Harrichhausen 2024

doi:10.26443/seismica.v2i4.
1177
Inner forearc faults in northern Cascadia do not

accommodate elastic strain driven by the megathrust
seismic cycle
Nicolas Harrichhausen � ∗ 1, 2 , Kristin D Morell � 2 , Christine Regalla � 3
1
Univ. Grenoble Alpes, Univ. Savoie Mont Blanc, CNRS, IRD, Univ. Gustave Eiffel, ISTerre, Grenoble, France, 2 Department of Earth Science,
University of California, Santa Barbara CA, USA, 3 School of Earth and Sustainability, Northern Arizona University, Flagstaff AZ, USA
Author contributions: Conceptualization: N. Harrichhausen, K.D. Morell, C. Regalla. Methodology: N. Harrichhausen. Formal Analysis: N. Harrichhausen.
Writing - Original draft: N. Harrichhausen. Writing - Review & Editing: N. Harrichhausen, K.D. Morell, C. Regalla. Visualization: N. Harrichhausen. Supervision: K.D.
Morell, C. Regalla. Project administration: K.D. Morell, C. Regalla. Funding acquisition: N. Harrichhausen, K.D. Morell, C. Regalla.
Abstract We employ numerical models to explore the connection between subduction zone coupling or Production Editor:
Alice-Agnes Gabriel
megathrust rupture and upper plate faults in the northern Cascadia forearc. Active forearc faults north of the
Christie Rowe
Olympic Peninsula exhibit similar characteristics: west-northwest strike, oblique right-lateral slip senses, and Handling Editor:
low slip rates (<1 mm/yr), but a potential to generate large (M ∼ 7) earthquakes. Previous hypotheses sug- Harold Tobin
gest stress in the upper plate from interseismic coupling or coseismic rupture on the subduction zone inter- Copy & Layout Editor:
face could drive permanent forearc strain. To test these hypotheses, we used a 3D boundary element method Théa Ragon
model to predict slip on the Leech River–Devils Mountain fault system if interseismic coupling or coeseismic
Received:
rupture cause deformation. Our model predicts reverse left-lateral slip if strain results solely from subduction
January 15, 2024
zone coupling, or normal right-lateral slip if the faults accommodate strain from a megathrust rupture. These Accepted:
results contradict observed fault kinematics. By modelling strain partitioning, where only the strain from the May 31, 2024
strike-slip component of subduction zone coupling is accommodated in the forearc, we also predict slip that Published:
is inconsistent with observed fault kinematics. These models challenge the hypothesis that subduction in- July 2, 2024
terface coupling or coseismic rupture are the primary driver of permanent forearc deformation in northern
Cascadia.
1 Introduction is thought to be driven by interseismic coupling on the

subduction zone interface (e.g., McCaffrey et al., 2000;
At subduction zones, partial to full coupling of the Wallace et al., 2004; Loveless et al., 2010; Feng et al.,
down-going oceanic slab against the upper plate results 2012; Delano et al., 2017). Some models argue that
in strain in the upper plate crust. Some of this strain strain resulting from oblique coupling on the megath-
is released as elastic rebound during megathrust earth- rust is partitioned spatially, with strike-slip faulting ac-
quakes (e.g., Dragert et al., 1994; Leonard et al., 2004), commodating margin-parallel strain in the onshore in-
while some may remain as permanent strain resulting ner forearc, and reverse faulting and folding accom-
in faulting and folding in the upper plate (e.g., Bellier modating margin-perpendicular strain occurring in the
et al., 1997; Marshall et al., 2000; Sieh and Natawidjaja, offshore outer forearc (e.g., Kimura, 1986; McCaffrey,
2000). This faulting in the upper plate can be a signifi- 1991; Wang et al., 2007). However, other stresses are
cant source of seismic hazard (e.g., Johnson et al., 1999; also thought to contribute to permanent strain. Some
Quigley et al., 2012; Ryder et al., 2012; Mouslopoulou models argue that forearc faulting could also be in-
et al., 2015), and constraining how strain is partitioned duced by stress changes related to coseismic rupture
into permanent and elastic components is important during large megathrust earthquakes (Loveless et al.,
in understanding this hazard. Additionally, quantify- 2010; Duckworth et al., 2021). Additionally, several sub-
ing permanent deformation in the upper crust can also duction zones have inner forearc maximum horizontal
provide information on what controls subduction zone stress (SHmax ) orientations derived from crustal seismic-
coupling, such as plate roughness and topography, age ity that are perpendicular to plate convergence at the
of the oceanic plate, curvature of the margin, and the margin, contrary to convergence-parallel SHmax orienta-
obliquity of convergence (e.g., Sitchler et al., 2007; All- tions that interseismic coupling is expected to induce
mendinger and González, 2010; Morell et al., 2013). (e.g., Wang, 2000; Townend and Zoback, 2006; Balfour
Despite the importance of understanding permanent et al., 2011; Townend et al., 2012; Dimitrova et al., 2016).
forearc deformation, the processes driving this strain These results suggest permanent forearc strain is not
are not completely understood. At several subduction controlled solely by subduction zone coupling or coseis-
zones, permanent deformation of the upper plate crust mic rupture of the megathrust, and instead may reflect
∗ Corresponding author:
the background stress state of the forearc and arc.
[email protected]
1 SEISMICA | ISSN 2816-9387 | volume 2.4 | 2024

SEISMICA | RESEARCH ARTICLE | Forearc faults and megathrust cycle elastic strain
Our models test if the stress induced by either inter-

seismic subduction zone coupling or coseismic megath-
rust rupture can explain the observed kinematics and
slip history of Holocene-active forearc faults. We use a
3D Boundary Element Method (BEM) model to predict
fault kinematics of the Leech River–Devils Mountain
fault system (LRDM) in northern Cascadia if they re-
sult solely from stress induced by geodetically-derived
interseismic coupling on the subduction zone megath-
rust, or conversely, stress resulting from coseismic slip
on the megathrust (e.g., Delano et al., 2017; Duckworth
et al., 2021). This fault system is unique in that it tran-
sects the entire ∼ 20 to 25 km wide onshore forearc
in northern Cascadia, allowing us to examine margin-
perpendicular variations in forearc strain. We analyze
the effects of strain partitioning by seperately consider-
ing the effects of only the strike-slip component of sub-
duction zone coupling versus unpartitioned coupling.
We then compare these results to previous paleoseismic
studies and instrumental seismicity along this fault sys-
tem to determine if interseismic strain, partitioned in-
terseismic strain, or coseismic strain alone can explain
the observed fault kinematics. These analyses test the
degree to which active forearc faults relieve the stress
imposed by the megathrust throughout its seismic cy-
cle.
