ACTIVE FAULTING ALONG A NEWLY FOUND SEGMENT OF THE SADDLE MOUNTAIN FAULT ZONE, SOUTHEASTERN OLYMPIC MOUNTAINS, WASHINGTON: A PALEOSEISMIC TRENCHING STUDY Jessica L. Czajkowski1, Timothy J. Walsh1, Trevor A. Contreras1, Kelsay Davis-Stanton1, Harvey M. Kelsey2, Elizabeth R. Schermer3, Robert J. Carson4

(1) Washington Division of Geology and Earth Resources, Olympia, WA 98504; (2) Department of Geology, Humboldt State University, Arcata, CA 95521; (3) Geology Department, Western Washington University, Bellingham, WA 98225; (4) Department of Geology, Whitman College, Walla Walla, WA 99362 LOCATION MAP

Concrete CANADA USA The Alligator trench was excavated across a LiDAR-detected, southeast-facing Holocene- Preliminary age estimates of a charcoal fragment from beneath this layer suggest burial after Lilliwaup Swamp

1200 1100 age fault scarp on a steep hillside near Price Lake in Mason County (Fig. 1). The scarp is 3,370 +/- years BP. The contacts between the bedrock, diamict and colluvium and the shear Port Angeles 1300 D Figure 4A - 4D: Equal-area stereonet projections of fault data U Forks Everett visible for approximately 0.5 miles from the top of Dow Mountain northeastward down the fabric within the diamict do not show the same amount of rotation into the fault, suggesting 1400 Alligator hillside (Photos A—C), located southeast of and roughly parallel to the Saddle Mountain the clasts were preferentially oriented during deposition. Farther from the fault surface what P Study A CI area trench Bremerton Seattle A) Active fault surface (poles to planes) B) Active fault surface slickenlines C) Bedrock fracture surfaces (poles to planes) D) Bedrock fracture slickenlines Miller F East and West fault scarps. is interpreted to be slope colluvium (4) does not show a preferred orientation of clasts, and I OCEC North Bend 1600 Pond is much less indurated. The age of burial, if interpreted to be caused by faulting, has not Active fault surface planes Active fault surface slickenlines Bedrock fracture surface planes Bedrock fracture slickenlines A

