1

Timing, Extent, and Continued Evolution of the Van Zandt Landslide Complex, Whatcom County, Washington

Thesis Proposal for the Master of Science Degree Department of Geology, Western Washington University

Bellingham, Washington

Geoffrey Malick September 2, 2015

Approved by Advisory Committee Members:

__________________________________________________________

Dr. Douglas H. Clark, Thesis Committee Chair

__________________________________________________________

Dr. Jackie Caplan- Auerbach, Thesis Committee Advisor

__________________________________________________________

Dr. Robert J. Mitchell, Thesis Committee Advisor

_______________________________________________________

Dr. Scott R. Linneman, Thesis Committee Advisor

2

Introduction and Problem Statement

The Pacific Northwest is particularly susceptible to large slope failures because of its high relief

terrain, seismic activity, and an abundance of precipitation (Blais-Stevens et al., 2011). Despite

recent tragedies like the Oso, Washington, slide (3/22/2014; 43 fatalities), our understanding of

the mechanics, timing, and possible triggers of such landslides remains poor. Although several

recent studies (e.g., Iverson et al., 2015; Keaton et al., 2014) have focused on large slope failures

that initiate in unconsolidated glacial deposits, little attention so far has been given to

catastrophic bedrock landslides. To help address this deficiency, I will characterize the Van

Zandt Landslide Complex (VZLC), a high-mobility bedrock landslide in the western foothills of

the North Cascades of Washington (Fig. 1); the goals of my research are to 1) determine ages,

volumes, and potential triggering events for the distinct lobes of the VZLC; 2) test hypotheses

that such large bedrock landslides are linked to other regional landslides, paleoseismic events, or

paleoclimatic trends, and 3) evaluate potential indicators of continued slope deformation in the

source zone.

Dating prehistoric landslides can be difficult because many lack natural exposures and

datable material linked to their formation. Fortuitously, the VZLC deposit contains an abundance

of naturally formed ponds on its debris lobes (Fig. 2); these ponds, as well as several natural

river-cut exposures, provide an exceptional opportunity to collect and date sediments related to

the various slides and to thereby establish a comprehensive chronology of failures at this site.

Such constraints are crucial to test proposed causal links between such slides and paleoseismic or

paleoenvironmental events. Furthermore, previous workers (e.g., Brunengo, 2001) and initial

reconnaissance efforts have identified evidence of ongoing deformation and small slope failures

in the source area. For this study, I will evaluate the relationship between shallow

3

microseismicity and tension gaps in the headwall region to quantify progressive strain and test

the potential for future hazards at the VZLC. The results will provide crucial new constraints for

hazard assessments of a poorly understood type of catastrophic landslide that threatens people

and property in the Pacific Northwest, complementing ongoing investigations of sediment-

sourced landslides in the region (Iverson et al., 2015; Keaton et al., 2014).

Landslide hazard assessment requires knowledge of the age, frequency, and volume of

previous events, the identification of slopes prone to failure, and data on active deformation in

potential initiation zones (Orwin et al., 2004). The recent devastating debris avalanche-flow in

Oso emphasized the great hazard represented by such high-mobility landslides in the Pacific

Northwest as well as our poor understanding of how and why they occur. The Oso landslide was

extremely destructive because of an exceptionally high travel distance compared to previous

historic events at that location (Keaton et al., 2014). Such long-runout landslides, however, are

not uncommon in the Pacific Northwest. According to lidar evidence, the VZLC appears to have

had at least two large failures with travel distances comparable to that of the Oso slide.

Triggering-event and threshold-condition constraints for long-runout landslides are crucial to

understanding hazard susceptibility throughout the region, especially in the face of growing

population in potentially hazardous areas and increased landslide potential related to greater

rainstorm intensities linked to forecasted climate change (Huggel et al., 2011).

Background

Location and Morphology

The VZLC is located approximately 25 km east of Bellingham, WA, in the western foothills of

the North Cascades (Fig. 1). Preliminary mapping based on 2013 lidar data (Fig. 2) indicates that

4

VZLC comprises at least three prehistoric cross-cutting debris lobes that initiated on the steep

western slope of a prominent north-south trending ridge known as the Van Zandt Dike (VZD). In

addition to these larger lobes, there are also several possible smaller deposits and colluvial lobes

formed from shallow slope failures.

The 11 km long VZD is shaped like a north-pointing arrowhead that is ~3 km wide at its

southern end and tapers to ~1.5 km to the north where it forms the VZLC source zone (Fig. 1).

The VZD rises steeply >500 m above the lower reaches of the South Fork Nooksack River

reaching a maximum elevation of 672 m above sea level (asl). The northern half of the VZD (and

the entire VZLC source zone) is made up of the Eocene Chuckanut Formation (sandstone,

mudstone, siltstone, and minor coal intervals) whereas the southern half is Darrington Phyllite

(Tabor et al., 2003).

The crescent-shaped headscarp region for the VZLC is located along the northwestern

edge of VZD and is close to 5 km long. A significant slope break defines the upper reaches of the

steep headwall from the relatively flat top of the VZD (Fig. 2). Along this flat upper region of the

VZD are a series of NNW-SSE trending gentle ridges separated by shallow troughs (Fig. 2). This

alternating ridge and swale topography reflects distinct beds of the Chuckanut Formation where

weaker layers were preferentially scoured by glacial ice (Brunengo, 2001).

The likely source zone for at least one of the large failure events is a wedge-shaped

bedrock hollow (Figs. 2, 3) in steeply dipping beds of the Chuckanut Formation with a top

elevation of 625 m asl and a bottom elevation of 270 m. The hollow is defined by two distinct

intersecting rupture surfaces: a steep (~48°) dip slope (~300 m high) on its southeast side and a

cross-cutting lateral escarpment on its northwest side (Figs. 2, 3) that varies in height from 75-

5

110 m. The base of the hollow is a large low-gradient bench formed roughly 240 m above the

valley floor.

The runout zone of the complex comprises at least three landslide debris lobes that

accumulated on the floor of the South Fork valley (Figs. 2, 3). The debris lobes are

distinguishable from the adjacent floodplain by their raised relief and irregular hummocky terrain

and from one another based on cross-cutting morphology of their outer margins and by distinct

morphologic textures. The total area covered by the landslide debris is approximately 4.7 km2.

The morphologically oldest deposit, Debris Lobe 1, is ~1.8km long, approximately 0.6-

0.9 km wide and has an area of at least 0.92 km2 (Fig. 3). This oldest lobe is characterized by a

relatively smooth surface with only a small area of hummocky terrain preserved, mainly in the

northern half of the deposit. The deposit exhibits post-deposition gullying and incision along its

western (distal) edge, characteristics that are largely lacking in the other two lobes.