2 Tectonic Setting
Figure 1 Tectonic setting of northern Cascadia with se-
At the Cascadia subduction zone, the Juan de Fuca lected GNSS velocity vectors (thin blue arrows) showing mo-
(JdF) plate subducts towards the northeast beneath the tion with respect to stable North America (UNAVCO, McCaf-
North America (NA) plate at a rate of ∼ 40 mm /yr frey et al., 2013). Juan de Fuca plate motion with respect to
(Fig. 1; Yoshioka et al., 2005). Relative plate motion, North America (DeMets et al., 2010) shown in thick black ar-
subduction zone geometry, and subduction obliquity rows. Juan de Fuca slab depth contours (Blair et al., 2011)
vary substantially along strike from south to north along are shown with dotted lines and are 10 km intervals. UTM
the margin. At the southern extent of the subduc- grid coordinates (m) are in NAD83 UTM Zone 10N.
tion zone, the trench is oriented ∼ N–S, and JdF-NA
plate motion is highly oblique, with an ∼2:1 ratio be-
tween slab-strike perpendicular and slab-strike parallel
(right-lateral) components of relative plate motion (e.g., reflecting a cessation of subduction at the boundary be-
DeMets et al., 2010; Finley et al., 2019). In northern tween the JdF and Explorer plates (Savard et al., 2020).
Cascadia, the trench is oriented ∼ NW–SE and JdF-NA The GNSS velocities (Fig. 1) show that the Cascadia
plate motion is near-orthogonal, with a ∼ 4 : 1 ratio subduction zone experiences partial to full interseismic
between the slab-strike perpendicular and left-lateral coupling along the megathrust for the entire length of
slab-parallel components (Fig. 1; e.g., DeMets et al., the subduction zone (Dragert et al., 1994; Hyndman and
2010; Finley et al., 2019). Wang, 1995; Wang et al., 2003; Schmalzle et al., 2014;
Global navigation system satellite (GNSS) surface ve- Li et al., 2018b). Subduction zone coupling occurs off-
locities of the forearc relative to the stable portion of shore of western North America, a region we term the
the NA plate reflect this change in obliquity and the ‘outer forearc’, while onshore beneath the ‘inner fore-
broader tectonics of the region (Fig. 1). At the southern- arc’, there is little to no measurable coupling from GNSS
most extent of Cascadia (Mendocino triple junction), data (Schmalzle et al., 2014; Li et al., 2018b). The latest
surface velocities are directed towards the northwest, geodetic models suggest that the width of megathrust
partially as a result of northwest motion of the Sierra coupling in map view varies from ∼ 50 km in south-
Nevada–Great Valley block and westward motion of the ern Cascadia, to ∼ 100 km near the Olympic Mountains,
Basin and Range (Wells et al., 1998; McCaffrey et al., to ∼ 60 km in northern Cascadia near Vancouver Is-
2000; Wells and McCaffrey, 2013; McKenzie and Fur- land (e.g., Li et al., 2018b) , although these coupling es-
long, 2021). GNSS velocities gradually rotate clock- timations contain substantial uncertainty due to insuf-
wise further north until they are approximately parallel ficient seafloor geodetic data (Wang et al., 2003; Wang
with JdF–NA relative plate motion in northern Oregon and Tréhu, 2016). The wider zone of coupling beneath
(Fig. 1). In northern Cascadia, at the northern limit of the Olympic Mountains is due to a lesser slab dip where
the JdF plate and the subduction zone, northeastward there is a bend in the subducting JdF Plate (Fig. 1; Li
directed velocities are significantly diminished (Fig. 1) et al., 2018b). Slab dip partially controls the width of the
2 SEISMICA | volume 2.4 | 2024
coupled zone because the downdip extent of coupling in Two of these faults, the Leech River and Devils Moun-
warm subduction zones, such as Cascadia, is thought to tain faults, form a continuous ∼W-striking fault system
be thermally controlled by the brittle-ductile transition (LRDM) that dips steeply north (Johnson et al., 2001;
(e.g., Hyndman and Wang, 1993; Hyndman et al., 1997; Li et al., 2018a), and transects the majority of the on-
Oleskevich et al., 1999). The up-dip extent of coupling shore Cascadia forearc across southern Vancouver Is-
may also be thermally controlled, in this instance by the land in Canada, and the San Juan Islands and Puget
depth at which clay-dehydration reactions occur (e.g., Lowland in Washington State (Fig. 2). We chose to use
Vrolijk, 1990; Hyndman and Wang, 1995; Moore and Saf- this structure to model megathrust-coupling induced
fer, 2001). In addition to the width of the coupled zone slip and help constrain the driving force of permanent
varying along strike, the degree to which the plates are deformation in northern Cascadia because of its wide
coupled to each other varies along-strike within the cou- transect and relatively long paleoseismic record span-
pled zone (Schmalzle et al., 2014; Li et al., 2018b). ning the Holocene (e.g., Johnson et al., 2001, 2004a; Per-
Partial coupling in northern Cascadia is character- sonius et al., 2014; Morell et al., 2017, 2018; Barrie and
ized by persistent subduction zone creep, and episodic Greene, 2018; Harrichhausen et al., 2021).
tremor and slip (ETS) that recover a portion of the elas-
tic deformation of the upper crust (McCaffrey, 2009;
Schmalzle et al., 2014). ETS may be influenced by in-situ
dehydration reactions of oceanic crust (Fagereng et al.,
2018; Condit et al., 2020), with increased pore-fluid pres- 2.1 Paleoseismic studies
sure resulting in tremor. Along-strike changes in sed-
iment thickness on the downgoing slab, and lithology Paleoseismic studies along the LRDM, summarized in
and porosity of the upper plate, can affect pore fluid Table 1 and denoted with letters i–vi on Fig. 2a, indi-
generation and pressure; therefore controlling ETS, cate this structure has hosted at least five surface rup-
persistent creep, and the degree of coupling (e.g., Au- turing earthquakes with oblique slip in the Holocene.