N Tacoma Mean plane girdle Girdle Mean plane girdle Girdle Cle The trench was approximately 20 m long, although only 16 m was logged (Fig. 2). The been seen elsewhere in the Puget Sound. Dow Elum Aberdeen Olympia Mountain trench contained two benches approximately 1 m wide on the uphill side for safety, with 1300 SMW SME 1800 BST Centralia D Chehalis over 6 m of apparent vertical stratigraphic exposure. However, the down-slope dip of units Directly above unit 4a in the footwall of the fault is a 3 to 5 cm thick soil horizon (4as), that Saddle Mountain west fault U Morton projected into the trench walls yielded less true stratigraphic exposure, and some contacts contains abundant charcoal. The presence of colluvium (4b) above this soil indicates a fault- 1700 Raymond were duplicated as a result of benching the face. (see Trenching Profiles; Fig. 3). ing event buried soil 4as. Preliminary age estimates from charcoal sampled from this layer Kelso Price Longview (AT-2) suggest burial occurred sometime after 1,360 +/- 40 BP. 1800 Lake PL The trench was logged on three 1 x 1 m photomosaic sections of trench wall. Stratigraphic CC units were numbered by interpreted age, oldest to youngest. Above unit 4b, there is another possible soil horizon (4bs?), although weakly developed, which lies beneath more colluvium (4c?). The layer contained abundant charcoal, yet some The trench exposed a southwest-striking, steeply-dipping fault which offsets bedrock samples had ambiguous origin (i.e. possibly modern). If soil 4bs? does not exist, colluvium No. of data = 9 No. of data = 8 No. of data = 23 No. of data = 6 AT Mean principal orientation = 236/79 Mean principal direction = 1-063 Mean principal orientation = 214/86 Mean principal direction = 20-218 Lilliwaup (Eocene Crescent Formation basalt), glacial deposits, soil horizons, and colluvium deposits. 4c? and 4b are derived from the same faulting event. Both colluvium 4b and 4c? are faulted. DM Mean resultant direction = 79-326 Mean resultant direction = 328/38 Mean resultant direction = 54-305 Mean resultant direction = 145/59 The vertical component of bedrock throw across the fault was measured at approximately Mean resultant length = 0.99 Mean resultant length = 0.59 Mean resultant length = 0.40 Mean resultant length = 0.61 (Variance = 0.01) (Variance = 0.41) (Variance = 0.60) (Variance = 0.39) 3.2 m. No piercing points were discovered to yield the net throw. From slickenlines mea- Weak soil development (4cs?) at the top of units marked 4c? and 4b, parallel to the fault U Calculated. girdle: 10/16 Calculated. girdle: 243/85 Calculated. girdle: 338/8 Calculated. girdle: 213/75 D sured on both the active fault surface and within the fractured bedrock surfaces in the hang- scarp surface show the ground surface before the last interpreted faulting event. The collu- Calculated beta axis: 74-280 Calculated beta axis: 82-248 Lake ? Cushman ? ingwall, the direction of fault movement is uncertain, although data is highly suggestive that vial wedge (4d?) from this event appears to be derived from faulting as well as slope wash, older movement on the fault (measured on fracture surfaces in bedrock adjacent to the and fills a fault-parallel stream channel cut into earlier fault-derived colluvial deposits and Saddle Mountain east fault active fault surface) was oblique reverse (Figs. 4C and 4D). Holocene-age movement soil horizons. The stream channel was identified based upon a prism of thinly bedded allu- (measured on the clay-lined active fault surface) was sub-horizontal to slope-parallel (Photo vium (4da?) at the base of a channel-shaped feature present on both uphill and downhill UNIT DESCRIPTION LEGEND D and Figs. 4A and 4B), which implies left-lateral strike-slip offset when coupled with the sides of the trench with a downslope axis. The colluvial fill (4d?) within the channel con- vertical sense of bedrock offset. However, the scarp facing direction implies right-lateral tained abundant cobbles and pebbles at the base of the deposit on the northwest side of the offset, making the determination of slip direction problematic. Further investigation is nec- channel, fining to the southeast, implying the source for the material was predominantly 5 Modern soil GS-1 Grab sample location Contact, approximately located essary to resolve this. from the fault scarp and not upslope. 4 Slope colluvium Projected location of radiocarbon sample Contact, concealed The relationships between faults, colluvium and soil development within the trench give It is interesting to note that few angular clasts of Eocene Crescent Formation Basalt are strong evidence for multiple (possibly four) Holocene-age faulting events on this newly found within colluvium deposits in the footwall of the fault. This suggests that bedrock was 4a Event colluvium found segment of the Saddle Mountain fault system, although there are potentially several never exposed in the topographic scarp, and faulting-related colluvium was predominantly AT-1 Radiocarbon sample location ? ? ? Contact, concealed/queried 0.5 mi 4as Buried soil possible faulting scenarios and sequences. Age-date testing of charcoal samples obtained derived from the overlying diamicton and slope colluvium in the hangingwall. from lower soil horizons (Fig. 2) confirm at least two events. Future testing will hopefully 4b Event colluvium AT-1 Radiocarbon sample with preliminary lab results Benches in trench walls Figure 1. Lidar image of fault scarps, mapped faults, and paleoseismic trenches in the Price Lake area, Mason County. CC = further constrain the younger faulting history REFERENCES Fault; dashed where inferred; Trench profile locations at Cargill Creek trench (Witter and others, 2008), DM = Dow Mountain trench, PL = Price Lake trench, SMW = Saddle Mountain 4bs? Possible buried soil horizon TP-1 arrows show apparent relative motion vertical grid coordinates west trench, SME = Saddle Mountain east trench (Wilson, 1975), BST = Banana Slug trench (E. A. Barnett, USGS, personal The variable thickness of the compact diamicton—interpreted to be lodgment till of the Wilson, J. R., 1975, Geology of the Price Lake area, Mason County, Washington: North 4c? Event colluvium communication, 2008), and AT = Alligator trench (this study). Inset map shows location of the Alligator trench on the southeast- NW Vashon stade—suggests that deposition of the diamicton occurred after the scarp was par- Carolina State University Master of Science thesis, 79 p., 2 plates. Contact facing scarp (approximately 2.5 m tall), and the extent of the fault scarp along the southern crest of Dow Mountain, where it is tially developed. The presence of preferentially-oriented cobbles and boulders within collu- Witter, R. C.; Givler, R.W.; and Carson, R.J., 2008, Two post-glacial earthquakes 4cs? Buried soil 7 approximately 4 to 5 m tall with the same facing direction. vium directly above a charcoal-rich surface above the compact diamicton adjacent to the on the Saddle Mountain west fault, southeastern Olympic Peninsula, Washington: fault surface implies that either the colluvium (4a) was deposited in response to a faulting Bulletin of the Seismological Society of America (December 2008), 98(6):2894- 4da? Buried channel alluvium Large clast event, or that faulting events after its deposition realigned clasts in portions of the deposit. 2917.