Debris Lobe 2 overlies the northern part of Debris Lobe 1 and is the largest lobe (Fig. 3).

It is 2.7 km long, approximately 1-1.5 km wide and has an area of ~3.5 km2. The South Fork

appears to have eroded the distal-most edge of this lobe, therefore its ultimate extent is uncertain.

The deposit is quite hummocky with a high density of steep-sided hills, closed depressions and

irregular arcuate ridges. Several of the closed depressions are occupied by small perennial bodies

of water.

Debris Lobe 3 is the youngest lobe and is 1.3 km long, 0.5-0.75 km wide, and has an area

of ~0.45 km2 (Fig. 3) This deposit clearly overprints the northwestern part of Debris Lobe 2 as

well as an alluvial terrace riser associated with the North Fork Nooksack River, both of which

project under the deposit. The individual hummocks on this lobe are smaller both in local relief

6

and in dimensions than the hummocks on Debris Lobe 2. This lobe contains at least three small

permanent bodies of water including two perennial lakes and a small bog.

Bedrock Geology

The source rock for the VZLC (Fig. 1) is the Eocene Chuckanut Formation, a sequence of

alternating fine-grained and coarse-grained alluvial strata (Dragovich et al., 1997). Several

studies (e.g., Johnson, 1982; Mustoe, 1997) have divided the Chuckanut Formation into the

Bellingham Bay, Slide, Padden, Maple Falls, and Governors Point members. The entire source

zone of the VZLC is mapped as Bellingham Bay Member (Dragovich et al., 1997); however it is

within 0.5 km of the contact with the Slide Member. Some workers, however, (e.g., Mustoe,

1997; Johnson, 1982) have suggested that these two members are at least partially coeval,

representing the same fluvial system. The only sedimentological difference between the two

members is an overall finer texture of the Slide Member, which rarely contains strata coarser

than medium-grained sandstone (Johnson, 1984). Both units are thick- to thin bedded, well

sorted, rounded to subrounded micaceous fine-grained arkose sandstone alternating with

abundant mudstone, siltstone, and minor coal intervals (Johnson, 1982, 1984; Dragovich et al.,

1997). The sandstone beds throughout the formation consist dominantly of plagioclase,

potassium feldspar, monocrystalline quartz, and biotite with lesser muscovite, chert, and lithic

fragments (Dragovich et al., 1997). The Bellingham Bay Member occasionally contains

polymictic conglomerate consisting of clasts from neighboring units such as the Darrington

Phyllite (Johnson, 1982). The most extensive intact bedrock outcrops in the region occur near the

landslide headscarp where alternating sandstone and finer grained strata strike to the NNE to NE

and dip between 40° and 53° to the NW.

7

Schmidt and Montgomery (1994) identified 34 mountain-front-scale landslides that

initiated within the Chuckanut Formation. This unit is particularly susceptible to failure because

extensive folding has caused widespread fracturing and created abundant steep dip-slopes with

highly variable material strengths exposed in a high-relief glacially scoured landscape (Schmidt,

1994; Brunengo, 2001). Despite its local susceptibility to failure, the resistant Chuckanut

Formation forms some of the steepest and most prominent ridges in the region, whereas weaker

low-grade metamorphic rocks (such as the Darrington Phyllite) make up many of the gentler

slopes (Brunengo, 2001). Mountain front-scale mechanical rock strength is highly influenced by

the material properties of its weakest zones (Badger, 2002). In the case of the Chuckanut

Formation, the dominant discontinuities that modulate slope stability are thin intervals of shale,

mudstone and coal, as well as extensive bedrock fractures (Schmidt, 1994).

Structural Geology and Regional Tectonics

Tertiary deformation compressed much of the region into large N-S and NW-SE trending folds,

primarily plunging towards the north and northwest (Brown and Dragovich, 2003).

There are also several Tertiary faults in the vicinity of the VZLC (Fig. 1), some of which have

been reactivated in the Holocene under a new tectonic stress regime (Sherrod et al., 2013). These

faults include the NE-SW trending Boulder Creek Fault, the NW-SE and N-S trending Canyon

Creek Fault, the NE-SW trending Smith Creek Fault, and the low-angle McCauley Creek Thrust

Fault (Sherrod et al., 2013).

Along the Cascadia margin, oblique subduction of the Juan de Fuca plate beneath the

North American plate has created a complex and seismically active volcanic arc and

convergence zone (Wells and McCraffrey, 2013). The foothills of northwestern Washington and

8

the Nooksack River Basin lie on the actively migrating fore-arc, an upper-plate block being

rotated and compressed against a rigid backstop of the Canadian Coast Mountains (Wells et al.,

1998). This block migration causes N-S shortening that is accommodated along numerous

shallow-crustal faults throughout the Puget Sound region (Mazzotti et al., 2002). In the foothills

of the North Cascades, this crustal seismicity and deformation is focused along reactivated

Tertiary faults including the Boulder Creek, Canyon Creek, McCauley Creek, and Smith Creek

Faults, all of which are in close proximity to the VZLC (Sherrod et al., 2013).

Glacial Geology

Episodic continental glaciation has also strongly influenced the present landscape. During much

of the Pleistocene, ice from the Cordilleran Ice Sheet extensively scoured and oversteepened the

valley walls of the Nooksack River Basin and surrounding areas (Kovanen and Easterbrook,

2001). This substantial glacial erosion combined with interglacial river incision, particularly at

the base of slopes, left many of the regional mountain fronts at or above their maximum

threshold hillslope height and gradient, making them highly susceptible to failure (Schmidt and

Montgomery, 1994; Montgomery, 2002).

Landslide Triggering

A landslide trigger is an external stimulus such as seismic shaking, intense rainfall, rapid

snowmelt, volcanic eruption, or a freeze-thaw event that causes immediate or progressive slope

failure. Triggering-event constraints are crucial to understanding hazard susceptibility because

they provide critical information about threshold conditions for previous failures and allow

9

investigators to estimate probability of failure and factor of safety at a given site (Turner and

Jayaprakash, 1996).