det and Bürgmann, 2014; Schmalzle et al., 2014; Wells Evidence for surface-rupturing earthquakes is present
et al., 2017; Savard et al., 2018). Along-strike changes in as offset Quaternary sediments and offset subaqueous
upper plate roughness also affect the degree of coupling banks observed in bathymetry along the eastern WNW-
along the subduction interface (e.g. Wang and Bilek, striking portion of the Leech River fault on southern
2014; Bassett and Watts, 2015; Van Rijsingen et al., 2019). Vancouver Island (Morell et al., 2017, 2018; Harrich-
These changes in subduction zone properties along- hausen et al., 2021), and along the Devils Mountain fault
strike have resulted in spatially variable coupling that offshore of Victoria, British Columbia and in Washing-
is dependant not only on slab dip. ton State (Johnson et al., 2001, 2004a; Personius et al.,
The presence and spatial variability of coupling along 2014; Barrie and Greene, 2018). Paleoseismic trenching,
the Cascadia subduction interface drives elastic defor- geomorphic mapping, and stratigraphic analyses con-
mation such as coastal uplift measured from tide gauges sistently indicate N-side-up motion along both faults
(Burgette et al., 2009), and to some degree, permanent (Table 1). Right-lateral offset of the LRDM is docu-
deformation such as long-term uplift on the Olympic mented in paleoseismic trenches and by offset banks
Peninsula deduced from river terraces (Delano et al., at three locations (ii, iii, and v), with the most un-
2017). Far-field tectonic stresses, such as those related ambiguous observation of strike-slip kinematics along
to Basin and Range extension (in southern Cascadia) the Devils Mountain fault where Personius et al. (2014)
and distributed right-lateral shear between the Pacific used 3D trenching to show right-lateral offset of glacial
and NA plates (e.g., Pezzopane and Weldon, 1993; Wells outwash channels and older bedrock faults. In con-
et al., 1998; McCaffrey et al., 2000, 2007; Wells and Mc- trast, left-lateral offset is documented in a paleoseismic
Caffrey, 2013; McKenzie and Furlong, 2021; Littel et al., trench excavated across a southward branch of the Dev-
2023), and stress related to margin parallel mantle con- ils Mountain fault, the Utsalady Point fault (vi). Earth-
vection (e.g., Sternai et al., 2014; Chen et al., 2016) may quake rupture ages and dated offset stratigraphy sug-
also play a role in driving permanent strain that is ob- gest a minimum slip-rate of 0.05 mm/yr and a max-
served in the upper plate. This permanent forearc de- imum of 1.6 mm/yr, with no spatial pattern in min-
formation is recorded in Cascadia by oroclinal bend- ima and maxima observed along strike of the fault sys-
ing, rock uplift, active crustal faulting, and instrumental tem (Fig. 2b; Table 1). No evidence for Holocene sur-
seismicity (Fig. 2; e.g., McCaffrey, 1992; Loveless et al., face rupture of the western-most W-striking section of
2010; Delano et al., 2017; Malatesta et al., 2018; Finley the Leech River fault has been observed, and this por-
et al., 2019; Harrichhausen et al., 2023). tion of the fault has been considered inactive due to
In the inner forearc of northern Cascadia, north of the observation of Oligocene marine sediments non-
the Olympic Mountains, several active faults accommo- conformably overlying the fault on the west coast of
dating permanent strain have been identified through Vancouver Island (MacLeod et al., 1977; Fairchild and
paleoseismic observations and instrumental seismicity. Cowan, 1982; Groome et al., 2003). However, pro-
These faults form a network of subvertical to steeply N- nounced topographic lineaments following the bedrock
dipping W- to NW-striking structures that have hosted surface trace of this portion of the fault and dense
Holocene surface rupturing earthquakes (Fig. 2a; e.g., vegetation potentially obscuring exposures of faulted
Johnson et al., 2001; Personius et al., 2014; Morell et al., Oligocene sediments, may suggest more recent fault
2018; Schermer et al., 2021; Harrichhausen et al., 2023). slip (Morell et al., 2017).
Figure 2 a) Surface traces and kinematics of Quaternary–Active crustal faults in northern Cascadia highlighting the Leech
River (LRF) and Devils Mountain (DM) faults, combined to make the Leech River-Devils Mountain fault system (LRDM). Paleo-
seismic study results and references (locations shown as red boxes i–vi) are shown in Table 1. Fault traces are from the United
States Geological Survey (USGS) Quaternary fault and fold database. Quaternary reverse slip has been observed on the Seat-
tle fault zone (SFZ) (Johnson et al., 1994, 1999; Nelson et al., 2003; Blakely et al., 2002), right-lateral slip on the North Olympic
fault zone (Nelson et al., 2017; Schermer et al., 2021), and right-lateral oblique slip on the southern Whidbey Island fault zone
(SWIF) (Johnson et al., 1996; Sherrod et al., 2008). b) Upper plate crustal seismicity (< 35 km depth) in northern Cascadia
from January 1, 1970 through April 29, 2015 compiled by Brocher et al. (2017) from the Pacific Northwest Seismic Network
(2022). SHmax directions calculated by Balfour et al. (2011) from clusters of crustal earthquake focal mechanisms are shown
by black arrow pairs. Adapted from Harrichhausen et al. (2021).

Fault Location Type Fault dip Kinematics Slip rate Earthquake ages Reference
LRF i Geo- Subvertical N-side up N/A N/A Morell et al. (2017)
morphology
LRF i Trench Subvertical N-side up 0.2-0.3 mm/yr (ver- 1.7 ± 0.1, 2.2 ± Morell et al. (2018)
tical slip) 0.1, 8.7 ± 0.3 ka
LRF ii Trench Subvertical Oblique N/A 9.4 ± 3.4 ka Harrichhausen et al.
right-lateral (2021)
DMF iii Bathymetric, Subvertical Oblique 0.8 − 1.3 mm/yr N/A Barrie and Greene
seismic reflec- right-lateral (2018)
tion data
DMF iv Well logs 45◦ − 75◦ N N-side up 0.05 − 0.30 mm/yr N/A Johnson et al. (2001)
(vertical)
DMF v Trench subvertical to Oblique 0.14 ± 0.1 mm/yr 2.2 ± 0.1, 8.1 ± 0.1 Personius et al.
N-dipping right-lateral (strike slip) ka (2014)
UPF vi Trench Subvertical Oblique left- Mininimum slip rate 100 − 400, 1100 − Johnson et al.
lateral = ∼ 1.6 mm/yr 2200, cal B.P. (2004a)
Table 1 Summary table of paleoseismic studies of the Leech River (LRF) and Devils Mountain (DMF) and Utsulady Point
(UPF) fault systems
2.2 Crustal seismicity and forearc stress field 2.3 GNSS derived upper plate strain
orientations
A combination of GNSS velocity vectors, geologic slip
rates, paleomagnetic declination anomalies, seafloor
spreading rates, and earthquake slip vector analyses
In northern Cascadia, instrumental crustal seismic- together have been used to produce block models of
ity (located at < 35 km depth, recorded by the Pa- northern Cascadia that predict rates and kinematics
cific Northwest Seismic Network between 1970–2015, on active forearc faults (McCaffrey et al., 2007; Evans,
Brocher et al., 2017) is concentrated in a region along 2022). These models separate the Pacific northwest re-
the eastern Leech River fault and the Devils Mountain gion of North America, which encompasses the Casca-
fault (Fig. 2b). In comparison, reduced seismicity along dia subduction zone, into regions of relatively cohesive,
the western Leech River fault (Fig. 2b) may suggest rigid blocks. The Devils Mountain fault is one of the
lower strain rates, which is consistent with this portion block boundaries and the McCaffrey et al. (2007) block
of the fault being inactive. However, crustal seismic- model predicts ∼ 0.4 mm/yr of right-lateral normal slip
ity increases again towards the trench from the onshore along this structure. The strike-slip component of this
portion of the western Leech River fault, resulting in slip-rate is consistent with both paleoseismic studies
a ’seismicity gap’ where there is the reduction in seis- and crustal seismicity. However, the 70% confidence
micity between the inner forearc and the offshore por- interval error ellipse for this vector is large (>> 0.4
tion of the forearc (Fig. 2b; Brocher et al., 2017; Bostock mm/yr) and a left-lateral reverse slip vector is within its
et al., 2019). Seismicity relocated on the eastern WNW- bounds. The Evans (2022) block model discretizes slip
striking portion of the Leech River fault suggests that rates and kinematics along this boundary, and predicts
it is a ∼ 60◦ N-dipping zone of subvertical en échelon ∼ 0.05 to 0.5 mm/yr of reverse right-lateral slip along
faults extending to a depth of ∼ 28 km (Li et al., 2018a). the Devils Mountain fault. The lowest slip-rates, and
the most substantial component of reverse slip, occur
Earthquake focal mechanisms along the LRDM pre- near the intersection between the South Whidbey Island
dominantly show right-lateral slip on steeply dipping fault zone and the Devils Mountain fault and increase
W–NW striking fault planes, and N–S shortening on to ∼0.5 mm/yr towards the west and ∼ 0.2 mm/yr to the
reverse faults (Fig. 2b; Brocher et al., 2017; Li et al., east. Again, the model-derived standard deviations of
2018a). These focal mechanisms are consistent with these slip rates are larger than the absolute values of
oblique right-lateral slip observed in paleoseismic stud- the vectors (∼ 0.1 to 1.0 mm/yr), indicating that oppo-
ies at sites ii, iii, and v (Table 1). NNW–SSE maximum site slip kinematics are within the error bounds of the
horizontal compressive stress (SHmax ) and σ1 directions model.
