Precipitation, landsliding, and erosion across the Olympic Mountains, Washington State, USA

    Precipitation, landsliding, and erosion across the Olympic Mountains, Washington State, USA Stephen G. Smith, Karl W. Wegmann PII: DOI: Reference:

S0169-555X(17)30431-2 doi:10.1016/j.geomorph.2017.10.008 GEOMOR 6194

To appear in:

Geomorphology

Received date: Revised date: Accepted date:

23 January 2017 11 October 2017 12 October 2017

Please cite this article as: Smith, Stephen G., Wegmann, Karl W., Precipitation, landsliding, and erosion across the Olympic Mountains, Washington State, USA, Geomorphology (2017), doi:10.1016/j.geomorph.2017.10.008

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Precipitation, landsliding, and erosion across the Olympic Mountains, Washington State, USA

Department of Marine, Earth, and Atmospheric Sciences, North Carolina State University, Box 8208,

RI

a

PT

Stephen G. Smitha,b and Karl W. Wegmanna

SC

Raleigh, NC, 27695, USA b

Center for Integrative Geosciences, 354 Mansfield Road, Unit 1045, University of Connecticut, Storrs,

NU

CT, 06269, USA

MA

Corresponding author: Stephen G. Smith; email address: [email protected]

TE

D

Abstract

AC CE P

In the Olympic Mountains of Washington State, landsliding is the primary surface process by which bedrock and hillslope regolith are delivered to river networks. However, the relative importance of large earthquakes versus high magnitude precipitation events to the total volume of landslide material transported to valley bottoms remains unknown in part due to the absence of large historical earthquakes. To test the hypothesis that erosion is linked to precipitation, approximately 1000 landslides were mapped from Google Earth imagery between 1990 and 2015 along a ~15 km-wide x ~85 km-long (1250 km2) swath across the range. The volume of hillslope material moved by each slide was calculated using previously published area–volume scaling relationships, and the spatial distribution of landslide volume was compared to mean annual precipitation data acquired from the PRISM climate group for the period 1981–2010. Statistical analysis reveals a significant correlation (r = 0.55; p < .001) between total landslide volume and mean annual precipitation, with 98% of landslide volume occurring along the windward, high-precipitation side of the range during the 25-year interval. Normalized to area,

ACCEPTED MANUSCRIPT this volume yields a basin-wide erosion rate of 0.28 ± 0.11 mm yr-1, which is similar to previous timevariable estimates of erosion throughout the Olympic Mountains, including those from river sediment 10

Be, fluvial terrace incision, and thermochronometry. The lack of large historic

PT

yield, cosmogenic

earthquakes makes it difficult to assess the relative contributions of precipitation and seismic shaking to

RI

total erosion, but our results suggest that climate, and more specifically a sharp precipitation gradient,

SC

plays an important role in controlling erosion and landscape evolution over both short and long

NU

timescales across the Olympic Mountains.

MA

Keywords

AC CE P

TE

D

Olympic Mountains; Olympic Peninsula; Landslides; Erosion rates; Climate; Precipitation

ACCEPTED MANUSCRIPT

PT

1. Introduction

The Olympic Mountains, located in the state of Washington, are situated along an active plate boundary

RI

in a region that receives abundant precipitation. Previous study of the Olympic Mountains has explored

SC

the roles of both tectonics and climate in the topographic and structural evolution of this forearc high

NU

(Willett, 1999; Montgomery and Greenberg, 2000; Montgomery, 2001; Montgomery and Brandon, 2002; Stolar et al., 2007), but the relative contribution of each mechanism to the surface response of the

MA

landscape remains unknown. The interplay among climate, tectonics, and mountain belt evolution is a research topic that has garnered much attention and debate in the 21st century (e.g. England and

D

Molnar, 1990; Burbank et al., 1996; Montgomery et al., 2001; Riebe et al., 2001; Bonnet and Crave,

TE

2003; Whipple, 2009; Binnie et al., 2010; Korup et al., 2010; DiBiase and Whipple, 2011; Ferrier et al.,

AC CE P

2013; Gasparini and Whipple, 2014), and although it is accepted that precipitation facilitates erosion in mountainous regions, deciphering the relative contribution of precipitation to overall rates and spatial patterns of erosion has proven difficult. Much of this difficulty likely arises due to complexity inherent in erosional processes across various spatial and temporal scales as well as inconsistencies in feedback cycles between climate, tectonics, and erosion across different regions of the globe. In addition to studies addressing mountain belt evolution over long (> 106 yr) timescales (e.g. Willett, 1999; Montgomery et al., 2001; Bishop, 2007; Whipple, 2009), recent research has also sought to better understand the relationship between climate and tectonics over much shorter timescales by attempting to constrain the feedbacks between rainfall, earthquakes, and erosion during and after singular meteorological or seismic events (e.g. Dadson et al., 2004; Guzzetti et al., 2004; Yanites et al., 2010; Hovius et al., 2011; Parker et al., 2011; Tang et al., 2011; Lira et al., 2013; Marc et al., 2015). These studies offer valuable information on the erosional response of particular areas to individual geologic

ACCEPTED MANUSCRIPT phenomena, but nevertheless researchers have yet to reach to reach an empirical consensus regarding

PT

the net effect of climate on erosion rates.

Previous studies have shown that the Olympic Mountain portion of the Cascadia forearc high is in a flux

RI

steady-state (Willett and Brandon, 2002) where rock exhumation, surface uplift, and erosion rates

SC

increase gradually from the Pacific Ocean toward the central massif and are balanced on >10 5 yr time

NU

scales (Brandon et al., 1998; Pazzaglia and Brandon, 2001). Over shorter timescales, less is known about patterns of erosion, with research limited to river incision and 10Be-derived catchment-averaged erosion

MA

rates from the Clearwater River basin (Wegmann and Pazzaglia, 2002; Belmont et al., 2007). Furthermore, since the last full-margin rupture of the Cascadia Subduction Zone occurred more than

D

300 years ago (Satake et al., 1996; Yamaguchi et al., 1997), the impact of large earthquakes on the

TE

surface processes of individual catchments within the Olympic Mountains also remains unclear. Based

AC CE P

on globally observed effects of seismic shaking in high-relief mountainous terrain, it is assumed that the AD 1700 event resulted in widespread slope failures that may have imparted a decadal-scale impact on fluvial networks and associated ecosystems (e.g. Dadson et al., 2004; Yanites et al., 2010; Hovius et al., 2011; Parker et al., 2011; Tang et al., 2011), but this assumption is unproven and it is unknown how the sediment flux from such a singular event would compare to the flux derived from, for instance, the 668 landslides inventoried in the Quinault River catchment of the Olympic Mountains during the seismically quiescent period 1939 to 1998 (Quinault Indian Nation, 1999). As a result, the relative roles of seismicity and precipitation in driving the long-term pace of erosion is a fundamental, yet unanswered question crucial for understanding mountain belt evolution and patterns of sediment flux for this and other mountain ranges in temperate, tectonically-active regions.

ACCEPTED MANUSCRIPT Here we utilize a ~1250 km2 swath across the Olympic Mountains that is oriented parallel to both tectonic (exhumation and uplift) and precipitation gradients of the range as a natural laboratory to test

PT

the following two hypotheses: 1) that the volume of hillslope rock and regolith displaced by landslides mimics patterns of precipitation across the range; and 2) that erosion via aseismic landsliding is a

RI

significant contributor to the long-term rate of erosion. We discuss our remote sensing results relative

MA

NU

upland response to past climatic and/or seismic events.

SC

to previous research conducted within the Olympic Mountains as well as field evidence for possible

D

2. Geology and climate of the Olympic Mountains

TE

The Olympic Mountains represent the sub-aerial portion of the accretionary wedge formed as a

AC CE P

consequence of oblique subduction of the Juan de Fuca plate beneath the North American plate across the Cascadia margin. Around 15 Ma, a decrease in accommodation space led to uplift and exposure of Olympic Mountain segment of the subduction wedge above sea level (Brandon et al., 1998). At the surface, the Olympic Mountains consist primarily of the Olympic subduction complex (OSC), which is characterized by highly deformed, stratigraphically discontinuous, and partially metamorphosed sedimentary rocks (Fig. 1; Tabor and Cady, 1978; Brandon et al., 1998). Juxtaposed against the rocks of the OSC is a horseshoe-shaped surface exposure of older oceanic basalts situated on the footwall of the eastward-dipping Hurricane Ridge thrust fault (Tabor and Cady, 1978; Brandon and Calderwood, 1990; Babcock et al., 1994). These basalts, which make up the Eocene Crescent Formation (ECF), form a border along the northern, eastern, and southern sides of the range.