Seismic Triggering

Earthquake scenario reports (e.g., WADNR, 2012; 2013) and my own preliminary limit

equilibrium factor-of-safety analysis demonstrate that local seismically active crustal faults can

potentially produce earthquakes with the peak ground accelerations (>0.2 g) required to trigger

deep-seated bedrock landslides. Trenching data from the Kendall scarp of the Boulder Creek

Fault shows evidence for at least three shallow-focus Holocene earthquakes that were likely Mw

> 6.3, capable of producing damaging ground shaking in northwest Washington (Barnett et al.,

2006; Siedlecki, 2008). Previous studies (e.g., Engebretson et al., 1996; Pringle et al., 1998) have

noted the high concentration of deep-seated bedrock landslides in the Nooksack River Basin

(Fig. 1) and have suggested a causal relationship with nearby shallow-crustal faults.

In addition to shaking from local crustal faults, a partial or complete rupture of the

Cascadia Subduction Zone will produce strong ground shaking throughout the region with a peak

ground acceleration of at minimum 0.2 g (WADNR, 2013; Schmidt, 1994). At least 19 full-

length ruptures of the subduction zone have occurred during the Holocene, along with numerous

partial ruptures (Goldfinger et al., 2012). The Bonneville (Bridge of the Gods) Landslide

Complex along the Columbia River in southern Washington and the Lake Washington Landslide

in Seattle are two examples of dated landslides in the Pacific Northwest that have been linked to

megathrust events (Pringle and Schuster, 1998).

10

Rainfall Infiltration Triggering

Although often associated with shallow slope failures in loose or unconsolidated materials,

meteorological events can also induce deep-seated bedrock landslides. Intense and/or long-

lasting storms with significant infiltration of rainwater can cause highly elevated groundwater

pressures, which increase the driving forces and reduces the resistive force of friction on

potential failure surfaces (Iverson, 2000). Furthermore, rainfall infiltration into bedrock fractures

can increase the mass of an unstable slope, which drives up the shear stress acting upon it,

reducing its factor of safety (FS = resisting forces/ driving forces) and increasing the probability

of failure.

The VZLC headscarp region is located in an area of recent and active timber harvesting.

Forest practices can exacerbate the effects of rainfall events and increase susceptibility to deep-

seated failure. The removal of vegetation significantly reduces rainfall interception and

evapotranspiration, leading to increased groundwater infiltration and pressures (Keaton et al.,

2014). Logging activity can also increase the risk of shallow failures such as rockfalls, topples,

and debris flows. Tree roots provide support and stability at the ground surface and the removal

of them can reduce “root cohesion” in loose soils and expose fractured rock (Legorreta-Paulin et

al., 2008).

The West Salt Creek Landslide (5/25/2014; 3 fatalities) located in western Colorado is a

recent example of a rainfall infiltration triggered bedrock landslide. This large failure was

induced by two days of heavy rain including a 30-minute period when rainfall intensity was over

20 mm/hr just before the failure (White et al., 2015). The heavy rainfall and overland runoff

infiltrated the shale and marlstone bedrock and rapidly increased pore pressure in and along rock

fractures, joints, and pre-existing shear surfaces (White et al., 2015). Failure occurred along a

11

rotational-slip surface causing the rock mass to rapidly disintegrate. The landslide travelled

nearly 3.5 km, making it the longest and one of the most highly mobile (H/L=0.14) landslides in

Colorado’s history (White et al., 2015).

To assess the potential for hydrometerological events as effective landslide triggers in the

region, my study aims to use 14C ages and the robust paleoclimate record (e.g., Steinman, 2014;

Nelson et al., 2011; Mathewes and Heusser, 1981) for the region to determine if the VZLC

failure events are potentially linked to paleoclimatic trends, events, or other dated landslides in

the region.

Although numerous studies (e.g., Geertsema and Schwab, 1997; Huggel et al., 2012;

Zerathe et al., 2014) have examined short-term (decadal) as well as longer-term (centennial to

millennial) paleoclimatic fluctuations and their implications for landslide activity in many of the

world’s mountain belts, little research of this type has been conducted in the North American

Cordillera. Some studies (e.g., Zerathe et al. 2014) have identified clusters of spatially and

temporally concordant landslides and suggest that they are contemporaneous with known periods

of increased precipitation in the paleoclimatic record (e.g., medieval climate anomaly, little ice

age, ENSO fluctuations). The 14C ages presented by this study will contribute to the expanding

inventory of dated landslides in the region and provide new constraints to help assess the

potential for climatic triggering of some paleo-landslides in northwestern Washington.

Other Possible Triggering Mechanisms

Several other natural factors can lead to the genesis of large rock slope failures including rapid

snowmelt, glacial debuttressing, freeze-thaw events, and volcanic eruptions

Rain-on-snow events and/or rapid snowmelt can often have the same effect as heavy

12

rainstorms by causing significant surface runoff and groundwater recharge (Hewitt et al., 2007).

However, elevated water output from snowpack typically produces a more continuous supply of

water for recharge compared to infiltration solely from rainfall (Horton, 1938). This is especially

pertinent in the western Washington where a high percentage of liquid precipitation is

intercepted and stored by dense forest vegetation and never reaches the ground surface and

allowed to infiltrate (Storck et al. 1998). Triggering of deep-seated landslides by rapid and

continuous recharge is usually caused by a combination of factors including increased mass of

the unstable block, decreased frictional strength, and the propagation of bedrock fractures. The

Racehorse Creek Landslide (1/7/2009; ~5x105 m3; Crider et al., 2009) is a recent example of a

rain-on-snow triggered mass movement in the Chuckanut Formation at a site that has many

similarities to the VZLC (prehistoric large failures, steeply dipping beds, recent ground cracks).

In January 2009, a particularly strong “pineapple express” rainstorm following snowfall caused

failure along existing bedrock joints, which debuttressed an entire adjacent hillside and initiated

failure along a bedding plane discontinuity (Crider et al., 2009).

Glacial debuttressing as a trigger occurs when a previously glaciated landscape, freshly

exposed and lacking the lateral support once provided by glacial ice, is no longer stable (Panek,

2014). Such paraglacial rock failure typically occurs within centuries to a millennium after

deglaciation and this delay can often make it difficult to conclusively rule out other external

triggering mechanisms (Panek, 2014). The sharply defined morphology of the VZLC debris

lobes expressed in lidar imagery, however, suggests that they are considerably younger than the

most recent (late-Pleistocene) deglaciation of the region (Clague et al., 1997; Kovanen and

Easterbrook, 2001).

