bench above the mouth of the Walla Walla River near Wallula, WA. Hwy 12 is seen at top of this June 2019 photo.

Mapped scarp. Yellow line highlights a feature USGS Geologist Stephen Angster interpreted to be a young fault scarp. The dissected bench traversed by the feature is composed of late Pleistocene Touchet Beds (Missoula flood slackwater deposits, 18-14 ka) capped by <1m of Holocene loess. A weak soil is developed into the loess. Local gullies are dry, but carry intermittent flow during high rainfall events. The water table is located several meters below the surface. Vegetation cover is sagebrush, grasses, and some forbs. Base image is a bare earth lidar hillshade.

Aerial photo interpretation. I used nine sets of aerial photo sets (1996-2021) to compile a map of features created by humans and livestock in the vicinity of the Starthistle Trench. Mapped features include ranch roads, cattle trails, fencelines, watering structures, a borrow pit, and a single small building. The same features appear in the lidar image. A network of two-track ranch roads (black lines) were established along fencelines and other intuitive routes. Some appear to have been built decades ago by a dozer, but are still in use today. The trace of Angster's "scarp" follows one such track (yellow line, black line).

Overlap. A road and the "scarp" appear to be the same feature.

Trench log of east wall. No faulting observed. Blue sediments are Touchet Beds. Tan sediment is latest Pleistocene-Holocene loess. Orange blobs closely resemble rodent burrows (krotovina), common throughout soils of the region. They look nothing like sheeted clastic dikes described by (Cooley, 2020).

Burrowing rodents not devastating earthquakes. Features like this are very common in modern and buried soil profiles developed in Touchet Beds and Holocene loess of south-central Washington. Irregular, funnel-shaped forms with unsheeted fills that connect to darker soil above are not sand blows, clastic dikes, or features related to liquefactiion. The space created for infilling the sediment was excavated by rodents. Photo from Angester et al. (2023).

Review and Commentary

No fault was found in the Starthistle Trench.

The spatial correspondence between a two-track ranch road and the "fault scarp" mapped and trenched by USGS is striking. The scarp appears to be a road created by a dozer sometime after WWII, not a break in the Earth's crust. Remnants of the road surface are seen in the trench log. A cultural feature, not a fault.

"Widespread liquefaction" was identified in dry Holocene loess positioned well above the water table. No evidence of wetland sediment or soil is present. The underlying silty-sandy Touchet Beds (18,000-14,000 years old) are undeformed.

In south-central Washington, modern soils (mapped locally as Sagemore Series fine sandy loam) are commonly developed into a half meter of post-glacial loess (windblown silt). While new discoveries are certainly possible in the Channeled Scablands, liquefaction in Holocene loess has never been reported. There are hundreds of excellent roadcuts, railcuts, and cutbanks within 100 km of the Starthistle Trench site that expose identical sediments that have been examined by thousands of geologists, researchers, soil mapping crews, road engineers, dam engineers, Hanford scientists, guide book authors, field trip participants, and hungover college students over the past century. Someone prior to Angster would have noticed liquefaction features in Holocene loess if present. None has. Alternative explanations should have been more thoroughly considered by USGS.

Angster reports finding MSH Set J tephra (~13,000 years old; 12-10.5 ka according to Mullineaux, 1996) in the trench. Ash beds provide precise age control on the sediments that enclose them. Finding the Set J ash suggests faulting at Wallula is a bit younger than at other sites in the region and shortens the recurrence interval. A younger date increases the hazard. However, it is unlikely the ash bed is actually the Set J. The distinctive Set J tephra is all but absent from south-central Washington due to its limited eastward drift during eruption and extensive soil mixing since that time. Geologists who live and work in Eastern Washington would be quite surprised to find Set J in any upland exposure this far east. More likely, Angster found the bright white MSH Set S tephra (16,000 years old), which is ubiquitous throughout the Walla Walla Valley, Yakima Valley, Umatilla Basin, Lewiston Basin, Tucannon Valley, and Pasco Basin. Set S ash is readily observable in roadcuts near Wallula today at 65 mph. The tephra in the Starthistle Trench is almost certainly Set S, not Set J.