along the fault system, derived from clusters of crustal GNSS velocities collected throughout the past several
seismicity by Balfour et al. (2011), would also predict decades also suggest N–S shortening across the Olympic
that the LRDM accommodates oblique reverse right- Mountains, which has been interpreted to result from a
lateral slip if the observed stress was relieved by these northward-migrating southern Cascadia forearc (Khaz-
faults (Fig. 2b). However, the SHmax direction derived for aradze et al., 1999; Mazzotti et al., 2002). Oroclinal bend-
the upper crust offshore the western end of the Leech ing, reflected by counterclockwise rotation of the north-
River fault would predict a left-lateral component to slip ern forearc and clockwise rotation of the southern fore-
(Fig. 2b). arc about an axial trace that bisects the Olympic Penin-
which, along with oroclinal bending, may accommo-

date the observed N–S shortening across the Olympic
Peninsula (e.g., Nelson et al., 2017; Finley et al., 2019;
Harrichhausen et al., 2021). This trenchward extrusion
of the forearc would also imply right-lateral slip on the
Leech River fault as the portion of Vancouver Island
south of the Leech River fault is moving west with re-
spect to the northern side of the fault (Fig. 3). Towards
the east along the Devils Mountain fault, a south-to-
north reduction in northward residual velocities sug-
gests N–S shortening perpendicular to the strike of the
Devils Mountain fault (Fig. 3; Khazaradze et al., 1999;
Mazzotti et al., 2002, 2011), predicting reverse slip on the
eastern segment of our modelled fault system.
3 Model description
We used a 3D BEM model (e.g., Crouch and Starfield,
Figure 3 Map showing the tectonic processes of Cascadia 1983; Thomas, 1993; Loveless et al., 2010; Delano et al.,
that may influence right-lateral slip on the LRDM. Clockwise 2017; Duckworth et al., 2021) to calculate stress in a lin-
rotation and northward motion of the southern Cascadia ear elastic half-space representing the upper plate of
relative northern Cascadia induces oroclinal bending and the Cascadia subduction zone (based on algorithms in
westward escape of the northwest Olympic Peninsula and Meade, 2007). The stress results from prescribed slip
western Vancouver Island (Nelson et al., 2017; Finley et al., rates on the subduction zone interface that mimic in-
2019; Harrichhausen et al., 2021). Inset shows how orocli- terseismic coupling, coseismic megathrust rupture, and
nal bending results in strike-slip faulting on each limb of partitioned forearc strain. To estimate the slip rates on
the orocline, similar to flexural slip on a fold. Small blue ar- the LRDM and associated faults, we calculated slip rates
rows show gridded interpolation of the residual GNSS veloc- on meshes representing these faults that are required
ity field with respect to stable North America from (Mazzotti to relieve the stress induced by the tractions prescribed
et al., 2011) and indicate west–northwestward escape of the on the subduction zone mesh. This model has previ-
Olympic Peninsula and western Vancouver Island. Relative ously been used to estimate slip on other crustal forearc
plate motion from (DeMets et al., 2010). Clockwise rotation structures in the Cascadia forearc in Washington State
shown in blue and counterclockwise rotation shown in red. (Delano et al., 2017; Duckworth et al., 2021; Loveless,
UTM grid coordinates (m) are in NAD83 UTM Zone 10N. 2021). We adapted the model to include the faults in our
study area, and to test how partitioned strain is accom-
modated by forearc faulting (archived model files avail-
sula (Fig. 3; Miller et al., 2001; Mazzotti et al., 2003; Mc- able in: Harrichhausen et al., 2024).
Caffrey et al., 2007; Finley et al., 2019), has been sug- The inputs to the BEM model are: A meshed surface
gested to accommodate this shortening (Finley et al., representing the Cascadia subduction zone with pre-
2019; Harrichhausen et al., 2021). Paleomagnetic and scribed dip-slip and strike-slip rates representing inter-
structural geology studies suggest oroclinal bending has seismic coupling, partitioned interseismic coupling, or
been ongoing since 18 Ma (Wells and McCaffrey, 2013; coeseismic rupture (Delano et al., 2017); A meshed sur-
Finley et al., 2019). Right-lateral fault slip on E–W- face(s) representing forearc fault(s); Coefficients and as-
striking faults in northern Cascadia is suggested to re- sumptions governing the elastic properties of the half-
sult from deformation similar to flexural slip on a fold space (Poisson’s ratio and shear modulus). The meshed
limb, during oroclinal bending (Fig. 3; Finley et al., surfaces are discretized as triangular dislocation ele-
2019; Harrichhausen et al., 2021). ments (TDEs) for which slip rates are prescribed or cal-
GNSS residual velocities, where the estimated elastic culated individually. This discretization allows for com-
effect of subduction zone coupling is removed from to- plex surface geometries, and spatially variable coupling
tal GNSS velocities, have also been used to constrain and slip-rate distributions on the modelled structures.
permanent forearc deformation in Cascadia (Khaz-
aradze et al., 1999; Miller et al., 2001; Mazzotti et al.,
3.1 Subduction zone mesh
2003, 2011). These residuals are based on subduction
zone coupling models and the lack of constraint on cou- The Cascadia subduction zone mesh and slip-rates are
pling (e.g., Wang et al., 2003; Wang and Tréhu, 2016) from Delano et al. (2017), with the subduction zone
may inhibit their usefulness. However, some areas of geometry based on slab contours from McCrory et al.
the residual velocity fields predict forearc deformation (2006). Each TDE of the mesh is defined by its own strike
that is consistent with observed forearc fault kinemat- and dip based on the slab contours, and these orien-
ics (Harrichhausen et al., 2021). GNSS residuals pre- tations are used to resolve the strike-slip and dip-slip
dict northwestward motion, or ”escape”, of the Olympic components of deformation. The slip rates on the Cas-
Peninsula and the west coast of Vancouver Island rela- cadia subduction zone mesh equate to the slip-deficit,
tive to stable North America (Mazzotti et al., 2002, 2011), or the rate of potential elastic slip accumulated on the
subduction interface due to coupling in mm/yr. These traces in Washington State were interpolated from John-
slip deficit values have been calculated using a geodeti- son et al. (2001); Personius et al. (2014); Greene et al.
cally constrained block model that uses a GNSS velocity (2018), Personius et al. (2014), and Greene et al. (2018).
field consisting of 1717 stations, in conjunction with in- The trace of the Leech River fault is based on geomor-
dependently (geologically) calculated relative block mo- phic mapping from Graham (2018), Morell et al. (2017,
tions to predict slip on the block boundaries (Meade 2018), and Harrichhausen et al. (2021), as well as ge-
and Loveless, 2009; Delano et al., 2017). This block ologic mapping of the fault from Massey et al. (2005).