13

Freeze- thaw cycles can also induce bedrock landslides in mountainous areas. When

stored or flowing water in bedrock fractures, joints, or other discontinuities freezes into ice, it

undergoes volumetric expansion. The increased volume of water causes significantly elevated

pore-water pressures and if complete freezing occurs, the ice itself can exert a pressure on the

surrounding rock (Arsenault and Meigs, 2005). In some cases, these pressures can be enough to

fracture the rock and/or cause slope failure (Matsuoka et al., 1998). Diurnal, seasonal, or long-

term freeze-thaw cycles are highly effective at propagating fractures, which can create

preferential flow paths for infiltrating water. Fractures can also greatly reduce bedrock material

strength and a given slope’s angle of friction and factor of safety (Matsuoka et al., 1998).

Although it cannot be ruled out as a contributing factor, freeze-thaw activity is not likely to have

occurred at the likely depths of the failure surface or at the relatively low elevation of the VZLC.

In particular, the initiation points of slides as massive as those at the VZLC must have been tens

to hundreds of meters below ground, well below the depths reached by transient freezing fronts

in the Pacific Northwest.

Volcanic eruptions can trigger large landslides through ground shaking and the

mobilization of large amounts of loose volcanic debris, although this seems unlikely in the case

of the VZLC because the nearest volcano (Mount Baker) is 25 km away and there are no

significant volcanic deposits reported in the landslide debris. Lastly, it is important to keep in

mind that sometimes there is no apparent attributable trigger for a landslide at all or that failure

may have been caused by a combination of processes but not one individual trigger (Hewitt et

al., 2007).

14

Long-runout landslides and transport mechanisms

High-mobility landslides have long intrigued and perplexed geologists and hazard planners.

Some of the most disastrous and confounding landslide events in history (e.g., Frank Slide, AB;

Oso, WA) were of this variety (Cruden and Hungr, 1996; Iverson et al., 2015). The long travel

distances are often unexpected and make these events exceedingly destructive (Cruden and

Hungr, 1996; Iverson et al., 2015). Despite considerable advances in understanding landslide

mechanisms, there is a general lack of knowledge as to why these events happen where they do

and what processes trigger and drive the rapid propagation of the debris mass (Hungr et al.,

2005).

A commonly used mobility index for landslides is the H/L ratio where H is equal to the

vertical height of the fall and L is equal to the horizontal displacement of debris measured from

the back of the headscarp (Coraminas, 1996). The majority of documented landslides worldwide

have an H/L > 0.6 and some literature (e.g., Hsu, 1973; Scheidegger, 1978) considers anything

with an H/L ≤ 0.3 as a long-runout landslide (Coraminas, 1996). The devastating Oso, WA

debris-avalanche flow (DAF) had an H/L= 0.105 (Iverson et al., 2015), whereas the two largest

events for the VZLC had only a slightly lower estimated mobility values of H/L=0.160 and H/L=

0.187.

Although my study will not test directly specific runout mechanisms, it will provide

insight into the types of triggering events and conditions that could possibly generate

catastrophic high-mobility landslides. Evaluating temporal links to contemporaneous regional

landslides may also shed light on why certain orogenic belts such as coastal ranges of

northwestern North America have large concentrations of long run-out landslides (Hewitt et al.,

2007).

15

Methods

Geomorphic Mapping

During the summer of 2015, I will conduct detailed field mapping of the entire study area

including the head scarp region, potential initiation zones, and runout zones. I will focus much of

my attention on identifying and documenting areas with evidence of recent or ongoing mass

wasting. I will be looking for natural and human-made debris exposures throughout the runout

zone and any other relevant geomorphic features (e.g., evidence of liquefaction, mechanical

striae, intact large boulders), as well as evaluating qualitative indicators of debris movement

(e.g., longitudinal ridges, transverse crevasses) and deposit age (e.g., surface soil development).

Lake Coring

Lidar data acquired in 2013 indicate the VZLC comprises at least three major overlapping debris

lobes and possibly several smaller deposits (Fig. 2). I will obtain minimum limiting ages for the

two younger main lobes of the VZLC by collecting sediment cores from small lakes on the

surface of these deposits. I will focus my coring on the deepest portions of these shallow lakes

because that is where the lacustrine record tends to be the most complete and best preserved. In

order to determine the deepest part of each lake, I will collect detailed bathymetric data using an

inflatable raft, differential GPS, and a handheld bathymetry meter and then create a bathymetric

map using Surfer (v.8.0) data visualization software.

I will collect lake sediments using a modified Livingstone piston corer (Wright, 1967)

from an anchored floating raft to collect multiple cores from each lake with a goal of recovering

the deepest (i.e., oldest) sediments possible. Basal sediments and any terrestrial macrofossils

within them will yield minimum 14C ages for the emplacement of the slide debris (Panek et al.,

16

2014; Lang et al., 1999). Although there can be a lag between the slide failure and initial lake

formation, such delays are typically within the margin of error and uncertainty (decades to

centuries) for radiometric dating techniques (Lang et al., 1999). More relevant to my study is the

potential for a lag between lake formation and initial organic sedimentation in the lake (Lang et

al., 1999). To minimize this potential error, I will collect terrestrial macrofossils (wood

fragments, seeds, needles) from the lowest possible strata of the sediment cores, closest to the

underlying landslide debris. Macrofossils are preferable to bulk sediments because they are less

likely to be affected by contamination or other types of errors (e.g., reservoir effects; Bertolini et

al., 2004). In addition to sampling the cores for 14C dating, I will log the visual stratigraphy as

well as the magnetic susceptibility to help establish more complete sedimentary context for the

post-slide deposition in the lakes. These records will help correlate the various cores in each

lake, as well as possibly correlating between lakes. It is also possible that lake sediments on older

slide surfaces may preserve event-strata related to emplacement of younger slide events.

Natural Exposure Dating

At the distal edges of the two largest landslide lobes (Fig. 2), there are natural fluvial incisions of

the landslide debris (Fig. 4) along the South and North forks of the Nooksack River. This

summer (2015), I will investigate these exposures and any others for visible stratigraphy and

collect any datable organic materials included within, or below any exposed slide debris. Basal

contacts, if exposed, will provide valuable constraints for thickness and volume estimates of the

landslide debris lobes (Orwin et al., 2004). Radiocarbon dating of terrestrial macrofossils or

buried logs embedded in the debris can provide direct age constraints on the actual failure event

(Bertolini et al. 2004). The accuracy of the event age depends on the wood fragment or log

17

coming from vegetation that was killed by the landside and incorporated into the debris (Lang et

al., 1999). Outermost rings of intact whole logs are preferable to wood fragments because they

would most closely record the actual kill-date of the tree (Bertolini et al., 2004). For fragments, it

is impossible to tell what part of the tree the wood came from. If the sample was sourced from

the inner rings where 14C had already started to decay, it would give an age that is some

unknown number of years older than the actual slide event (Lang et al., 1999). If I am only able

to find and date wood fragments, the ages will be considered a maxima. I will travel to Lawrence

Livermore National Laboratory (LLNL) in the fall or winter of 2015 to prepare and conduct

accelerator mass spectrometry analysis on all 14C samples from both natural exposures and lake

sediment cores.