4d? Event/slope colluvium 5 GS-4-1 4 3 Ancient ground surface above compact TRENCH PROFILES diamicton with abundant charcoal 6 3 2 TP-1 AT VERTICAL GRID COORDINATE 3,0 2 Compact diamicton NE SW Figure 2 (below): Alligator trench log across fault scarp. 2a 5 2a Water-deposited glacial deposits within compact diamicton 4 Figure 3 (left): Trench profiles TP-1 (top) and TP-2 (bottom), located along 1 Eocene Crescent Formation basalt 5 3 vertical green dashed grid lines shown on Figure 2. Benching 4 of the trench face produced an apparent duplication of strat- 2 1b Cataclastic material derived from 4cs? 3 igraphy, due to the downslope dip of contacts. Eocene Crescent Formation basalt 4d? AT-16 GS-4cs GS-4d 4 2 AT-14 ELEVATION (m) SE 2 TP-1 TP-2 AT-15 GS-4c 4 4c? 4 1 16 4 5 4bs? GS-4bs 3 AT-7 GS-1c 4 AT-4 B 1 B) View of the scarp at the base of the treeline in the upper right 3 1 GS-4b A portion of the photo (red dashed line). Scarp height near trench 2 5 4b ? ? (TM) location was approximately 2.5 m. Photo by Trevor Contreras. ? A) Google Earth oblique air photo of the southeast- 0 3 ? 3 facing unnamed fault scarp, visible along treeline in fresh 0 1 2 3 4 5 6 7 ? 13 14 15 DISTANCE (m) 4d? logging cut. View to the north. AT-10 4 AT-8 AT-9 5 AT-2 AT-3 AT-1 4as GS-4as C) Photo of geologists standing on top of the fault TP-2 AT VERTICAL GRID COORDINATE 7,0 NE SW 5 scarp. Viewed to the southwest. Photo by Trevor 4d? 4a GS-4-2 Contreras. GS-4a 1,360 +/- 40 BP 2 2 4da?

DISTANCE (m) 2 2 3 4a 5 GS-4? AT-5 2 AT-11 4d? ? AT-12 AT-6 ? 3 AT-13 3 4 3 5 4d 3 2 2 3 2 4da? GS-4al 3,370 +/- 40 BP 2 Alligator lizard, trench namesake 4da 1 GS-2 1 1 2 ELEVATION (m) 4d 4 3 2 1 ACKNOWLEDGMENTS 4d 1 1 1 D) Sub-horizontal GS-1a GS-1b 1 slickenline within clay- 4da 2 1a We would like to thank both Elizabeth A. Barnett and Brian 4 lined active fault surface. 1 Sherrod of the U.S. Geological Suvey (USGS), who funded 3 2 Photo by Elizabeth 1 this research and generously provided field assistance and D 0 C Schermer. 0 1 2 3 4 5 6 0 0 useful commentary. We also benefited from insightful field 1 2 3 4 5 6 7 8 9 10 11 12 DISTANCE (m) 0 visits by Isabelle Sarikhan (DNR), Ray Wells (USGS), and DISTANCE (m) Don West and Dave Finley from Golder Associates.