model uses potentially active faults as block bound- We also extended the Leech River fault ∼ 40 km off-
aries, and includes a block boundary that is close to shore to the west based on a map of the Leech River
our mapped trace of the LRDM. The coupling model re- fault by MacLeod et al. (1977). Although this portion
solves ∼ −13 to 35 mm/yr of dip-slip deficit and ∼ −35 of the fault may not be active (MacLeod et al., 1977;
to 10 mm/yr strike-slip deficit on the subduction zone Fairchild and Cowan, 1982; Groome et al., 2003), SHmax
interface, where positive slip rates are reverse and left- orientations have been calculated from crustal seismic-
lateral, respectively. The greatest coupling is located ity in this region (Balfour et al., 2011), providing ob-
where the subducting slab geometry changes from ∼N- servations to compare to our modelled fault kinemat-
striking to ∼NW-striking at UTM 530000 m N and the ics. We extracted UTM coordinates (NAD83 Zone 10N)
slab dip is reduced (Fig. 4a). At this location, the strike- from the vertices of these surface traces and used a pub-
slip deficit also changes from right-lateral to left-lateral lished MatLab script (available at https://github.com/
from south to north due to a change in the obliquity jploveless/meshing) to produce a mesh constructed from
of subduction that occurs at the same location. At the 3000 m sided triangles. We used a depth of 28 km and a
southern extent of the modelled slab, the Delano et al. dip of 60◦ N for the fault system based on relocated seis-
(2017) model prescribes an excess of dip slip on the sub- micity on the Leech River fault from Li et al. (2018a).
duction interface (i.e., where the slip deficit is negative). We also use three additional multi-fault models us-
The negative slip deficit here has also been described ing the same subduction zone slip rate distributions
by other coupling models and it may result from fore- as Models A–C: Model D–interseismic coupling, Model
arc GNSS velocities in this region being significantly in- E–partitioned interseismic coupling, and Model F–
fluenced by complex interactions with the Mendocino coseismic rupture. These multi-fault models include
triple junction (e.g., Saux et al., 2022). This region is far the eastern Leech River fault, the Devils Mountain fault,
enough away (> 500 km) from the LRDM that it likely Utsalady Point fault (Johnson et al., 2004b), Strawberry
has little impact on our results. More details on the Point fault (Johnson et al., 2001), and the South Whidbey
block model used to calculate the slip deficit and the Island fault (Johnson et al., 1996; Sherrod et al., 2008).
Cascadia subduction zone mesh used in our model are The purpose of including the additional fault zones was
described in Delano et al. (2017) (Supplementary Mate- to test how multiple structures interact with each other.
rial). We chose to include these faults as their surface traces
We prescribed three different slip rate distributions all converge towards each other. We also chose to ex-
to the subduction zone mesh to predict slip rates on clude the western W-striking segment of the Leech River
the LRDM due to unpartitioned interseismic coupling fault in this model, as there is no reported evidence of
(Model A), interseismic strike-slip coupling (to model offset Upper Oligocene sediments that overlie this seg-
partitioned forearc strain, Model B), and coseismic ment of the fault on the west coast of Vancouver Is-
megathrust rupture (Model C). For Model A, we pre- land (MacLeod et al., 1977; Fairchild and Cowan, 1982;
scribe a negative (normal) slip-deficit rate and a cor- Groome et al., 2003), and we wanted to test how remov-
responding left- or right-lateral slip-deficit rate in the ing this portion of the fault would affect its kinemat-
opposite direction of relative plate motion (Fig. 4a, b). ics. We limited the down-dip projection of each fault
These slip deficit rates imposed on the upper plate to 20 km and set them all to dip 75◦ towards the north
mimic the dragging of the upper plate along the subduc- or northwest. We used these new parameters to ensure
tion interface during interseismic coupling. To model that the surfaces did not cross-cut one another.
partitioned forearc strain (Model B), we prescribe zero
dip-slip deficit rates to the subduction zone mesh, thus
3.3 Elastic half-space assumptions
only the strike-slip deficit rates remain as model inputs
that deform the elastic half-space. Finally, to model co- The BEM model employs the prescribed slip deficit rate
seismic megathrust rupture (Model C), we prescribe a distributions to analytically calculate the stress rate at a
positive (reverse) slip rate and a corresponding strike- given point in a linear, homogeneous elastic half-space,
slip rate parallel with relative plate motion (Fig. 4c, d), characterized by a Poisson’s Ratio of 0.25 and shear
mimicking the full elastic recovery of the upper plate modulus of 3x108 Pa. The BEM model then predicts
during, or immediately after, a rupture of the entire Cas- slip-rates on crustal faults by solving for the slip-rate on
cadia subduction zone. each crustal fault TDE required to completely relieve the
stress rate on a traction-free surface at that point in the
3.2 Forearc fault meshes elastic half-space. Displacement normal to any of the
TDEs is not allowed in the model to prevent fault open-
We constructed the meshed surface representing the ing. The model assumes a linear homogeneous crust,
LRDM by extrapolating surface traces of the faults at and that the crustal faults are traction-free surfaces that
a specified dip to a maximum depth. The surface relieve all of the stress imposed on them, including the
Figure 4 Prescribed slip rates (tractions) on a mesh representing the Cascadia subduction zone interface that drive elastic
deformation and forearc fault slip in the boundary element method (BEM) model. Interseismic dip-slip (a) and strike-slip
(b) deficit rates represent interseismic coupling. Coseismic dip-slip (c) and strike-slip (d) rates represent coseismic slip. The
mesh representing the Leech River-Devils Mountain forearc fault system (LRDM) is shown in black, and the Juan de Fuca (JDF),
North America (NA), and Pacific plate boundaries are shown in green. UTM grid coordinates (m) are in NAD83 UTM Zone 10N.
Input slip deficit and slip rates are from (Delano et al., 2017).
stresses induced by slip on neighboring TDEs on the of fluid pore pressure, or fault slip mechanics and earth-
same or neighboring faults. Therefore, where we model quake recurrence intervals. However, the simplicity
multiple crustal faults, the slip rates are lower where of the model allows us to conduct a first-order test of
the faults are in close proximity to each other as the whether stress due to elastic deformation of the fore-
strain is distributed. Additionally, slip rates are lower arc by megathrust coupling or rupture, results in the ob-
at the edges of the faults where there are no free sur- served fault kinematics along the LRDM.
faces. The model does not take into account viscous de-
formation below the brittle-ductile transition, changes

4 Model Results 0.05 mm/yr of normal slip while the rest of the faults
show < 0.2 mm/yr of reverse slip. Finally, in Model
Model A, where unpartitioned coupling on the Cascadia F, coseismic rupture on the megathrust drives the op-
subduction interface (including both the strike-slip and posite kinematics of the interseismic coupling model
dip-slip components) drives strain in the upper plate, (Model D), with < 0.7 mm/yr of normal slip and < 1.6
predicts reverse left-lateral slip along the entire LRDM mm/yr of right-lateral slip.