Slide Volume Estimations

I will estimate the volume of each debris lobe based on thickness estimates from residential well

and gas-pipeline logs, natural exposures, and projections of underlying fluvial terraces imaged in

lidar data (Orwin et al., 2004). Surface topography and areal extents of each lobe are well

defined in the lidar data and high-resolution aerial photography. I will also estimate the volume

from the source area by estimating the pre-event topography and using the cut-fill spatial analyst

tool in ArcGIS on a lidar derived DEM. These calculations will provide an independent test of

my debris-field volume calculations. I expect some discrepancy in the two volume calculations

because erosion has likely removed some debris at distal edge of the largest lobe (Fig. 2).

Counteracting that effect however is the increase in volume that would accompany

disaggregation of the original bedrock source during sliding.

18

Active Monitoring

For years, residents of the rural community of Van Zandt, built on top of and adjacent to the

main VZLC debris (Fig. 2), have reported loud noises as well as fresh exposures of rock and

debris in the head scarp area (Oral Communication, Andy Wiser; Wayne Harrison, 2014).

Previous workers (e.g., Brunengo, 2001) have also identified recent tension cracks (Figs. 5, 6) in

various locations near the head scarp, suggesting ongoing deformation and the potential for

progressive slope failure. For this study, I will evaluate the relationship between shallow

microseismicity and tension gaps in the headwall region to test the potential for future hazards,

because large landslides often initiate with slow deformation or creep that culminates in

catastrophic failure (Hewitt et al., 2007). Landslide hazard assessment requires the knowledge of

the age, frequency, and volume of previous events, the identification of slopes prone to failure,

and data on active deformation in potential initiation zones (Orwin et al., 2004).

In the spring of 2015, an array of four high-sensitivity Guralp CMG-6TD seismometers

was installed in and around the VZLC source zone (Figs. 2, 7). Over the following year, these

instruments will record shallow microseismicity related to mass movements including fracturing

of the rock mass, rockfall, or deep-seated gravitational shearing. One of the major advantages of

seismic monitoring over other methods is that it allows for continuous measurement and precise

event timing (Helmstetter and Garambois, 2010).

The seismic data will be downloaded monthly using a field laptop computer and

Seismometer Configuration Real-Time Acquisition and Monitoring (SCREAM) software. I will

use a Matlab script to produce spectrogram graphs and Antelope software to produce waveform

graphs of the seismic data. I will evaluate them to estimate relative location and the timing of any

possible “events of interest” that are recorded on at least two of the four stations. After

19

attempting to relate these “events of interest” to any global, regional, and local earthquakes as

well as any known anthropogenic activities, I will eliminate any signals that are clearly not

sourced from the landslide mass. Remaining signals will be categorized based on their

spectrogram and waveform signals. I will then compare the results to preceding and concurrent

hydrometeorological records from the nearby Lawrence and Clearbrook rain gauges to

investigate if the frequency (number of) of seismic events is related to rainfall duration, intensity,

cumulative rainfall, or any combination of these parameters.

To further constrain progressive strain in the source zone, in early October 2015, I will

install Unimeasure JX-PA potentiometric extensometers in several tension gaps in the headwall

region to record displacement and activity of deforming blocks (Fig. 7). Because these

extensometers will be connected to timed dataloggers making hourly measurements, I will be

able to evaluate both whether the fractures are actively forming as well as whether such

movement is episodic or continuous. In addition, I will be able to test if they are linked to

microseismicity recorded by the seismometers. The extensometers have maximum spans of 2-4

m, a standard measurement range of 120 mm, and repeatable precision of 0.15 mm. The main

instrument body can be bolted directly to bedrock and has an extendable draw wire with an

anchor that can likewise be bolted onto rock (Fig. 7). In reconnaissance mapping, I have located

three tension gaps in the head scarp region (Figs. 5, 6) that are well suited for measuring

displacement in the dominant failure direction (WNW-directed slip down the dip of bedding

planes). Sheared soil and tree roots (Fig. 6) indicate that these tension fractures have been

progressively forming and recently active, probably within the last few years. I will download

the data monthly from two Campbell Scientific dataloggers using a field laptop computer and

Campbell Scientific Loggernet software.

20

Anticipated Results

The radiocarbon ages I hope to produce in my study will provide new constraints on the ages of

the two largest debris lobes of the VZLC. The efforts of this project will produce at least limiting

ages and potentially “event” ages for the triggering of large rock slope failures of the VZLC.

These ages may correlate with known paleoseismic events, paleoclimatic intervals and/or other

regional dated landslides (e.g., Pringle et al., 1998; Orwin et al., 2004; Pringle and Schuster,

1998; Engebretson et al., 1996). I expect to find similar minimum ages from the two coring

targets on the largest debris lobe and for these dates to be distinct from the minimum age found

for the youngest debris lobe (Fig. 2). I anticipate that the two main debris lobes will be younger

than ca. 2,000 years B.P. based on their sharply defined topography, thin soils, and poorly

developed drainage network as apparent both on lidar imagery (Fig. 2) and in the field.

Although it is possible that the source zone may remain quiescent over the relatively

short length of my study, I am hopeful that the active monitoring component of this study will

provide evidence for ongoing deformation in the source area of the VZLC. I expect that the

seismometers will show intermittent high-frequency seismic events at all four stations,

potentially indicating local rockfalls, and bedrock fracturing or creep. I expect that the installed

wire extensometers will show that tension gaps in the head scarp region are expanding as active

blocks are gravitationally slumping along discontinuity slip surfaces. If active, it is reasonable to

expect that any movement will be seasonally variable, with increased activity in the wetter and

colder winter months, especially during strong storms with heavy precipitation and rapid

groundwater recharge, or during periods of diurnal freeze-thaw cycles. It will be particularly

crucial to determine if displacement along tension cracks is synchronous with shallow ground

movements recorded by the four seismometers, particularly if it occurs during or shortly after a

21

large rain event. Although it may be difficult to precisely constrain rainfall threshold conditions,

this study will shed light on which parameters (intensity, duration, cumulative rainfall) have the

greatest influence on mass movements and are most likely to trigger larger failures in the VZLC

and other regional deep-seated bedrock landslides.