Out Over Your Skis

No one benefits from inflated seismic hazard ratings. Mistaking a road for a fault, confusing Set J for Set S, and taking common soil features for liquefaction are major errors. Type I blunders. Alternative explanations for faint, irregular features in soil - a key observation of the report - should have been discussed in detail. Alternatives might include piping, rodent burrows, hoof shear, an irregular oxidation front, dry ravel, recent rainfall, rooting zone phenomena, etc. Chalk the mistakes up to irrational exuberance, youth, or - my preference - ambition. Whatever you choose, know this: the 'end of all things sublunary' will not be brought about by an Eastern Washington earthquake. It is entirely possible the rise of Horse Heaven Hills was accompanied by dozens of quakes never exceeding Magnitude 5.9.

Vincent School. The Vincent School was built in 1911 in Umapine, OR. Umapine was the epicenter of one of the region's largest quakes, the 1936 Stateline Earthquake (M5.8, VII). Anyone looked into the school's records of damage from that quake? Photo taken 2019.


Angster's team presented their Starthistle Trench findings, including the claim of "widespread liquefaction" and the discovery of "Mount St. Helens Set J tephra", to the 2020 American Geophysical Union meeting (poster session) and via Zoom to the Wenatchee Valley Erratics in 2021. Northwest Public Broadcasting ran a story on their trench work at Wallula in December 2019. The word has gotten out. Links to reports and media below.

Who's In Charge?

I am mystified how managers at the USGS Seattle Field Office-Earthquake Science Center on the University of Washington campus let this through review. A casual ten minutes spent surfing around in Google Earth would surely have raised questions about the nature of Angster's "scarp". Internal reviewers should have terminated his trenching proposal early on. Instead of, "Find a better prospect, young man," Angster was told, "Load up the backhoe." Local geologists were invited to the site after the trench had already been opened, rather than during the planning stages. Many thousands of tax payer dollars were wasted.

Publish or Abandon

At this point it is unclear what Angster and Sherrod will do. Will they a.) publish an article on their work as it stands, b.) submit a heavily-revised version for peer review, or c.) exclude the Starthistle Trench from regional seismic hazard syntheses altogether? I'm guessing the SUK-Starthistle trench findings, cleverly de-emphasized, will make their way into the next paleoseismic report.

What am I missing?

Your comments are always welcome:



Angster, S.J.; Sherrod, B.L.; Lasher, J.P., 2020, Constraining the paleoseismic record on the Wallula fault from widespread liquefaction features in southeastern Washington: Nice scarp, where's the fault...?, Link to AGU poster below:

Angster_AGU2020_Wallula Fault and Liquifaction_D3_FINAL
Download PDF • 1.02MB

Angster, S.J.; Sherrod, B.L.; Lasher, J.P., 2023, Logs and data from the Starthistle Trench across a scarp within the Wallula Fault Zone, southeastern Washington, USGS Scientific Investigations Map 3495



Brocher, T.M.; Sherrod, B.L., 2018, Intensities, aftershocks, and location of the 1936 Milton‐Freewater Earthquake near the Oregon–Washington Border, USA Intensities, Bulletin of the Seismological Society of America, v. 108, p. 2594-2613

Courtney Flatt, "In southeastern Washington, geologists dig into the past to find fault signs of future shaking", Northwest Public Broadcasting (December 6, 2019)

Mullineaux, D.R., 1996, Pre-1980 tephra-fall deposits erupted from Mount St. Helens, WA, USGS Professional Paper 1563

YouTube: Stephen Angster talks to Wenatchee Valley Erratics - December Meeting 2021

Washington Standard, 11 January 1873:

According to the Olympia-based Washington Standard newspaper, shaking caused by the 1872 quake centered near Entiat - Eastern Washington's largest on record - was "insignificant" to people of Puget Sound. Fabulous newspaper geology at its best.

It appears that our earthquake experience…although it awakened considerable interest in the future state, was insignificant compared to that of our neighbors east of the mountains, who were forced to believe at the time that the end of all things sublunary had indeed come.

A letter from Klikitat county says that the earthquake...was very violent in that vicinity, but did no damage. The writer…formerly of this county, gives a very amusing account of the conduct of [our reporter] Mr. Shazer at the time. Greatly excited he sprung from bed, and ran out to his chicken-coop, and soon returned with the gratifying information that the chickens were all safe! Earthquakes will never injure such men.


Last 50 Posts
All Posts by Month
    bottom of page