(Fig. 5a). Reverse slip rates range from ∼ 0 to 3 mm/yr
with the highest rates at the far western end of the Leech
River fault, and where the Leech River fault bends and
5 How do the BEM models compare
strikes WNW. This western increase in slip rate is par- with previous studies?
tially due to higher slip rates being produced where the
Our models of upper plate fault kinematics indicate that
fault is located closer to the subduction interface. Left-
the elastic response of the upper plate to subduction
lateral strike slip rates range from ∼ 0 to 3.5 mm/yr with
zone coupling, or coseismic rupture, cannot be the sole
a steady increase in slip rates towards the western end
explanation for the permanent deformation accommo-
of the fault system, again due to proximity to the sub-
dated by the modelled upper plate faults in northern
duction interface. Strike-slip rates decrease down-dip
Cascadia. Predominantly left-lateral reverse slip pre-
on the LRDM mesh away from the Earth’s surface.
dicted in the interseismic models (Models A, B, D, and
Model B, where the prescribed strike-slip deficit rates E) differs in strike-slip sense from the right-lateral fault
on the megathrust mimic partitioned forearc strain, kinematics inferred on these structures in paleoseis-
predicts that the LRDM accommodates left-lateral (red) mic studies (Fig. 2a, Table 1a), instrumental seismic-
slip, and both normal (blue) and reverse (red) compo- ity (Fig. 2b), and geodetic block modelling. The coseis-
nents of dip slip (Fig. 5b). Modelled dip-slip rates range mic models (Models C and F) predict normal slip, which
from ∼ 0.6 mm/yr of normal slip on the W-striking west- is not inferred for any of the structures in this study.
ern end of the Leech River fault to ∼ 0.15 mm/yr of re- In contrast to these inconsistencies, the single fault in-
verse slip along the W-striking Devils Mountain fault. terseismic coupling model (Model A) and instrumen-
Very little to no slip is calculated on the NW-striking tal seismicity suggest similar reverse left-lateral strain
middle segment of the fault system, except for one ∼ regimes off the west coast of Vancouver Island, and
5 km section in the middle of this segment where < 0.1 may indicate subduction zone coupling is an important
mm/yr of reverse slip is estimated. Modelled strike-slip driver of faulting in the outer forearc.
rates are also lower for the partitioned coupling model:
∼ 0 to 0.7 mm/yr of left-lateral strike slip with the rate
decreasing towards the east away from the subduction 5.1 Paleoseismic studies
zone. The overall reduction in slip-rates are expected The fault kinematics produced in the BEM models are
due to the lower absolute value of the slip deficit rates inconsistent with most of the observed active kinemat-
prescribed on the subduction interface. ics from paleoseismic studies of the LRDM, whether it
In Model C (Fig. 5c), coseismic rupture on the is modelled as a single structure (Models A-C) or mul-
megathrust results in calculated slip-rate distributions tiple faults (Models D-F). Paleoseismic investigations
for the LRDM that are the opposite of Model A. Instead predominantly observe reverse right-lateral slip aside
of reverse left-lateral slip, normal and right-lateral slip from a study on the Utsalady Point fault (Fig. 2a; Ta-
with slip rates ranging from ∼ 0 to 3 mm/yr increasing ble 1). These observations are in contrast with the sin-
towards the western end of the Leech River fault are pre- gle structure and multi-fault unpartitioned interseismic
dicted. coupling models predicting reverse left-lateral slip on
In Models D–F (Fig. 6), where we include multiple all faults (Models A and D). In the strain partitioning
faults (the eastern Leech River, Devils Mountain, South model (Model B), the sense of dip slip switches along
Whidbey Island, Strawberry Point, and Utsalady Point the western portion of the Leech River fault. However,
faults), we see similar slip-rate distributions compared as Model B still predicts left-lateral slip along the length
to the previously described models that only include the of the fault, it is still inconsistent with the paleoseis-
LRDM. In all of the multi-fault models, slip rates are mic observations. Finally, both the coseismic models
lower where the faults overlap each other (they are sub- (Models C and F) produce right-lateral slip on all fore-
parallel in space) and deformation is partitioned onto arc faults, similar to paleoseismic observations (Fig. 2a;
multiple surfaces. Also, the NW-striking South Whid- Table 1). However, they also predict normal slip on N-
bey Island fault has greater slip rates than the the more dipping faults, inconsistent with the observed N-side up
westerly-striking Utsalady Point, Strawberry Point, and motion along the LRDM.
Devils Mountain faults. In Model D, where unparti- Despite the inconsistencies between the BEM mod-
tioned interseismic coupling drives strain, all of the els and paleoseismic observations in slip sense, most of
modelled faults are predicted to accommodate < 0.7 the modelled absolute slip rates compare relatively well
mm/yr of reverse dip slip and < 1.6 mm/yr of left-lateral with most of the observed slip rates (Table 1). Model
slip. Model E, where only strike-slip coupling drives de- B, and Models D through F estimate absolute slip rates
formation, predicts < 0.45 mm/yr of left-lateral slip on (0 to 1.5 mm/yr) that are similar to the slip rates ob-
the forearc faults. The far-eastern down-dip portion of served in paleoseismic studies. Given this similarity,
the Leech River fault, and where the Leech River fault we speculate that the paleoseismic slip sense observa-
overlaps with the South Whidbey Island fault, show < tions could incorporate alternating slip histories result-
Figure 5 Dip-slip and strike-slip rates on the LRDM resulting from unpartitioned coupling in Model A (a), strike-slip coupling
in Model B (b), and coseismic slip in Model C (c) on the Cascadia subduction zone mesh (Fig. 4). Red denotes reverse slip-rates
and left-lateral slip rates, while blue denotes normal slip rates and right-lateral slip rates. Surface trace of the LRDM shown
with solid black line. Fault dips away from the surface trace to the north. UTM grid coordinates (m) are in NAD83 UTM Zone
10N.
ing from stress reversals during the megathrust cycle the fault system along strike, such as the Devil’s Moun-
(e.g., Hasegawa et al., 2012; Regalla et al., 2017; Cortés- tain Fault (Fig. 2a; Table 1). Although this possibil-
Aranda et al., 2022), and the kinematics observed today ity exists, the most unambiguous observation of strike-
are the net sum of these alternating slip sense events. slip displacement are right-laterally offset channels ob-
This hypothesis could explain why left-lateral slip sense served in 3D trenching along the Devils Mountain fault
is observed on the Utsulady Point fault in contrast to (Personius et al., 2014), and the oblique right-lateral slip
right-lateral slip sense or pure dip slip along the rest of is consistent with interseismic GNSS and seismicity ob-

Figure 6 Maps showing modelled dip-slip (left) and strike-slip (right) rates on on multiple faults in the northern Cascadia
forearc. Slip in the models results from unpartitioned coupling in Model D (a), only strike–slip coupling in Model E (b), and
unpartitioned coseismic slip in Model F (c) on the Cascadia subduction zone mesh (Fig. 4). Red denotes reverse slip-rates
and left-lateral slip rates, while blue denotes normal slip rates and right-lateral slip rates. Modelled faults are the eastern
Leech River (LRF), Devils Mountain (DM), Strawberry Point (SP), Utsalady Point (UP), and South Whidbey Island (SWIF) faults.