Potential Problems

One of the greatest hurdles to completing all of the goals of this project is landowner permission

and access. The Washington State Department of Natural Resources (DNR) owns the majority of

the source area and has granted me permission to access, install instruments, and conduct lake

coring on their land (Appendix A). The runout zone, however, is divided into numerous small

parcels owned by dozens of different private landowners. Permission to core at least one lake on

each of the two main debris lobes is the most critical, along with access to any natural exposures

of the landslide debris. General access to land in the runout zone is also important for completing

geomorphic mapping. To date, I have had good success gaining access to many of the crucial

regions on the slide.

Lake sediment coring can present a range of challenges involving the sediment and lake

characteristics, the equipment, and sediment recovery. For example, one of the lakes that is

located at the base of a steep slope (Fig. 2) may have been repeatedly affected by post-slide

rockfall or tree-falls. The Livingston coring system is not well suited to core through sediments

that contain coarse debris or large tree trunks; in such a situation, it may be difficult or

impossible to reach the basal most sediments most closely tied to the slide.

Reservoir effects are often a concern with 14C dating in terrestrial lakes or ponds. Even

small amounts of carbonate minerals or other sources of “old” carbon (radiocarbon dead)

22

percolating into the lake can cause 14C ages on aquatic plants to be inaccurate (Wohlfarth et al.,

1998). Although the Chuckanut Formation does not contain anything more than trace amounts of

carbonate minerals, it does have an abundance of ancient organic-rich strata (Mustoe, 1997).

The sensitive nature of the Guralp CMG-6TD seismometers makes them highly effective

for recording small-scale proximal ground movements. However, this sensitivity can also cause

the instruments to record unwanted anthropogenic “noise” which can dilute the seismic signal.

All four instruments are within 1.5km of human habitations and are in an area that includes

active logging, hunting, agriculture, a railroad, and a state highway (Figs. 2, 3).

I also acknowledge that external factors (e.g., weather) over the course of my study may

not reach threshold conditions needed to cause responses measurable with my instrumentation.

For example, the strong El Nino forecast for the coming winter (2015-2016) will likely be

warmer and dryer than average, significantly reducing the effects related to both heavy

precipitation as well as freeze-thaw events.

Project Timeline

The array of four seismometers was installed during the spring of 2015. I will complete monthly

data downloads from these instruments through at least the spring of 2016. Lake coring, field

mapping, and sample collection from debris exposures will commence in the summer of 2015

and continue into the fall. I will install the wire extensometers in early October and will collect

those data at the same time I download data from the seismometers. 14C analysis will likely take

place during the fall or winter of 2015, depending on the schedule of Lawrence Livermore

National Lab. Interpretation of both the 14C results and active monitoring data will take place

during the spring of 2016.

23

References Cited

Arsenault, Ann M., and Andrew J. Meigs. 2005."Contribution of deep‐seated bedrock landslides to erosion of a glaciated basin in southern Alaska." Earth Surface Processes and Landforms 30, no. 9: 1111-1125. Atwater, Brian F., Bobb Carson, Gary B. Griggs, H. Paul Johnson, and Marie S. Salmi. 2014.

“Rethinking Turbidite Paleoseismology along the Cascadia Subduction Zone.” Geology 42 (9): 827–30.

Badger, Thomas C. 2002. “Fracturing within Anticlines and Its Kinematic Control on Slope

Stability.” Environmental & Engineering Geoscience 8 (1): 19–33. Barnett, E. A., H. M. Kelsey, B. L. Sherrod, R. J. Blakely, J. F. Hughes, E. R. Schermer, R. A.

Haugerud, C. S. Weaver, and E. Siedlecki, 2006, Active faulting at the northeast margin of the greater Puget lowland: A paleoseismic and magnetic-anomaly study of the Kendall scarp, Whatcom County, Northwest Washington, Eos Trans. AGU, 87(52), Fall Meet. Suppl., Abstract S31A-0183.

Bertolini, Giovanni, Nicola Casagli, Leonardo Ermini, and Claudio Malaguti. 2004.

“Radiocarbon Data on Late glacial and Holocene Landslides in the Northern Apennines.” Natural Hazards 31 (3): 645–62.

Blais-Stevens, Andrée, Reginald L. Hermanns, and Courtney Jermyn. 2011. “A 36Cl Age

Determination for Mystery Creek Rock Avalanche and Its Implications in the Context of Hazard Assessment, British Columbia, Canada.” Landslides 8 (4): 407–16.

Brown, Edwin H., and Joe D. Dragovich, 2003. Tectonic elements and evolution of northwest Washington. Washington Division of Geology and Earth Resources. Brunengo, M.J. Guidebook for the Northwest Geological Society Field Trip: June 2-3, 2001: Part 2. The Van Zandt Dike Landslide. Northwest Geological Society (2001). Clague, John J., R. W. Mathewes, J.-P. Guilbault, I. Hutchinson, and B. D. Ricketts. 1997. “Pre-

Younger Dryas Resurgence of the Southwestern Margin of the Cordilleran Ice Sheet, British Columbia, Canada.” Boreas 26 (3): 261–78.

Corominas, J, J Moya, A Lloret, J. A Gili, M. G Angeli, A Pasuto, and S Silvano. 2000.

“Measurement of Landslide Displacements Using a Wire Extensometer.” Engineering Geology 55 (3): 149–66. doi:10.1016/S0013-7952(99)00086-1.

Crider, Juliet G., Tucker, David, Linneman, Scott L. Clark, Doug, 2009. "The 2009 Racehorse Creek Landslide: Forensic Dynamics of a Large, Complex Catastrophic Mass Movement." 2009 Portland GSA Annual Meeting.

24

Cruden, D. M., and O. Hungr. 1986. “The Debris of the Frank Slide and Theories of Rockslide–avalanche Mobility.” Canadian Journal of Earth Sciences 23 (3): 425–

Dragovich, J.S. 1997, Geologic map and interpreted geologic history of the Kendall and Deming 7.5-minut quadrangles, western Whatcom County, Washington. Washington, Division of Geology and Earth Reseurces. Engebretson, D.C., Easterbrook, D.J., and Kovanen, D.J. 1996, Triggering of very large, deep-

seated, bedrock landslides by concentrated, shallow earthquakes in the North Cascades, Washington: Geological Society of America Abstracts with Programs, v. 28, no. 5, p. 64.

Geertsema, Marten, John J. Clague, James W. Schwab, and Stephen G. Evans. 2006. “An

Overview of Recent Large Catastrophic Landslides in Northern British Columbia, Canada.” Engineering Geology, Large Landslides: dating, triggering, modelling, and hazard assessment, 83 (1–3): 120–43.