ACCEPTED MANUSCRIPT As the Juan de Fuca plate continues to converge with the North American plate at 36 mm yr-1, rates of uplift increase landward from ~0.3 km Myr-1 near the Pacific Coast of Washington to ~1 km Myr-1 at the

PT

summit of Mount Olympus, which, at an elevation of 2430 m (7980 ft), is the highest point in the Olympic Mountains (Fig. 1; Brandon et al., 1998; DeMets and Dixon, 1999). Previous research has shown

RI

that net uplift, or material influx into the range, is balanced by the processes of denudation, or material

SC

flux out of the range, and thus the Olympic Mountains currently maintain a flux steady state in which

NU

there is no net addition or loss of mass over geologic timescales (Pazzaglia and Brandon, 2001; Willett

MA

and Brandon, 2002).

The surficial geology of the Olympic Mountains is dominated by the effects of glacial and fluvial erosion

D

into an uplifting massif and the formation of threshold hillslopes whose gradients are maintained by

TE

both shallow and deep-seated mass wasting (Montgomery, 2001). The Olympic Mountains are just

AC CE P

south of the maximum extent of Cordilleran Pleistocene ice sheets, but the valleys draining the central massif were repeatedly filled with alpine glacial ice that facilitated the carving of broad U-shaped valleys with steep bedrock valley walls (Thackray, 2001; Dragovich et al., 2002). Modern alpine glaciers, which have been retreating rapidly, are restricted to the highest elevation cirque basins and to the Mount Olympus massif. Hillslope processes throughout the Olympic Mountains are dominated by bedrockinvolved landslides, debris flows, and colluvial creep down the ubiquitous steep hillslopes (Reid and Dunn, 1984; Swanson et al., 1987). Hillslope soils are rocky and thin. Shallow landslides initiating in residual soils near ridge-crests often transition into erosive debris flows that deliver sediment onto Holocene alluvial fans at the base of the steep valley sides or directly into tributary streams and axial river channels (Reneau et al., 1989; Wegmann and Pazzaglia, 2002; Benda et al., 2003). Deep-seated bedrock landslides occur both along the steep valley walls and at lower evaluations within glacio-fluviallacustrine units destabilized by post-glacial stream incision (Quinault Indian Nation, 1999; Gerstel, 1999).

ACCEPTED MANUSCRIPT At the highest elevations, gravitational spreading of steep-sided ridges (sackung) is common where sedimentary units of the OSC dip steeply towards the valley floor below (Tabor, 1971). Hillslope surficial

PT

geology and geomorphic processes vary little from the windward to leeward side of the range.

RI

Geographically, the Olympic Mountains constitute a large portion of the Olympic Peninsula and are

SC

unique in the United States for the combination of rugged, glacially-sculpted topography and large

NU

swaths of temperate rain forest (Fig. 2). Rainforest conditions exist predominantly on the western, windward side of the range due to orographic forcing as moisture-laden, land-falling Pacific storms

MA

encounter the steep, mountainous terrain (e.g. Minder et al., 2008; Neiman et al., 2011; Gavin and Brubaker, 2015). The bulk of precipitation, which can exceed 5 m annually in the higher-elevation

D

western side of the range (Fig. 3), falls during the winter season as the Aleutian low pressure system

TE

intensifies and drifts south from the Gulf of Alaska, subsequently steering moisture and winds toward

AC CE P

the Pacific Northwest (PRISM Climate Group, 2015; Gavin and Brubaker, 2015). While snow in lowland areas is very rare, during cold storms a significant amount of precipitation may fall as snow atop high ridges (Minder et al., 2008). Modeling of orographic precipitation patterns across the range show that deposition of snow at high elevations has a minor role in rainfall distribution, but that cold storms with low freezing levels can lead to increased leeward precipitation produced via downwind advection of frozen hydrometers generated in the orographic cloud (Zängl, 2007; Minder et al., 2008). The typically cool, rainy winters transition to warm and dry summers once the Aleutian low weakens and the North Pacific high pressure system strengthens and brings northwesterly air flow, low humidity, and sunny skies to the region (Ware and Thomson, 2000; Patterson et al., 2013).

3. Methods

ACCEPTED MANUSCRIPT

The study area encompasses a ~1250 km2 NW–SE swath across the entirety of the Olympic Mountains

PT

that was subdivided into a grid of 40 blocks of roughly equal area (~30 km 2) for comparison between landsliding and precipitation (Fig. 2). Within each block, individual landslides were mapped from 1990 to

RI

2015 using all satellite imagery available in Google Earth (Fig. 4), which included full coverage of the

SC

study area for the years 1990, 1994, 2005, 2006, 2009, 2011, 2012, 2013, and 2015. For the purposes of

NU

this study, a “landslide” consists of a large (area ≥ 5000 m2) mass movement with a well-developed rupture surface. We considered neither the failure rate nor transport distance of landslide material, as

MA

these factors are irrelevant to the purpose of the study. For instance, slumped material that has not been carried far from its point of origin was mapped as a complete landslide since it is now more readily

D

available for transport throughout the channel network. Landslide borders were carefully drawn to

TE

avoid grouping multiple scars as a single unit; flow paths, chutes, and other features outside of the main

AC CE P

scar were not mapped as part of the total landslide area. Potential landslides in high elevation areas with minimal-to-no vegetation were excluded from the assessment due to difficulties in identifying individual slide boundaries. Landslides affiliated with anthropogenic activity, (e.g. initiating along logging roads) were also excluded from the study. Mapped landslides were exported from Google Earth into ArcMap 10.3, where individual landslide areas and volumes were calculated, using the area-volume scaling parameters reported in Larsen et al. (2010) (Fig. 4). Based on field observations and satellite imagery, landslides with areas ≥ 5000 m2 were considered deep-seated bedrock landslides, whereas those with areas < 5000 m2 were categorized as shallow failures consisting predominately of soil and vegetation. A total landslide volume was calculated for each survey block (1–40), and volumes were normalized to block area to correct for minor differences in block size. Uncertainties in the normalized volumes include an allowed 15% mapping error for individual landslides as well as reported standard deviations (σ) for α and ϒ in the formula VL = αAϒ, where VL is landslide volume, A is landslide area, and α

ACCEPTED MANUSCRIPT and ϒ are empirically-derived constants used to estimate the volume of a landslide based on its surface area. For bedrock slides, σα = 0.03 and σϒ = 0.02, and for shallow soil slides σα = σϒ = 0.005 (Larsen et al.,

PT

2010). The total volume uncertainty is dependent upon the landslide size distribution within each grid block; when total volume is spread among many similarly-sized landslides, uncertainty is reduced as

RI

individual landslides are unlikely to be biased in the same direction, whereas when the total landslide

NU

less likely to be balanced (e.g., Emberson et al., 2016).

SC

volume of a grid block is dominated by one or several large landslides, uncertainty is high as errors are

MA

Grid block averages for the modeled 30-year mean annual precipitation (MAP) data (1981–2010; PRISM, 2015) were compared to total landslide volume for each block of the study area. The spatial distribution

D

of precipitation is a community model output of the PRISM Climate Group, and data covering the entire

TE

Olympic Peninsula were obtained in ASCII format and imported to ArcMap, where MAP was averaged

AC CE P

for each grid block (Figs. 2 and 3) (Table 1). The relationship between landslide volume and MAP was assessed via a linear regression model using MATLAB in which statistical significance was determined by the results of a permutation test with 107 iterations. Mean elevation and mean slope were also calculated for each grid block using the USGS 10 m digital elevation models (DEMs), and their relationship to precipitation was assessed via the same linear regression model in MATLAB.

A 25-year erosion rate for the study area was calculated by isolating landslides in the study area that occurred during the years 1990–2015 (~250 landslides; Fig. 4). These landslides were identified by viewing the 1990 imagery for each block of the study grid and progressively stepping through each later image. Any landslide that was not present in the 1990 imagery, but was identified in a later image, was included in the 25-year erosion rate calculation. The timespan between images in 1994 and 2005 requires that a landslide occurring just after the 1994 imagery still be readily identifiable 11 years later,

ACCEPTED MANUSCRIPT which is a valid assumption given that landslide scars present in the 1990 imagery persist through the most recent imagery (in 2015) used for the study. The total volume and associated erosion rates derived

PT

from these landslides were partitioned into “high” and “low” precipitation regions of the study area with a cutoff of 2.5 m yr-1 in which corresponding to blocks 1-18 and 20-21 receive “high” (≥ 2.5 m yr-1)

RI

precipitation whereas blocks 19 and 22-40 receive “low” (< 2.5 m yr-1) precipitation. We chose 2.5 m yr-1

SC

as the cutoff value since it corresponds to the minimum rate of precipitation received by all windward

NU

areas of the Olympic Mountains and allows for a simple, yet straightforward comparison of landsliding

MA

and erosion between the two sides of the orographic precipitation gradient.