Surface trace of the crustal faults are shown with solid black lines. Faults dip away from the surface trace to the north. UTM
grid coordinates (m) are in NAD83 UTM Zone 10N.
servations (Fig. 2b; Fig. 3). This consistency between slip rates occur where the LRDM is thought to be inac-
trenching and interseismic observations is indicative tive (MacLeod et al., 1977; Fairchild and Cowan, 1982;
that permanent interseismic strain is right-lateral, and Groome et al., 2003). Thus this inconsistency may be
is in contrast with all of our interseismic BEM models further evidence that our simple model of elastic defor-
(Models A, B, D, E). Finally, Models A and C predict max- mation resulting from the megathrust cycle cannot be
imum slip rates of up to 3 to 3.5 mm/yr at the western used to explain the observed forearc faulting in north-
end of the LRDM, which have not been observed in pale- ern Cascadia.
oseismic investigations. In fact, these highest modelled
5.2 Crustal seismicity block boundaries (Meade and Loveless, 2009; Evans,
2022). Thus, these models represent an inversion of the
Similar to the paleoseismic observations, recorded
GNSS velocities, while our model ”drives” forearc de-
crustal seismicity, which has occurred during an inter-
formation using the stress that results from subduction
seismic period in the Cascadia subduction zone megath-
zone coupling. As the block models are inferred to be
rust cycle, is also inconsistent with the BEM models of
a prediction of the current deformation being observed
inner forearc (onshore) fault kinematics. Upper plate
at the block boundaries, they are a useful comparison
focal mechanisms (Brocher et al., 2017; Li et al., 2018a)
with our modelled fault kinematics.
and SHmax directions derived from crustal seismicity
Two of the block models that include the Devils Moun-
(Balfour et al., 2011) predict right-lateral reverse slip
tain fault as a block boundary estimate right-lateral slip
along the LRDM (Fig. 2b). This reverse right-lateral slip
along this structure (McCaffrey et al., 2007; Evans, 2022),
is inconsistent with the interseismic coupling models
which is inconsistent with our prediction of left lateral
(Models A, B, D, and E) that predict a left-lateral compo-
slip on the Devils Mountain fault, the entire LRDM, or
nent of strike slip along the modelled upper plate faults,
the network of forearc faults (Models A, B, D, and E).
as well as the coseismic BEM models (Models C and F)
The McCaffrey et al. (2007) block model indicates a nor-
that yield a normal component of dip slip on these struc-
mal, right-lateral slip-rate slip rate of ∼ 0.4 mm/yr on
tures.
the Devils Mountain fault, consistent with our model
In contrast to the inner forearc, the reverse left-
driven by coseismic rupture (Model C). However, the
lateral slip predicted by our unpartitioned interseis-
McCaffrey et al. (2007) block model reflects interseis-
mic coupling model (Model A) in the outer forearc off
mic GNSS velocities, and the error on the block model
the west coast of Vancouver Island is consistent with
is large enough that left-lateral reverse slip is also pos-
SHmax directions derived from upper plate seismicity
sible. Therefore we hesitate to use this comparison
in the same region. At the farthest west end of the
to make any deductions about coseismic elastic defor-
Leech River fault, off the west coast of Vancouver Is-
mation controlling permanent forearc deformation on
land, the SW–NE SHmax direction from crustal seismicity
this segment of our modelled fault system. Finally, the
would predict left-lateral reverse slip on a W-trending
Evans (2022) block model predicts ∼0.05 to 0.5 mm/yr
fault (Fig. 2b), consistent with slip produced by our un-
(standard deviation of up to 1 mm/yr) of reverse right-
partitioned interseismic coupling model (Model A). Al-
lateral slip along the Devils Mountain fault, which is in-
though the western on-land portion of the Leech River
consistent with all of our models of upper plate fault
fault has formerly been deemed inactive (MacLeod
kinematics.
et al., 1977; Fairchild and Cowan, 1982; Groome et al.,
2003), crustal seismicity increases offshore towards the
west (Fig. 2b), suggesting the stress rate also increases 6 What drives the observed oblique
offshore (Stevens and Avouac, 2021), and potentially right-lateral slip?
indicates offshore active faulting. There is no evi-
dence that this offshore faulting is along the Leech Our BEM models suggesting that elastic response of
River fault. However, our interseismic model shows the upper plate to interseismic subduction zone cou-
left-lateral oblique slip on a W-striking fault, which im- pling and coseismic rupture alone cannot explain the
plies the same SW–NE SHmax direction as calculated by observed fault kinematics of the LRDM may be repre-
Balfour et al. (2011). This SW–NE SHmax direction is sentative of a larger area in Northern Cascadia. Simi-
consistent with inferred kinematics on seismically im- lar BEM modelling by Duckworth et al. (2021) predict-
aged structures in the forearc basin sediments west ing slip on the North Olympic fault zone ∼ 50 km south
of Vancouver Island. The imaged structures are late of the Leech River fault (Fig. 1b), also concludes that
Pliocene NE-verging reverse faults and folds (Hayward stress arising from interseismic coupling does not drive
and Calvert, 2007) and may still be active. This area west the right-lateral oblique motion observed on this fore-
of Vancouver Island is immediately above where cou- arc fault zone. They suggest that coseismic strain may
pling on the megathrust increases towards the trench have resulted in reverse right-lateral slip; or, that the
(Schmalzle et al., 2014; Li et al., 2018b, Fig. 4a, b;). We North Olympic fault zone acts as the northern bound-
speculate that these observations indicate subduction ary of a rotating tectonic block (”Olym” block, McCaf-
zone coupling may be important in controlling perma- frey et al., 2013) and that strain along the boundary is
nent strain in the outer forearc closer to the trench, the result of a deformation field larger than the subduc-
which is proximal or immediately above the locked portion zone. Their coseismic BEM model, like ours, pre-
tion of the megathrust. dicts normal slip and contrasts with the reverse slip that
has been observed in paleoseismic studies (Schermer
et al., 2021) and bedrock mapping (Polenz et al., 2004).
5.3 GNSS derived strain and previous BEM
Therefore, we conclude that these observations suggest
modelling
coupling or coseismic rupture also do not drive fault slip
Geodetic block models of Cascadia use GNSS velocities on the Olympic Peninsula, and may indicate our results
to constrain the relative rotations of microplates, inter- are not just confined to the LRDM and may reflect most
seismic elastic deformation due to locked faults along of northern Cascadia.