Geertsema, M., and Schwab, J.W., 1997 "Retrogressive flowslides in the Terrace-Kitimat, British Columbia area: from early post-deglaciation to present and implications for future slides." Proceedings of 11th Vancouver geotechnical society symposium, Vancouver. Goldfinger, Chris, C. Hans Nelson, and Joel E. Johnson. 2012. “Cascadia Great Earthquake

Chronology Based on the Turbidite Event Record.” Boulder, CO, United States: Geological Society of America GSA.

Hewitt, Kenneth, John J. Clague, and John F. Orwin. 2008. “Legacies of Catastrophic Rock

Slope Failures in Mountain Landscapes.” Earth-Science Reviews 87 (1–2): 1–38. Horton, R.E., 1938. “Phenomena of the contact zone between the ground surface and a layer of

melting snow” Assoc. Internationale d’Hydrol Scientifique, Paris 244: 545-561 Hsü, K. J. 1975. “Catastrophic Debris Streams (Sturzstroms) Generated by Rockfalls.”

Geological Society of America Bulletin 86 (1): 129–40. Hsu, K.J. , 1978. Albert Heim: observations on landslides and relevance to modern interpretations. In: Voight, B. (Ed.), Rockslides and Avalanches. In: Natural Phenomena, vol. 1. Elsevier, Amsterdam, pp. 71–93. Huggel, Christian, John J. Clague, and Oliver Korup. 2012. “Is Climate Change Responsible for

Changing Landslide Activity in High Mountains?” Earth Surface Processes and Landforms 37 (1): 77–91.

Hungr, O., Evans, S.G., 2004. Entrainment of debris in rock avalanches: an analysis of the long- runout mechanism. Geol. Soc. Am. Bull. 116, 1240–1252. Hungr, Ol., Fell, R.,, Couture, R.,and Eberhardt, E. 2005. Landslide Risk Management. CRC

Press.

25

Johnson, S.Y., 1984, thesis, University of Washington, Seattle (1982); Can. J. Earth Sci. 21, 92 Iverson, R. M. 2000. “Landslide Triggering by Rain Infiltration.” Water Resources Research 36

(7): 1897–1910. Iverson, R. M., D. L. George, K. Allstadt, M. E. Reid, B. D. Collins, J. W. Vallance, S. P.

Schilling, et al. 2015. “Landslide Mobility and Hazards: Implications of the 2014 Oso Disaster.” Earth and Planetary Science Letters 412 (February): 197–208.

Johnson, Samuel Y. 1984. “Stratigraphy, Age, and Paleogeography of the Eocene Chuckanut

Formation, Northwest Washington.” Canadian Journal of Earth Sciences 21 (1): 92–106. Keaton, J. R. et al. The 22 March Oso Landslide, Snohomish County, WA; GEER Report

Summary. NSF Geotechnical Extreme Events Reconnaissance, 172 pp. (2014). Kovanen, D. J., and D. J. Easterbrook. 2001. “Late Pleistocene, Post-Vashon, Alpine Glaciation

of the Nooksack Drainage, North Cascades, Washington.” Geological Society of America Bulletin 113 (2): 274–88. doi:10.1130/0016-

Lang, Andreas, José Moya, Jordi Corominas, Lothar Schrott, and Richard Dikau. 1999. “Classic

and New Dating Methods for Assessing the Temporal Occurrence of Mass Movements.” Geomorphology 30 (1–2): 33–52.

Legorreta-Paulin, Gabriel, Carol Serdar, and Venice Goetz. 2008. “An Overview of the

Washington State Department of Natural Resources, Forest Practices Division’s Landslide Hazard Zonation Project Methodology and Products for Identification of Unstable Slopes.” Boulder, CO, United States: Geological Society of America GSA.

Mathewes, Rolf W., and Linda E. Heusser. 1981. “A 12 000 Year Palynological Record of

Temperature and Precipitation Trends in Southwestern British Columbia.” Canadian Journal of Botany 59 (5): 707–10.

Matsuoka, Norikazu, Kazuomi Hirakawa, Teiji Watanabe, Wilfried Haeberli, and Felix Keller. "The role of diurnal, annual and millennial freeze-thaw cycles in controlling alpine slope instability." In Proceedings of the Seventh International Conference on Permafrost, Yellowknife, Canada, Collection Nordicana, vol. 57, pp. 711-717. 1998. Montgomery, David R. 2002. “Valley Formation by Fluvial and Glacial Erosion.” Geology 30

(11): 1047–50. Mustoe, G.E, 1997 Paleogeography and Paleontology of the Early Tertiary Chuckanut Formation, Northwestern Washington, Washington Geology, v.25, no. 3, pg.3 Nelson, Daniel B., Mark B. Abbott, Byron Steinman, Pratigya J. Polissar, Nathan D. Stansell,

26

Joseph D. Ortiz, Michael F. Rosenmeier, Bruce P. Finney, and Jon Riedel. 2011. “Drought Variability in the Pacific Northwest from a 6,000-Yr Lake Sediment Record.” Proceedings of the National Academy of Sciences 108 (10): 3

Orwin, John F., John J. Clague, and Robert F. Gerath. 2004. “The Cheam Rock Avalanche,

Fraser Valley, British Columbia, Canada.” Landslides 1 (4): 289–98. Pánek, Tomáš. 2014. “Recent Progress in Landslide Dating A Global Overview.” Progress in

Physical Geography, October, 0309133314550671. Pringle, P. T.; Schuster, R. L., 1998, A new radiocarbon age for the Bonneville landslide, Columbia River Gorge, Washington [abstract]: Association of Engineering Geologists, 41st Annual Meeting, Program with Abstracts, p. 116. Pringle, P.T., Schuster, R.L., and Logan, R.L.1998, New radiocarbon ages of major landslides in

the Cascade Range, Washington, Washington Geology, v.26, p. 31-39. Scheidegger, A., 1973. On the prediction of the reach and velocity of catastrophic landslides. Rock Mech. 5, 231–236 Schmidt, Kevin M., 1994, Mountain Scale Strength Properties, Deep-Seated Landsliding, and Relief Limits, Thesis (M.S), University of Washington, Seattle, WA--Digital, Available at: http://www.dnr.wa.gov/publications/fp_tfw_sh10_95_001.pdf Schmidt, Kevin M., and David R. Montgomery. 1995. “Limits to Relief.” Science 270 (5236): 617–20. Sherrod, Brian L., Elizabeth Barnett, Elizabeth Schermer, Harvey M. Kelsey, Jonathan Hughes, Franklin F. Foit, Craig S. Weaver, Ralph Haugerud, and Tim Hyatt. 2013. “Holocene Tectonics and Fault Reactivation in the Foothills of the North Cascade Mountains, Washington.” Geosphere 9 (4): 827–52. Siedlecki, E. M. 2008, The geometry and earthquake history of the Boulder Creek fault: A

paleoseismic, seismic refraction and ground penetrating radar study, Whatcom County, Washington, Thesis (M.S)--Western Washington University, Bellingham, WA-Print