TE

D

4. Results

AC CE P

The 25-year inventory includes 937 mapped landslides that encompass an area of 4.5×106 m2 with an estimated volume of 2.0×107 m3 (Table 1). Of the 937 landslides mapped, 655 occurred in the “high” precipitation region of the study area, whereas 282 landslides occurred in the “low” precipitation region of the study area. MAP and landslide volume are significantly positively correlated (p = 0.0002; Fig. 5) via the relationship

, where the values used for landslide volume (VL(n)) have

been normalized to grid block area by dividing the total landslide volume (m3) in any given grid block by the corresponding area of that grid block (m2). In this manner, VL(n) has units of m and can be thought of as a vertical thickness of rock and regolith that is transported via landsliding in each grid block. MAP is measured as m yr-1, and the uncertainty in the slope is 1.5×10-5 ± 3.7×10-6.

Landslide volume is also positively correlated to mean slope (p = 0.008), but this correlation is mainly controlled by grid blocks 37–40 that, relative to the entire dataset, exhibit both comparatively low mean

ACCEPTED MANUSCRIPT slopes and low landslide volumes. When these grid blocks are removed, there is no correlation (p = 0.9) between landslide volume and mean slope, but landslide volume and MAP maintain a strong positive

PT

correlation (p = 0.002). There is no correlation between landslide volume and mean elevation (p = 0.8).

RI

For the years 1990–2015, total estimated landslide volume for the study area is 4.7×106 m3 (0.0047

SC

km3). Calculating erosion rate (Re) via the equation Re = VLA-1t-1, where VL is total landslide volume, A is

NU

the area considered, and t is the corresponding time interval (25 years) yields a basin-wide erosion rate of 0.15 ± 0.06 mm yr-1 (Table 2). Nearly all of this volume (98% or 4.6 x 106 m3) is derived from the

MA

windward, high precipitation side of the range (Fig. 6; Table 2). Calculating separate Re for high and low precipitation regions of the range via the same equation reported above results in an area-wide surface

D

lowering rate of 0.28 ± 0.11 mm yr-1 for the high precipitation western flank of the range and 0.005 ±

5. Discussion

AC CE P

TE

0.002 mm yr-1 for the leeward “low precipitation” region.

Based on prior studies of post-landslide vegetation recovery (e.g. Guariguata, 1990; Smale et al., 1997; Francescato et al., 2001; Chou et al., 2009), the temperate maritime climate of the Olympic Peninsula, and observations made from satellite imagery of the study area spanning 1990–2015, we are confident that the majority of mapped landslides originally present in the 1990 imagery likely occurred during the past century, and that landslides potentially triggered by the last rupture of the Cascadia subduction zone in AD 1700 have long been revegetated and rendered mostly unidentifiable via imagery-based remote sensing techniques. As a result, we consider the volume of hillslope material moved by the mapped landslides to reflect aseismic climate-driven erosion in the Olympic Mountains, as there have

ACCEPTED MANUSCRIPT been no earthquakes large enough or in close enough proximity (c.f. Keefer, 1984) to generate coseismic landsliding on slopes of the study area during the period of seismic instrumentation (Pacific Northwest

PT

Seismic Network, 2016). In the absence of seismic shaking, regional landslides are the probable result of intense precipitation or rapid snow melt contributing additional moisture to wet-season near-saturated

RI

regolith on hillslopes, thereby decreasing the effective normal stress and leading to slope instability (e.g.

SC

Swanson et al., 1986; Gerstel, 1999; Quinault Indian Nation, 1999; Benda et al., 2003; Wegmann, 2004).

NU

The significant correlation between landslide volume and MAP supports this conclusion, as more and larger landslides are found in portions of the Olympic Mountains that receive greater amounts of rain

MA

and snowfall. Given that landsliding is the dominant mode of sediment transport from hillslopes to fluvial networks in mountainous environments (Hovius et al., 1997; Hovius et al., 2000; Korup et al.,

D

2004; Korup et al., 2010; Wenske et al., 2012; Bennett et al., 2013), this suggests that the windward,

TE

wetter side of the Olympic Mountains erodes more quickly than the leeward, drier side during periods

AC CE P

of seismic quiescence. This result is consistent with the pattern of long-term exhumation and rates of river incision from the range (e.g, Brandon et al., 1998; Pazzaglia and Brandon, 2001).

Calculating the 1990–2015 erosion rate from landslides allows for an assessment of climate-driven erosion with respect to long-term estimates of Olympic Mountain landscape evolution. A rate of 0.28 ± 0.11 mm yr-1 for areas of high precipitation is similar to those calculated for the Clearwater River basin using cosmogenic

10

Be that integrate the catchment-averaged signal of erosion over the past 1.5 to

4×103 yrs depending on the rate associated with each individual sample (Belmont et al., 2007). This timespan includes several great earthquake events identified in paleoseismic studies focused on the CSZ (e.g. Atwater, 1987; Atwater, 1992; Atwater and Hemphill-Haley, 1997; Atwater et al., 2003; Kelsey et al., 2005; Nelson et al., 2006; Goldfinger et al., 2011; Goldfinger et al., 2012; Graehl et al., 2015), and therefore the comparable rate of modern aseismic erosion suggests that the spatial distribution of

ACCEPTED MANUSCRIPT precipitation has plays a significant role in setting the pace of late Holocene erosion. The 0.28 ± 0.11 mm yr-1 rate is also comparable to short term (annual-to-decadal) estimates of erosion based on Hoh River

PT

sediment yield (Nelson, 1986) as well as studies of longer-term Olympic Mountain evolution, including a ~140 ka record of river incision derived from Clearwater River strath terraces, the spatial pattern of rock

RI

exhumation derived from Zircon and Apatite fission track and U-Th/He thermochronology across the

SC

range, and estimates of erosion based on slope and mean local relief for a 10 km-wide swath across the

MA

Brandon, 2001; Montgomery and Brandon, 2002).

NU

range parallel to the Cenozoic vector of tectonic convergence (Brandon et al., 1998; Pazzaglia and

The 0.28 ± 0.11 mm yr-1 landslide-derived rate is, in fact, nearly equivalent to the mean long-term

D

exhumation rate of ~0.28 mm yr-1 for the entire Olympic Peninsula (Brandon et al., 1998), which

TE

provides compelling evidence that climate exerts a major control on the pace of erosion. Nevertheless,

AC CE P

this conclusion also raises questions about the nature of erosion on the leeward side of the range. Although every grid block assessed in this “low precipitation” portion of the range has a value of MAP that exceeds the U.S. national average (PRISM Climate Group, 2015; NOAA, 2016), aseismic landslide volume is limited and appears to have a negligible contribution to overall erosion. One possible explanation for this negligible contribution is that since slope failures in the “low precipitation zone” are not frequently triggered by local climatology and hydrology, more slopes may instead remain close to the critical threshold of stability over longer periods of time and thus fail more rapidly in the event of an earthquake as compared to slopes in the “high precipitation zone.” In this hypothesized scenario, relatively fewer landslides would occur throughout the windward side of the range since a lower percentage of slopes would be at or near the critical threshold, and thus the contribution of coseismic landslide volume to the long-term rate of erosion would be comparatively greater for the leeward side

ACCEPTED MANUSCRIPT of the range. This idea may be tested in the event of future earthquake impacts in this or other similarly

PT

mountainous regions having sharp precipitation gradients, such as the South Island of New Zealand.

However, confounding the above scenario is a ~0.17 mm yr-1 erosion rate derived for this study using

RI

sediment volume (1.56×107 m3) impounded behind the former Glines Canyon dam that was on the

SC

Elwha River from 1927–2012 (Bountry et al., 2011). This rate was calculated by assuming a bulk reservoir

NU

sediment density of 1.5 g cm-3 (e.g. Avnimelech et al., 2001; Audry et al., 2004; Snyder et al., 2004; Lazzari et al., 2015), an average rock density of 2.7 g cm-3, and dividing by the total upstream drainage

MA

area (632 km2) and the reservoir in-fill time (83 yrs). Although this rate is ~40% lower than our calculated rate for the windward side of the range, our hypothesis would predict a much lower rate of erosion for

D

the Elwha River basin since it is located on the leeward side of the range. One explanation may be that

TE

climate is still an important driver of erosion for this side of the range, but that high-elevation mass

AC CE P

wasting processes such as snow and debris avalanches and/or flows are the primary means of sediment release rather than landslides. We reject the possibility of anthropogenic influence because the entire Elwha River catchment above the former dam site is protected within Olympic National Park; this unexplained discrepancy is a target of future work.

Assuming the results of this study are accurate, the question of the potential impact that large earthquakes have on surface processes of upland areas, and subsequently how this manner of erosion compares to that driven by climate, still remains. Although it is impossible to quantify coseismic landsliding associated with the most recent CSZ rupture in 1700 CE, field data and observations in the upper Quinault River valley do provide evidence for possible impacts of this event. Roughly 20 km upstream of Lake Quinault, a cut-bank along the East Fork of the Quinault River exposes a massive 15 mthick, poorly-sorted unit consisting of large, angular boulders within a gravel and sand matrix. This

ACCEPTED MANUSCRIPT massive unit both overlies and is capped by ~1 m-thick beds of well-sorted fluvial gravels, a stratigraphy suggestive of disturbance in fluvial deposition caused by a large landslide event. The location of this

PT

deposit also corresponds with a prominent convexity in the Quinault River profile, which provides additional support for the possibility of a large, valley-blocking landslide (Fig. 7; Leithold et al., in press).

RI

Immediately downstream of the landslide deposit, and coincident with the confluence of Graves Creek,

SC

is a 5 m-high alluvial terrace that is continuous along both banks of the Quinault River for hundreds of

NU

meters and includes multiple buried and rooted conifer tree stumps in growth position from which limited 14C dating of outer growth rings yields a calibrated calendar age of AD 1760 ± 90 (Table 3; Fig. 7).

MA

This terrace deposit likely represents a relatively short period of rapid sediment accumulation that occurred as a result of breaching of the landslide dam created during the AD 1700 CSZ earthquake, after

TE

D

which the Quinault River gradually incised through the terrace and created the modern-day exposures.

AC CE P

In addition to the landslide and terrace deposits at Graves Creek, the U.S. Bureau of Reclamation mapped and dated a series of well-pronounced fluvial terraces located along a ~15 km span of the Quinault River immediately upstream of Lake Quinault, with the lowest terrace yielding an age of AD 1610 ± 90 from detrital Populus charcoal and the intermediate terrace yielding an age of 1460 ± 50 cal BP from detrital conifer charcoal (Bountry et al. 2005). The age of the lower terrace, like the deposit at Graves Creek, brackets the AD 1700 CSZ earthquake and in this case suggests the possibility of channel aggradation as the result of increased sediment delivery initiated by widespread coseismic landsliding. The formation of the intermediate terrace may be similarly described, as the 1460 ± 50 cal BP age is consistent with evidence of paleoseismic sediment disruption in Lake Quinault that may have been the result of an earlier CSZ rupture or a large earthquake triggered by slip along a nearby upper crustal fault (Leithold et al., in press). Collectively, these terrace deposits indicate that seismic shaking most likely does have an appreciable impact on erosion within Olympic Mountain catchments, but that the degree

ACCEPTED MANUSCRIPT of impact may depend on both the geomorphic conditions at the time of the event (e.g., antecedent soil moisture, duration since previous event; e.g., Wegmann and Pazzaglia, 2002) as well as the degree and

PT

spatial extent of ground shaking produced by each earthquake, which are related to the magnitude and location of the earthquake, but are also influenced by the surrounding topography (e.g. Meunier et al.,

RI

2008; Buech et al., 2010; Gallen et al., 2015). This is highlighted by particularly strong evidence for

SC

significant Lake Quinault sediment disruption ~1500 cal BP, but limited signs of lake-wide disturbance

NU

resulting from any known large subsequent earthquakes (Wegmann et al., 2014; Leithold et al., in

MA

press).

Coupling the precipitation/landslide relationship with evidence of seismically-induced erosion suggests

D

the long-term erosional evolution of the Olympic Mountains is governed in part by a balance between

TE

gradual, consistent climate forcing and abrupt, infrequent tectonic forcing. During times of seismic

AC CE P

quiescence, areas of the range that receive abundant precipitation continue to erode at a rate comparable to the long-term average, whereas the leeward portion of the range erodes much more slowly. When large earthquakes do occur, inducing strong ground shaking, it is probable that landsliding is widespread throughout all portions of the range, but it may also be the case that landsliding is of a greater density, area, and volume in those areas that are sheltered from receiving excessive orographic precipitation. Furthermore, even the season in which a large earthquake happens may matter in terms of hillslope sediment production. For instance, if a large earthquake takes place during the winter to early spring months when slopes are saturated, the shaking may generate more hillslope failures than if the same magnitude earthquake occurred when slopes are drier, such as in August or September. We acknowledge that this hypothesis is difficult to test in the absence of seismic activity, and thus future studies should focus on developing a more complete understanding of short and long-term erosion rates for individual catchments on either side of the Olympic orographic precipitation gradient.

ACCEPTED MANUSCRIPT

PT

6. Conclusions

Sediment volumes calculated for mapped landslides from a ~15 km-wide swath across the Olympic

RI

Mountains reveal a significant positive correlation with mean annual precipitation. The windward side of

SC

the range, which receives abundant orographic precipitation, is characterized by a greater number of

NU

landslides and associated volume of transported material compared to the drier leeward side of the range. For the period 1990–2015, the total volume of landslide sediment from the windward, high-

MA

precipitation portion of the study area yields a mean erosion rate of 0.28 ± 0.11 mm yr -1. This rate is similar to other estimates of erosion throughout the Olympic Mountains, including those from river 10

Be, fluvial terrace incision, and thermochronometry. The similarity

D

sediment yield, cosmogenic

TE

between our rates and those published previously suggests that climate is an important driver of both

AC CE P

short and long-term landscape evolution in regions of the Olympic Mountains receiving heavy annual precipitation. Additional work is necessary to determine the precise relative contributions of climate (i.e. precipitation) and tectonics (i.e. earthquakes) to overall erosion, but our results provide compelling evidence that climate plays an important and measurable role in delivering sediment from hillslopes to the fluvial network. This has significant implications for understanding mountain belt evolution in areas with similarly sharp precipitation gradients and for comprehending the mechanisms controlling source to sink sediment fluxes in such environments.

Acknowledgements

ACCEPTED MANUSCRIPT

We thank NC State students Robert Lane, Deanna Metevier, Catherine Opalka, and Sharese Roberts for

PT

their assistance with landslide mapping, and to the staff of Olympic National Park for their assistance with fieldwork logistics. This work was supported in part by a Geological Society of America Graduate

RI

Student Research Grant and by the Geomorphology and Land Use Dynamics Program National Science

AC CE P

TE

D

MA

NU

SC

Foundation (EAR-1226064).

ACCEPTED MANUSCRIPT References

PT

Atwater, B.F., 1987. Evidence for great Holocene earthquakes along the outer coast of Washington

RI

state. Science 236, 942-944.

Atwater, B. F., and Moore, A. L., 1992. A tsunami about 1000 years ago in Puget Sound, Washington.

SC

Science 258, 1614-1617.

NU

Atwater, B.F., and Hemphill-Haley, E., 1997. Recurrence intervals for great earthquakes of the past 3500

MA

years in northeastern Willapa Bay, Washington. U.S. Geological Survey Professional Paper 1576, 108 p.

Atwater, B.F., Tuttle, M.P., Schweig, E.S., Rubin, C.M., Yamaguchi, D.K., Hemphill-Haley, E., 2003.

D

Earthquake recurrence inferred from paleoseismology. The Quaternary Period in the United States, 331-

TE

350.

AC CE P

Audry, S., Schäfer, J., Blanc, G., Jouanneau, J., 2004. Fifty-year sedimentary record of heavy metal pollution (Cd, Zn, Cu, Pb) in the Lot River reservoirs (France). Environmental Pollution 132, 413-426.

Avnimelech, Y., Ritvo, G., Meijer, L.E., Kochba, M., 2001. Water content, organic carbon and dry bulk density in flooded sediments. Aquacultural Engineering 25, 25-33.

Babcock, R., Suczek, C., and Engebretson, D., 1994, The Crescent ‘‘Terrane’’, Olympic peninsula and southern Vancouver Island. Washington Division of Geology and Earth Resources Bulletin, v. 80, 141157.

Belmont, P., Pazzaglia, F., Gosse, J.C., 2007. Cosmogenic 10Be as a tracer for hillslope and channel sediment dynamics in the Clearwater River, western Washington State. Earth and Planetary Science Letters 264, 123-135.

ACCEPTED MANUSCRIPT Benda, L., Veldhuisen, C., and Black, J., 2003. Debris flows as agents of morphological heterogeneity at low-order confluences, Olympic Mountains, Washington. Geological Society of America Bulletin 115,

PT

1110-1121.

RI

Bennett, G.L., Molnar, P., McArdell, B.W., Schlunegger, F. and Burlando, P., 2013. Patterns and controls

SC

of sediment production, transfer and yield in the Illgraben. Geomorphology 188, 68-82.

NU

Binnie, S.A., Phillips, W.M., Summerfield, M.A., Fifield, L.K., Spotila, J.A., 2010. Tectonic and climatic controls of denudation rates in active orogens: The San Bernardino Mountains, California.

MA

Geomorphology 118, 249-261.

TE

Processes and Landforms 32, 329-365.

D

Bishop, P., 2007. Long‐term landscape evolution: linking tectonics and surface processes. Earth Surface

AC CE P

Bonnet, S., Crave, A., 2003. Landscape response to climate change: Insights from experimental modeling and implications for tectonic versus climatic uplift of topography. Geology 31, 123-126.