Steinman, Byron A., Mark B. Abbott, Michael E. Mann, Joseph D. Ortiz, Song Feng, David P. Pompeani, Nathan D. Stansell, Lesleigh Anderson, Bruce P. Finney, and Broxton W. Bird. 2014. “Ocean-Atmosphere Forcing of Centennial Hydroclimate Variability in the Pacific Northwest.” Geophysical Research Letters 41 (7): 2014GL059499. Steinman, Byron A., Mark B. Abbott, Michael E. Mann, Nathan D. Stansell, and Bruce P.

Finney. 2012. “1,500 Year Quantitative Reconstruction of Winter Precipitation in the Pacific Northwest.” Proceedings of the National Academy of Sciences 109 (29): 116. 19–23.

Storck, P., Bowling, L., Wetherbee, P., and Lettenmaier, D., 1998, Application of a GIS-based

27

distributed hydrology model for prediction of forest harvest effects on peak stream flow in the Pacific Northwest: Hydrological Processes, v. 12, p. 889–904. Tabor, R., Haugerud, R., Hildreth, W., and Brown, E., 2003, Geologic map of the Mount Baker 30- by 60-minute quadrangle, Washington: Geologic Investigations Series I-2660. US Geological Survey. Turner, A. Keith, and G. P. Jayaprakash. 1996. “LANDSLIDES: INVESTIGATION AND

MITIGATION. CHAPTER 2 - Assessment.” Transportation Research Board Special Report, no. 247. http://trid.trb.org/view.aspx?id=462499.

Varnes D.J.,1978. Slope movement types and processes. In: Schuster RL, Krizek RJ (eds). Landslides, analysis and control, special report 176: Transportation research board, National Academy of Sciences, Washington, DC., pp. 11–33 WADNR seismic scenario report for a M=6.8 Earthquake on the Boulder Creek Fault, Northwest Washington, 2012, Division of Geology and Earth Resources. WADNR seismic scenario report for a M=9.0 Cascadia Subduction Zone Earthquake, 2013, Division of Geology and Earth Resources. WADNR Significant Deep-Seated Landslides in Washington State – 1984 to 2014, 2014, Division of Geology and Earth Resources. http://wa-dnr.s3.amazonaws.com/publications/ger_list_large_landslides.pdf Wells, Ray E., Craig S. Weaver, and Richard J. Blakely. 1998. “Fore-Arc Migration in Cascadia and Its Neotectonic Significance.” Geology 26 (8): 759–62. Wells, Ray E., and Robert McCaffrey. 2013. “Steady Rotation of the Cascade Arc.” Geology 41 (9): 1027–30. White, J.L., Morgan, M.L, and Berry, K.A., “The West Salt Creek Landslide” A Catostrophic Rockslide and Rock/Debris Avalanche in Mesa County, CO” Colorado Geological Survey Bulletin 55:1-55 Wohlfarth, B., Skog, G., Possnert, G., Holmquist, B., 1998. Pitfalls in the AMS radiocarbon- dating of terrestrial macrofossils. J. Quat. Sci. 13, 137–145. doi:10.1002/(SICI)1099- 1417(199803/04)13:23.0.CO;2-6 Wright, H.E., 1967. A square-rod piston sampler for lake sediments. Sediment. Res. 37, 975– 976. Zerathe, S., Lebourg, T, Braucher T., , and Bourlès, D. 2014. “Mid-Holocene Cluster of Large-

Scale Landslides Revealed in the Southwestern Alps by 36Cl Dating. Insight on an Alpine-Scale Landslide Activity.” Quaternary Science Reviews 90 (April): 106–27.

28

Figures

Figure 1. Simplified regional geologic map of the northern Nooksack River Basin. The VZLC is one of several large landslide complexes in the Nooksack River Basin. Multiple reactivated Tertiary faults are within close proximity to the VZLC and the other large landslides in the area (Sherrod et al., 2013). Map data (geologic units and faults) from Dragovich, 1997.

29

0

100

200

300

400

500

600

700

0200400600800100012001400160018002000

Elev

atio

n (m

)

Horizontal Distance (m)

H

L

A A'

Figure 2. Preliminary mapping (a) and longitudinal topographic profile (b) of the Van Zandt Landslide Complex. a. Airborne lidar hillshade map indicates three discrete overlapping landslide debris lobes, a general landslide source zone, several lacustrine coring targets (at least three on each of the younger two lobes), and a colluvial lobe that was likely formed by small-scale failures. Natural exposures of landslide debris along the South Fork Nooksack River may also provide datable material and will aide in debris thickness and volume estimations. Road system provides ready access and indicates residential areas potentially at risk. Trench logs of the pipeline excavation should provide further subsurface constraints. b: Longitudinal topographic profile (A-A’) indicates runout length of debris-lobe 2 and factors H and L along transect A-A’. Vertical exaggeration is zero. Source Information: Lidar data (PSLC, 2006; 2013). Mapping by G. Malick.

a.

b.

30

Figure 3. Headscarp and main debris lobes of the VZLC. Imagery from Google Earth, capture date 5/4/2014, facing west at oblique angle.

Fig 4. Contact between the VZLC debris and underlying alluvial deposits in a river-incised natural exposure along the South Fork Nooksack River.

31

Fig 5. Tension fracture in the headscarp region of the VZLC. Note soil probe for scale.

Fig. 6. Tension Fracture in the headscarp region of VZLC showing torn roots and disturbed trees.

32

Fig 7. Schematic diagram of the VZLC headscarp region with tension fracture perpendicular to the dominant failure direction (down the dip of bedding planes). Thin intervals of shale, mudstone, and coal represent potential rupture surfaces and together with extensive bedrock fractures modulate slope stability in the VZLC. A: Example of seismometer configuration. The instrument is buried ~1 m below the surface and is preferably resting on bedrock. Power is sourced from a 12 V battery and solar panel. B: Example of extensometer configuration. The main instrument body is bolted directly into bedrock and a draw wire extends across the gap and is anchored on the opposite wall of the tension fracture. A 12 V battery supplies power to a data logger and the potentiometric sensor.