Bountry, J.A., Randle, T.J., Piety, L.A., Lyon, E.W., Jr., Abbe, T., Barton, C., Ward, G., Fetherston, K., Armstrong, B. and Gilbertson, L., 2005. Geomorphic Investigation of Quinault River, Washington: 18 Km Reach of Quinault River Upstream from Lake Quinault. U.S. Department of the Interior, Bureau of Reclamation, Denver, 175 p.

Bountry, J., Ferrari, R., Wille, K., and Randle, T.J., 2011. 2010 Survey Report for Lake Mills and Lake Aldwell on the Elwha River, Washington. U.S. Bureau of Reclamation Technical Report No. SRH-2010-23, 73 p.

ACCEPTED MANUSCRIPT Brandon, M.T., Calderwood, A.R., 1990. High-pressure metamorphism and uplift of the Olympic subduction complex. Geology 18, 1252-1255.

PT

Brandon, M.T., Roden-Tice, M.K., Garver, J.I., 1998. Late Cenozoic exhumation of the Cascadia

RI

accretionary wedge in the Olympic Mountains, northwest Washington State. Geological Society of

SC

America Bulletin 110, 985-1009.

NU

Buech, F., Davies, T.R., Pettinga, J.R., 2010. The Little Red Hill seismic experimental study: topographic effects on ground motion at a bedrock-dominated mountain edifice. Bulletin of the Seismological

MA

Society of America 100, 2219-2229.

D

Burbank, D.W., Leland, J., Fielding, E., Anderson, R.S., Brozovic, N., Reid, M.R., Duncan, C., 1996. Bedrock

TE

incision, rock uplift and threshold hillslopes in the northwestern Himalayas. Nature 379, 505-510.

AC CE P

Chou, W., Lin, W., Lin, C., 2009. Vegetation recovery patterns assessment at landslides caused by catastrophic earthquake: A case study in central Taiwan. Environmental Monitoring and Assessment 152, 245-257.

Dadson, S.J., Hovius, N., Chen, H., Dade, W.B., Lin, J., Hsu, M., Lin, C., Horng, M., Chen, T., Milliman, J., 2004. Earthquake-triggered increase in sediment delivery from an active mountain belt. Geology 32, 733-736.

DeMets, C., Dixon, T.H., 1999. New kinematic models for Pacific-North America motion from 3 Ma to present, I: Evidence for steady motion and biases in the NUVEL-1A model. Geophysical Research Letters 26, 1921-1924.

ACCEPTED MANUSCRIPT DiBiase, R.A., Whipple, K.X., 2011. The influence of erosion thresholds and runoff variability on the relationships among topography, climate, and erosion rate. Journal of Geophysical Research: Earth

PT

Surface 116, F04036, 1-17.

RI

Dragovich, J.D., Logan, R.L., Schasse, H.W., Walsh, T.J., Lingley Jr, W.S., Norman, D.K., Gerstel, W.J.,

SC

Lapen, T.J., Schuster, J.E., Meyers, K.D., 2002. Geologic map of Washington, Northwest quadrant.

NU

Washington Division of Geology and Earth Resources, GM-50, 1:250000 scale.

Emberson, R., Hovius, N., Galy, A., Marc, O., 2016. Chemical weathering in active mountain belts

MA

controlled by stochastic bedrock landsliding. Nature Geoscience 9, 42-45.

D

England, P., Molnar, P., 1990. Surface uplift, uplift of rocks, and exhumation of rocks. Geology 18, 1173-

TE

1177.

AC CE P

Ferrier, K.L., Perron, J.T., Mukhopadhyay, S., Rosener, M., Stock, J.D., Huppert, K.L., Slosberg, M., 2013. Covariation of climate and long-term erosion rates across a steep rainfall gradient on the Hawaiian island of Kaua ‘i. Geological Society of America Bulletin 125, 1146-1163.

Francescato, V., Scotton, M., Zarin, D.J., Innes, J.C., Bryant, D.M., 2001. Fifty years of natural revegetation on a landslide in Franconia Notch, New Hampshire, USA. Canadian Journal of Botany 79, 1477-1485.

Gallen, S.F., Clark, M.K., Godt, J.W., 2015. Coseismic landslides reveal near-surface rock strength in a high-relief, tectonically active setting: Geology 43, 11-14.

Gasparini, N.M., Whipple, K.X., 2014. Diagnosing climatic and tectonic controls on topography: Eastern flank of the northern Bolivian Andes. Lithosphere 6, 230-250.

ACCEPTED MANUSCRIPT Gavin, D.G., Brubaker, L.B., 2015. Late Pleistocene and Holocene Environmental Change on the Olympic Peninsula, Washington. Ecological Studies 222, Springer International Publishing.

PT

Gerstel, W.J., 1999, Deep-seated landslide inventory of the west-central Olympic Peninsula. Washington

RI

Division of Geology and Earth Resources Open File Report 99-2, 36 p., 2 plates.

SC

Goldfinger, C., 2011. Submarine paleoseismology based on turbidite records. Annual Review of Marine

NU

Science 3, 35-66.

MA

Goldfinger, C., Nelson, C.H., Morey, A.E., Johnson, J.E., Patton, J.R., Karabanov, E., Gutierrez-Pastor, J., Eriksson, A.T., Gracia, E., Dunhill, G., 2012. Turbidite event history: Methods and implications for

TE

Geological Survey, Reston, 170 p.

D

Holocene paleoseismicity of the Cascadia Subduction Zone. USGS Professional Pape 1661-F, U.S.

AC CE P

Graehl, N.A., Kelsey, H.M., Witter, R.C., Hemphill-Haley, E., Engelhart, S.E., 2015. Stratigraphic and microfossil evidence for a 4500-year history of Cascadia subduction zone earthquakes and tsunamis at Yaquina River estuary, Oregon, USA. Geological Society of America Bulletin 127, 211-226.

Guariguata, M.R., 1990. Landslide disturbance and forest regeneration in the upper Luquillo Mountains of Puerto Rico. The Journal of Ecology, 814-832.

Guzzetti, F., Cardinali, M., Reichenbach, P., Cipolla, F., Sebastiani, C., Galli, M., Salvati, P., 2004. Landslides triggered by the 23 November 2000 rainfall event in the Imperia Province, Western Liguria, Italian Journal of Engineering Geology and Environment 73, 229-245.

Hovius, N., Stark, C.P., Allen, P.A., 1997. Sediment flux from a mountain belt derived by landslide mapping. Geology 25, 231–234.

ACCEPTED MANUSCRIPT Hovius, N., Stark, C.P., Hao‐Tsu, C. and Jiun‐Chuan, L., 2000. Supply and removal of sediment in a landslide‐dominated mountain belt: Central Range, Taiwan. The Journal of Geology 108, 73-89.

PT

Hovius, N., Meunier, P., Lin, C., Chen, H., Chen, Y., Dadson, S., Horng, M., Lines, M., 2011. Prolonged

RI

seismically induced erosion and the mass balance of a large earthquake. Earth and Planetary Science

SC

Letters 304, 347-355.

NU

Hua, Q., Barbetti, M., Rakowski, A.Z., 2013. Atmospheric radiocarbon for the period 1950–2010.

MA

Radiocarbon 55, 2059-2072.

Keefer, D.K., 1984. Landslides caused by earthquakes: Geological Society of America Bulletin 95, 406-

D

421.

TE

Kelsey, H.M., Nelson, A.R., Hemphill-Haley, E., Witter, R.C., 2005. Tsunami history of an Oregon coastal

AC CE P

lake reveals a 4600 yr record of great earthquakes on the Cascadia subduction zone. Geological Society of America Bulletin 117, 1009-1032.

Korup, O., McSaveney, M.J. and Davies, T.R., 2004. Sediment generation and delivery from large historic landslides in the Southern Alps, New Zealand. Geomorphology 61, 189-207.

Korup, O., Densmore, A.L., Schlunegger, F., 2010. The role of landslides in mountain range evolution. Geomorphology 120, 77-90.

Larsen, I.J., Montgomery, D.R., Korup, O., 2010. Landslide erosion controlled by hillslope material. Nature Geoscience 3, 247-251.

ACCEPTED MANUSCRIPT Lazzari, M., Gioia, D., Piccarreta, M., Danese, M., Lanorte, A., 2015. Sediment yield and erosion rate estimation in the mountain catchments of the Camastra artificial reservoir (Southern Italy): A

PT

comparison between different empirical methods. Catena 127, 323-339.

RI

Leithold, E.L., Wegmann, K.W., Bohnenstiehl, D.R., Smith, S.G., Noren, A., O’Grady, R., 2017. Slope

SC

failures within and upstream of Lake Quinault, Washington, as uneven responses to Holocene

NU

earthquakes along the Cascadia Subduction Zone. Quaternary Research, In press.

Lira, C., Lousada, M., Falcão, A., Gonçalves, A., Heleno, S., Matias, M., Pereira, M., Pina, P., Sousa, A.,

MA

Oliveira, R., 2013. The 20 February 2010 Madeira Island flash-floods: VHR satellite imagery processing in support of landslide inventory and sediment budget assessment. Natural Hazards and Earth System

TE

D

Sciences 13, 709-719.

Marc, O., Hovius, N., Meunier, P., Uchida, T., Hayashi, S., 2015. Transient changes of landslide rates after

AC CE P

earthquakes. Geology 43, 883-886.

Meunier, P., Hovius, N., Haines, J.A., 2008. Topographic site effects and the location of earthquake induced landslides. Earth and Planetary Science Letters 275, 221-232.

Minder, J.R., Durran, D.R., Roe, G.H., Anders, A.M., 2008. The climatology of small‐scale orographic precipitation over the Olympic Mountains: Patterns and processes. Quarterly Journal of the Royal Meteorological Society 134, 817-839.

Montgomery, D.R., Greenberg, H.M., 2000. Local relief and the height of Mount Olympus. Earth Surface Processes and Landforms 25, 385-396.

ACCEPTED MANUSCRIPT Montgomery, D.R., 2001. Slope distributions, threshold hillslopes, and steady-state topography. American Journal of Science 301, 432-454.

PT

Montgomery, D.R., Balco, G., Willett, S.D., 2001. Climate, tectonics, and the morphology of the Andes.

RI

Geology 29, 579-582.

SC

Montgomery, D.R., Brandon, M.T., 2002. Topographic controls on erosion rates in tectonically active

NU

mountain ranges. Earth and Planetary Science Letters 201, 481-489.

MA

National Oceanic and Atmospheric Administration, National Centers for Environmental Information. https://www.ncdc.noaa.gov/temp-and-precip/climatological-rankings/index.php [accessed March 15,

D

2016].

TE

Neiman, P.J., Schick, L.J., Ralph, F.M., Hughes, M., Wick, G.A., 2011. Flooding in Western Washington:

AC CE P

the connection to atmospheric rivers. Journal of Hydrometeorology 12, 1337-1358.

Nelson, L. M., 1986. Fluvial sediment in the Hoh River. In Lum, W. E. (ed.), Reconnaissance of the water resources of the Hoh Indian Reservation and the Hoh River basin, Washington, U.S. Geological Survey Water Resources Investigations Report 85–4018, 18–19.

Nelson, A.R., Kelsey, H.M., Witter, R.C., 2006. Great earthquakes of variable magnitude at the Cascadia subduction zone. Quaternary Research 65, 354-365.

Pacific Northwest Seismic Network. https://www.pnsn.org [accessed April 7, 2016].

Parker, R.N., Densmore, A.L., Rosser, N.J., De Michele, M., Li, Y., Huang, R., Whadcoat, S., Petley, D.N., 2011. Mass wasting triggered by the 2008 Wenchuan earthquake is greater than orogenic growth. Nature Geoscience 4, 449-452.

ACCEPTED MANUSCRIPT Patterson, R.T., Chang, A.S., Prokoph, A., Roe, H.M., Swindles, G.T., 2013. Influence of the Pacific Decadal Oscillation, El Niño-Southern Oscillation and solar forcing on climate and primary productivity

PT

changes in the northeast Pacific. Quaternary International 310, 124-139.

RI

Pazzaglia, F.J., Brandon, M.T., 2001. A fluvial record of long-term steady-state uplift and erosion across

Climate

Group,

Oregon

State

University,

30-yr

Normal

Precipitation:

1981-2010,

NU

PRISM

SC

the Cascadia forearc high, western Washington State. American Journal of Science 301, 385-431.

MA

http://prism.oregonstate.edu [accessed September 1, 2015]

Quinault Indian Nation, 1999. Quinault River Watershed Analysis. U.S. Forest Service; U.S. Park Service;

D

U.S. Geological Survey.

TE

Ramsey, C.B., 2009. Bayesian analysis of radiocarbon dates. Radiocarbon 51, 337-360.

AC CE P

Riebe, C.S., Kirchner, J.W., Granger, D.E., Finkel, R.C., 2001. Minimal climatic control on erosion rates in the Sierra Nevada, California. Geology 29, 447-450.

Reid, L.M., and Dunne, T., 1984. Sediment production from forest road surfaces. Water Resources Research 20, 1753-1761.

Reimer, P.J., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Ramsey, C.B., Buck, C.E., Cheng, H., Edwards, R.L., Friedrich, M., 2013. IntCal13 and Marine13 radiocarbon age calibration curves 0–50,000 years cal BP. Radiocarbon 55, 1869-1887.

Reneau, S.L., Dietrich, W.E., Rubin, M., Donahue, D.J., and Jull, T.A.J., 1989. Analysis of hillslope erosion rates using dated colluvial deposits. Journal of Geology 97, 45-63.

ACCEPTED MANUSCRIPT Satake, K., Shimazaki, K., Tsuji, Y., Ueda, K., 1996. Time and size of a giant earthquake in Cascadia inferred from Japanese tsunami records of January 1700. Nature 379, 246-249.

PT

Smale, M., McLeod, M., Smale, P., 1997. Vegetation and soil recovery on shallow landslide scars in

RI

tertiary hill country, East Cape region, New Zealand. New Zealand Journal of Ecology, 31-41.

SC

Snyder, N.P., Rubin, D.M., Alpers, C.N., Childs, J.R., Curtis, J.A., Flint, L.E., Wright, S.A., 2004. Estimating

MA

northern California Water Resources Research 40.

NU

accumulation rates and physical properties of sediment behind a dam: Englebright Lake, Yuba River,

Stolar, D., Roe, G., Willett, S., 2007. Controls on the patterns of topography and erosion rate in a critical

D

orogen. Journal of Geophysical Research 112, F4. DOI: 10.1029/2006JF000713.

TE

Swanson, F.J., Benda, L.E., Duncan, S.H., Grant, G.E., Megahan, W.F., Reid, L.M., and Ziemer, R.R., 1987.

AC CE P

Mass failures and other processes of sediment production in Pacific Northwest forest landscapes. In Salo, E.O., Cundy, T.W. (eds.), Streamside Management: Foresty and Fishery Interactions, Proceedings of a Symposium held at University of Washington, 12-14 February, 1986: Seattle, Washington, Contribution no. 57, Institute of Forest Resources.

Tabor, R.W., 1971. Origin of Ridge-Top Depressions by Large-Scale Creep in the Olympic Mountains, Washington. Geological Society of America Bulletin 82, 1811-1822.

Tabor, R.W., Cady, W.M., 1978. The Structure of the Olympic Mountains, Washington: Analysis of a Subduction Zone. U.S. Geological Survey Professional Paper 1033, 39 p.

Tang, C., Zhu, J., Ding, J., Cui, X.F., Chen, L., Zhang, J.S., 2011. Catastrophic debris flows triggered by a 14 August 2010 rainfall at the epicenter of the Wenchuan earthquake. Landslides 8, 485-497.

ACCEPTED MANUSCRIPT Thackray, G.D., 2001. Extensive early and middle Wisconsin glaciation on the western Olympic Peninsula, Washington, and the variability of the Pacific moisture delivery to the northwestern United

PT

States. Quaternary Research 55, 257-270.

RI

U.S. Geological Survey, The National Map, 2016. 3DEP products and services: The National Map, 3D

SC

Elevation Program Web page, http://nationalmap.gov/3DEP/3dep_prodserv.html [accessed October 15,

NU

2015]

Ware, D.M., Thomson, R.E., 2000. Interannual to multidecadal timescale climate variations in the

MA

Northeast Pacific. Journal of Climatology 13, 3209-3220.

D

Wegmann, K.W., Pazzaglia, F.J., 2002. Holocene strath terraces, climate change, and active tectonics:

AC CE P

114, 731-744.

TE

The Clearwater River basin, Olympic Peninsula, Washington State. Geological Society of America Bulletin

Wegmann, K.W., 2004. Landslide hazard zonation project—Mass wasting assessment—Quinault Lake, Quinault River, and Cook-Elk watersheds, Jefferson and Grays Harbor Counties, Washington: Washington Department of Natural Resources, Forest Practices, 12 p., 4 plates, scale 1:24000.

Wegmann, K.W., Leithold, E.L., Bohnenstiehl, D.R., 2014. How important is seismically-induced erosion above the Cascadia subduction zone? Insights from the stratigraphy of large lakes on the Olympic Peninsula, Washington State. In Gillespie, A., Montgomery, D. (eds.), Proceedings of the 23rd Biennial Meeting of the American Quaternary Association, Volume 23: University of Washington, Seattle, WA, American Quaternary Union, 35-37.

ACCEPTED MANUSCRIPT Wenske, D., Jen, C.H., Böse, M. and Lin, J.C., 2012. Assessment of sediment delivery from successive erosion on stream-coupled hillslopes via a time series of topographic surveys in the central high

PT

mountain range of Taiwan. Quaternary International 263, 14-25.

RI

Whipple, K.X., 2009. The influence of climate on the tectonic evolution of mountain belts. Nature

SC

Geoscience 2, 97-104.

NU

Willett, S.D., 1999. Orogeny and orography: The effects of erosion on the structure of mountain belts.

MA

Journal of Geophysical Research: Solid Earth 104, 28957-28981.

Willett, S.D., Brandon, M.T., 2002. On steady states in mountain belts. Geology 30, 175-178.

D

Yamaguchi, D.K., Atwater, B.F., Bunker, D.E., Benson, B.E., Reid, M.S., 1997. Tree-ring dating the 1700

TE

Cascadia earthquake. Nature 389, 922-923.

AC CE P

Yanites, B.J., Tucker, G.E., Mueller, K.J., Chen, Y., 2010. How rivers react to large earthquakes: Evidence from central Taiwan. Geology 38, 639-642.

Zängl, G., 2007. Interaction between dynamics and cloud microphysics in orographic precipitation enhancement: A high-resolution modeling study of two north Alpine heavy-precipitation events. Monthly Weather Review 135, 2817–2840.

ACCEPTED MANUSCRIPT

Table 1 Topographic, precipitation, and landslide data for each block of the study grid. mean

mean

mean annual

#

landslide

area

elevation

slope

precipitation

of

area

2

2a

(degrees)

(m yr )

landslides

(m )

1

36.5

383

20.7

3.6

12

1.81 x 10

2

30.4

568

24.0

4.0

7

3.33 x 10

3

39.2

608

25.0

4.0

20

5.63 x 10

4

41.4

622

27.5

3.7

33

3.38 x 10

5

39.5

534

24.0

3.7

13

6

16.2

698

27.0

3.8

26

7

25.8

853

26.4

4.1

23

8

40.3

877

28.2

4.2

49

9

41.6

524

27.0

3.9

10

40.0

649

26.5

3.7

11

13.3

838

28.4

4.3

12

41.5

967

28.2

4.8

13

22.4

1072

31.0

14

41.2

1183

28.3

15

41.8

638

25.0

16

38.7

1112

28.0

17

11.2

1239

18

19.9

19

40.0

MA

1.80 x 10 1.09 x 10

17

2.14 x 10

27

1.45 x 10

59

1.74 x 10

4.8

92

6.37 x 10

3.9

23

9.31 x 10

4.2

10

1.03 x 10

4.7

44

4.13 x 10

27.5

3.9

23

1.62 x 10

1095

26.4

2.9

24

1.12 x 10

1309

25.1

2.1

19

2.73x 10

AC

CE

PT ED

3.30 x 10 7.07 x 10

27.0

3.7

41

4 4 5

NU 2.04 x 10

26

1239

4

SC

(m)

40.0

landslide 3

volume

(km )

20 a

-1

landslide

PT

Block

RI

Block #

4 5 5 5 4 5 5 5 5 4 5 5 5 5

4

1.42 x 10

5

volume (m )

3 b

(m )

1990–2015

3

3

5

4

4

4

6

5

4

4

5

5

5

5

6

5

5

4

6

5

5

5

5

5

6

6

5

5

5

5

6

6

5

5

5

5

4

4

5

5

8.35 x 10 ± 3.17 x 10 1.19 x 10 ± 4.51 x 10 7.04 x 10 ± 2.67 x 10 1.94 x 10 ± 7.37 x 10 3.25 x 10 ± 1.23 x 10 9.34 x 10 ± 3.55 x 10 3.93 x 10 ± 1.50 x 10 1.98 x 10 ± 7.54 x 10 1.47 x 10 ± 5.57 x 10 1.84 x 10 ± 6.98 x 10 5.54 x 10 ± 2.10 x 10 3.86 x 10 ± 1.47 x 10 3.25 x 10 ± 1.24 x 10 2.90 x 10 ± 1.10 x 10 4.97 x 10 ± 1.89 x 10 2.83 x 10 ± 1.08 x 10 7.01 x 10 ± 2.66 x 10 3.44 x 10 ± 1.31 x 10

3

3

5

4

4

4

5

5

4

4

5

4

5

5

4

4

4

4

4

4

5

5

5

4

6

5

2

1

5

4

5

5

5

4

5

4

6.22 x 10 ± 2.36 x 10 1.15 x 10 ± 4.36 x 10 6.97 x 10 ± 2.65 x 10 7.92 x 10 ± 3.01 x 10 2.75 x 10 ± 1.04 x 10 2.12 x 10 ± 8.07 x 10 3.29 x 10 ± 1.25 x 10 6.03 x 10 ± 2.29 x 10 6.73 x 10 ± 2.56 x 10 8.51 x 10 ± 3.23 x 10 3.55 x 10 ± 1.35 x 10 1.80 x 10 ± 6.84 x 10 1.41 x 10 ± 5.37 x 10 1.17 x 10 ± 4.45 x 10 1.34 x 10 ± 5.08 x 10 3.67 x 10 ± 1.40 x 10 1.42 x 10 ± 5.40 x 10 2.30 x 10 ± 8.72 x 10

3.43 x 10 ± 1.30 x 10 3.71 x 10 ± 1.41 x 10

0 3

4.68 x 10 ± 1.78 x 10

3

A 15% error is assumed for the mapping procedure.

b

Uncertainty includes propagated errors from the determination of landslide area and standard deviation for α and ϒ in the formula, V = αAϒ, used to calculate landslide volume (VL).

ACCEPTED MANUSCRIPT

Table 1 (continued) Topographic, precipitation, and landslide data for each block of the study grid.

2

(m)

(degrees)

mean annual precipitation -1 (m yr )

# of

landslide area

landslides

(m )

5

35.7

1253

28.1

2.9

86

2.08 x 10

22

9.3

1070

25.4

2.4

46

1.19 x 10

23

38.1

1449

26.9

1.9

14

4.41 x 10

24

26.5

1541

26.6

2.0

4

25

41.4

985

29.6

2.1

19

26

39.5

995

25.5

1.9

22

27

33.3

1640

26.5

1.9

12

28

22.8

1669

27.2

1.9

16

33 34 35

24.1 22.1 23.0 24.0

1477 1310 1256 1444 1020

27.6 28.6 28.9 28.3 26.8 25.3

1.2

13 2 4

SC

MA

3.52x 10

2.19 x 10 1.16 x 10 2.64 x 10 5.83 x 10 1.14 x 10

2.84 x 10

728

13.2

1.2

12

6.71 x 10

489

15.6

1.0

10

1.76 x 10

714

38

39.5

39

41.8

435 n/a 1010

15.3 n/a 25.7

1.0 n/a 2.8

3 937 23.4

5 4 4 5 4 4 4 4

4

2

42.2

29.8

1.6

27

1.22 x 10

1.2

37

AVERAGE

1.3

37

5.75 x 10

17.7

23.6

1258

1.5

8

2.83 x 10

2.61 x 10

741

SUM

1.8

2

8.17 x 10

10

34.6

18.4

1.8

1.02 x 10

1.2

36

40

1.8

2.53 x 10

PT ED

32

25.1

1545

30.4

CE

31

32.3

1619

AC

30

22.8

landslide volume

2a

21

29

a

mean slope

NU

(km )

mean elevation

PT

Block area

RI

Block #

7.73 x 10

5 5 4 3 4 4 3 3 4 3

4.53 x 10

6

landslide 3 volume (m )

3 b

(m )

1990–2015

5

5

5

4

5

4

4

4

5

5

5

4

4

4

5

4

4

4

5

4

5

5

5

5

4

3

3

3

4

4

4

4

3

2

3

3

3

3

3

3

7

6

5

5

5.39 x 10 ± 2.05 x 10 2.37 x 10 ± 8.99 x 10 1.52 x 10 ± 5.77 x 10 9.99 x 10 ± 3.79 x 10 3.48 x 10 ± 1.32 x 10 2.50 x 10 ± 9.48 x 10 5.91 x 10 ± 2.25 x 10

3

3

3

3

2

1

2

2

3

2

2

2

2

2

4.05 x 10 ± 1.54 x 10 7.95 x 10 ± 3.02 x 10 2.37 x 10 ± 9.01 x 10 2.67 x 10 ± 1.01 x 10 1.76 x 10 ± 6.67 x 10 7.01 x 10 ± 2.66 x 10 3.56 x 10 ± 1.35 x 10

1.49 x 10 ± 5.67 x 10

0

4.84 x 10 ± 1.84 x 10

0

1.14 x 10 ± 4.32 x 10

0

8.85 x 10 ± 3.36 x 10

0

4.95 x 10 ± 1.88 x 10

0

1.28 x 10 ± 4.88 x 10

0

2.89 x 10 ± 1.10 x 10

0

2.66 x 10 ± 1.01 x 10 6.85 x 10 ± 2.60 x 10 1.29 x 10 ± 4.91 x 10 2.76 x 10 ± 1.05 x 10

0 4

4

2

2

2

1

6.27 x 10 ± 2.38 x 10 2.72 x 10 ± 1.04 x 10 2.15 x 10 ± 8.15 x 10

8.25 x 10 ± 3.13 x 10

0

3.83 x 10 ± 1.46 x 10

2.02 x 10 ± 7.69 x 10 5.06 x 10 ± 1.92 x 10

0 6

6

5

4

4.67 x 10 ± 1.77 x 10 1.17 x 10 ± 4.43 x 10

A 15% error is assumed for the mapping procedure.

b

Uncertainty includes propagated errors from the determination of landslide area and standard deviation for α and ϒ in the formula, V = αAϒ, used to calculate landslide volume (VL).

ACCEPTED MANUSCRIPT Table 2 Landslide volumes and erosion rates for high and low precipitation areas of the study region. Area

region

2

elevation

Mean slope

mean

landslide

precip

volume (m )

rate

-1

(mm yr )

(m)

(degrees)

(m yr )

1990–2015

4.60 x 10 ± 1.75 x 10

657

848

27.0

3.93

Low Precipitation (leeward)

601

1172

24.7

1.63

7.45 x 10 ± 2.83 x 10

All

1258

1010

25.7

4.67 x 10 ± 1.77 x 10

SC

RI

High Precipitation (windward)

2.78

MA

a

erosion

3 a

NU

(km )

Mean

PT

Study

-1 a

6

6

0.280 ± 0.106

4

4

0.005 ± 0.002

6

6

0.149 ± 0.056

AC CE P

TE

D

Uncertainty includes propagated error from landslide area and 1σ standard deviation for α and ϒ in the formula, VL ϒ = αA , for landslide volume (VL).

ACCEPTED MANUSCRIPT

Table 3 Radiocarbon (14C) geochronology data for East Fork Quinault River landslide deposit.

(‰)a

%MC

-22.5

97.31

Radiocarbon age

1 error 0.33

BP 219

1 error 27

Calibrated ageb

cal AD range

probability

1640–1685 1735–1805 1935–

37.20% 44.50% 13.70%

SC

RI

D-AMS 007634

Fraction of modern

PT

Lab ID

(13C)

All results have been corrected for isotopic fractionation with 13C values measured on the prepared graphite using the AMS spectrometer

NU

a

b

AC CE P

TE

D

MA

Radiocarbon calibration was performed using the OxCal 4.2 online calculator using the IntCal 13 calibration curve (Bronk Ramsey, 2009). Results were rounded by 5 years.

ACCEPTED MANUSCRIPT Figures (if not included as separate file) and Figure Captions

PT

Fig. 1. Geologic setting of the study area. Top Inset: Simplified tectonic map of the Cascadia subduction zone. The Olympic Mountains (OM) study area is outlined by the dashed box. Main map: Simplified

RI

geologic map of the Olympic Peninsula (from Dragovich et al. 2002) draped over a hillshade of digital topography. Sedimentary rocks of the Olympic Subduction complex (OSC) are in green, while basalts of

SC

the Eocene Crescent Formation (CF) are colored in brown. Faults are shown by solid or dashed (inferred) black lines. The Hurricane Ridge fault (HRF) marks the contact between the OSC and peripheral rocks

NU

such as the CF. Faults with Holocene surface rupture are as follows: Seattle fault zone (SFZ); Saddle Mountain fault zone (SMFZ); Canyon River fault (CRF); Hood Canal fault (HCF); Tacoma fault zone (TFZ);

MA

and Lake Creek – Boundary Creek fault (LC-BCF). The black-lined polygon delineates the study area

D

swath.

TE

Fig. 2. Study area outlined on a satellite image of Washington’s Olympic Peninsula. The locations of Mount Olympus, Lake Crescent, Lake Quinault, the Clearwater River basin, the Glines Canyon dam,

AC CE P

Graves Creek, and the downstream reaches of major rivers (white lines) are provided for regional context and also for references made in the discussion section of this paper. The study area swath is delineated by the white-lined polygon consisting of 40 numbered grid blocks. Boundaries of individual grid blocks are based on Public Land Survey Township and Range subdivisions for western Washington State; the average block area is 31 km2.

Fig. 3. Precipitation distribution in the study area. Left: Distribution of mean annual precipitation (MAP; cm yr-1) for the period 1981–2010 draped over hillshaded digital topography of the Olympic Peninsula (PRISM, 2015). The study area is outlined in black and downstream reaches of major rivers are identified and labeled. Right: Precipitation and elevation profiles across the Olympic Mountains from A to A’ (see the left panel for location). Note the abrupt decrease in precipitation to the east of km ~60 that separates the western windward from the eastern leeward sides of the range.

ACCEPTED MANUSCRIPT Fig. 4. Google Earth images of landslides. A: Perspective-view example of delineated landslides within the Upper Quinault River catchment using Landsat imagery available in Google Earth (grid block 13). B: Area calculations derived by importing the landslide polygons originally determined in Google Earth into

PT

ArcMap 10.3. In the lowest-elevation polygon, landslide volume is written in red font below the surface area determination. Volumes were calculated using power-law scaling parameters reported in Larsen et

RI

al. (2010). C and D: Perspective-view example of a post-1990 landslide utilized for erosion rate

SC

calculation. C shows Imagery from 1990 with scar of future landslide outlined by the white dashed line. D shows post-landslide imagery from 2013 with the same white dashed line delineating the landslide. The depicted landslide occurred along Rustler Creek in the North Fork Quinault River catchment (grid

NU

block 13). Coordinates for the centroid of each photo are as follows: A/B: 47.6739° N, 123.6204° W; C/D:

AC CE P

TE

D

MA

47.6358° N, 123.5958° W.

Fig. 5. Plot of area–normalized landslide volume (VL(n); m) versus mean annual precipitation (MAP; m yr1

) for each of the 40 blocks within the study grid. Vertical error bars represent propagated uncertainty in

the volume calculation (see Methods section for details). The equation, VL(n) = 1.5×10-5MAP, represents the significant positive relationship between MAP and VL(n) (p = 0.0002) as determined using a permutation test.

Fig. 6. Relative contribution of high-precipitation (grid blocks 1–18 and 20–21) and low-precipitation (grid blocks 19 and 22−40) areas to total landslide volume for the period 1990–2015. The study area (including the division between “high” and “low” areas of precipitation) is outlined in black, and

ACCEPTED MANUSCRIPT downstream reaches of major rivers are marked with black lines. Over the 25-year span, 98% of landslide volume has been produced in the high precipitation, windward portion of the study grid,

AC CE P

TE

D

MA

NU

SC

RI

PT

yielding an erosion rate of 0.28 ± 0.11 mm yr-1 for this area.

ACCEPTED MANUSCRIPT Fig. 7. Field and geochronological evidence for the impact of the AD 1700 Cascadia subduction zone earthquake on the upper Quinault River catchment. A: Contour map of the area surrounding the confluence of Graves Creek and the East Fork of the Quinault River (see arrow on Fig. 2 for location).

PT

Hypothesized valley-blocking landslide is mapped in yellow and downstream event terrace is mapped in orange; consult discussion for additional details. Black star corresponds to the location of (C); white star

RI

corresponds to the location of (D). B: Panoramic photograph of the northern cut-bank of the Quinault

SC

River, just upstream of the confluence with Graves Creek. The 5 m-thick event terrace is visible on the downstream (left) side of the image and a portion of the 15 m-thick landslide deposit is visible on the right, upstream side of the photo. C: Photograph of the 15 m-thick landslide deposit; the boulder

NU

outlined in white has a long axis of approximately 1 m. D: Photograph of the event terrace from the opposite bank of the Quinault River, just downstream of Graves Creek. The blue star indicates the

MA

location of the 14C sample acquired from the outermost growth rings of one of several large trees buried in place by the aggradation of sediment now recorded as the event terrace upon which a mature conifer

AC CE P

TE

D

forest is now growing.

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Figure 1

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Figure 2

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

Figure 3

Figure 4

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

TE AC CE P

Figure 5

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC CE P

TE

D

Figure 6

TE AC CE P

Figure 7

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT Highlights 

Landslide volume is positively correlated with mean annual precipitation.



A calculated basin-wide erosion rate of 0.28 ± 0.11 mm yr-1 is consistent with previous research.



Between 1990-2015, 98% of landslide volume occurred in areas of high precipitation.

PT

1.1

AC CE P

TE

D

MA

NU

SC

RI

1.2