microplate boundaries, and the distribution of coupling Since our models indicate that subduction zone cou-
along partially locked faults. Slip rates are calculated pling and coseismic rupture alone cannot explain the
using the difference between the rotation rates at the observed strike-slip fault kinematics in the inner fore-
arc of northern Cascadia, we propose that permanent lower slip rates, especially where faults are in close
deformation accommodated by the LRDM combined is proximity to each other. These results are not consis-
the combined effect of subduction zone processes and tent with paleoseismic studies, instrumental seismicity,
far-field tectonic processes. This interpretation is sup- and geodetic block models which predominantly indi-
ported by the McCaffrey et al. (2007) and Evans (2022) cate right-lateral, reverse fault kinematics on the mod-
block models, which predict right-lateral slip on the elled faults in northern Cascadia. This result allows
Devils Mountain fault and is in agreement with crustal us to deduce that observed permanent deformation in
seismicity and paleoseismic studies (Fig. 2). As these the inner forearc in northern Cascadia does not solely
block models are based on GNSS data covering the en- accommodate elastic strain driven by subduction cou-
tirety of the Pacific Northwest region (McCaffrey et al., pling or coseismic rupture. SHmax directions predicting
2007), they support our conclusion that including far- left-lateral slip along the furthest west, offshore portion
field deformation is needed to reproduce the observed of the modelled Leech River fault, are consistent with
fault kinematics in forearc strain models. Similar con- our interseismic coupling models. We speculate that
clusions from studies of other subduction zones have this result suggests some of the elastic strain from sub-
been made where permanent inner forearc deforma- duction zone coupling may be reflected by permanent
tion is thought to reflect the background stress state deformation in the outer forearc, closer to the coupled
of the forearc and is not directly related to subduction portion of the megathrust. In the inner forearc, we sug-
zone coupling (e.g., Townend and Zoback, 2006; Tow- gest that oroclinal bending, and west-northwestward
nend et al., 2012; Dimitrova et al., 2016). extrusion of the Olympic Peninsula and the southwest-
A far-field tectonic process that we suspect is a sub- ern tip of Vancouver Island induce right-lateral slip on
stantial factor in driving forearc strain in northern Cas- the LRDM.
cadia is the oroclinal bending of the Cascadia forearc,
which accommodates north-south shortening across
the Olympic Peninsula and may promote its westward
Acknowledgements
escape (Fig. 3; Nelson et al., 2017; Finley et al., 2019;
Funding to support this research was from a Natural
Harrichhausen et al., 2021). Right-lateral slip on the
Sciences and Engineering Research Council of Canada
LRDM, in the case of oroclinal bending, is a result of
(NSERC) Post Graduate Scholarship and Centre National
flexural slip along the northern limb of the orocline
d’Études Spatiale (CNES) postdoctoral fellowship to N.
(Fig. 3). Westward escape of the Olympic Peninsula rel-
Harrichhausen, National Science Foundation Earth Sci-
ative to the northern forearc would also induce right-
ences (NSF EAR) grants #1756943 and #2046278 to K.
lateral motion on roughly west- and east-striking upper
Morell, and #1756834 to C. Regalla. This manuscript
plate faults. Because residual GNSS velocities south of
benefited greatly from discussions on the tectonics of
the Devils Mountain fault do not indicate westward es-
northern Cascadia with J. Loveless, E. Nissen, L. Lu-
cape immediately south of the Devils Mountain fault
cinda, T. Finley, S. Bennett, E. Lynch, and E. Schotten-
(Fig. 3), we suggest that the primary driver of right-
fells. Finally we thank H. Tobin, E. Lindsay, and an
lateral slip, at least at the eastern end of the LRDM, is
anonymous reviewer for their thoughtful insight and
flexural slip, while westward escape may explain right-
comments on our manuscript.
lateral motion further west on the Leech River fault.
In addition to the previously described tectonic pro-
cesses, we have not considered other drivers of forearc Data and code availability
strain, such arc-parallel flow in the mantle wedge (e.g.,
Sternai et al., 2014; Chen et al., 2016) or buoyancy con- Codes and input files for the BEM models are available
trasts along the forearc (e.g., Rippke, 2020), that could in Harrichhausen et al. (2024). These codes are adapted
potentially result in right-lateral slip along the LRDM from Loveless (2021). GNSS velocity data in Fig. 1a
fault. Although it is beyond the scope of this study, con- is from https://www.unavco.org/software/visualization/
sidering these other sources of forearc strain in future GPS-Velocity-Viewer/GPS-Velocity-Viewer.html.
models is essential to determine what processes control
permanent deformation of forearcs at different subduc-
tion zones. Competing interests
7 Conclusions There are no competing interests pertaining to the au-

thors.
Our BEM models suggest that elastic forearc strain
from interseismic coupling would result in predomi- References
nantly left-lateral reverse slip along the LRDM, and that
elastic strain from coseismic megathrust rupture re- Allmendinger, R. W. and González, G. Invited review paper:
sults in right-lateral normal fault slip. In addition, our Neogene to Quaternary tectonics of the coastal Cordillera,
models using only the strike-slip component of inter- northern Chile. Tectonophysics, 495(1-2):93–110, 2010. doi:
seismic coupling to drive fault slip, mimicking fore- 10.1016/j.tecto.2009.04.019.
arc strain partitioning, also predict predominantly left- Audet, P. and Bürgmann, R. Possible control of subduction zone
lateral forearc faulting. Multi-fault models predict the slow-earthquake periodicity by silica enrichment. Nature, 510
same fault kinematics as single fault models, albeit with (7505):389–392, 2014. doi: 10.1038/nature13391.

Balfour, N. J., Cassidy, J. F., Dosso, S. E., and Mazzotti, S. Mapping 2010. doi: 10.1111/j.1365-246x.2009.04491.x.
crustal stress and strain in southwest British Columbia. Jour- Dimitrova, L., Wallace, L., Haines, A., and Williams, C. High-
nal of Geophysical Research: Solid Earth, 116(B3), 2011. doi: resolution view of active tectonic deformation along the Hiku-
10.1029/2010jb008003. rangi subduction margin and the Taupo Volcanic Zone, New
Barrie, J. V. and Greene, H. G. The Devils Mountain Fault zone: An Zealand. New Zealand Journal of Geology and Geophysics, 59
active Cascadia upper plate zone of deformation, Pacific North- (1):43–57, 2016. doi: 10.1080/00288306.2015.1127823.
west of North America. Sedimentary Geology, 364:228–241, Dragert, H., Hyndman, R., Rogers, G., and Wang, K. Current defor-
2018. doi: 10.1016/j.sedgeo.2017.12.018. mation and the width of the seismogenic zone of the northern
Bassett, D. and Watts, A. B. Gravity anomalies, crustal structure, Cascadia subduction thrust. Journal of Geophysical Research:
and seismicity at subduction zones: 1. Seafloor roughness and Solid Earth, 99(B1):653–668, 1994. doi: 10.1029/93jb02516.
subducting relief. Geochemistry, Geophysics, Geosystems, 16(5): Duckworth, W. C., Amos, C. B., Schermer, E. R., Loveless, J. P., and
1508–1540, 2015. doi: 10.1002/2014gc005684. Rittenour, T. M. Slip and strain accumulation along the Sadie
Bellier, O., Sebrier, M., Pramumijoyo, S., Beaudouin, T., Har- Creek fault, Olympic Peninsula, Washington. Journal of Geo-
jono, H., Bahar, I., and Forni, O. Paleoseismicity and seismic physical Research: Solid Earth, 126(3):2020JB020276, 2021. doi:
hazard along the Great Sumatran Fault (Indonesia). Journal 10.1029/2020jb020276.
of Geodynamics, 24(1-4):169–183, 1997. doi: 10.1016/s0264- Evans, E. L. A dense block model representing western continental
3707(96)00051-8. United States deformation for the 2023 update to the National
Blair, J. L., McCrory, P., Oppenheimer, D., and Waldhauser, F. A Seismic Hazard Model. Seismological Society of America, 93(6):
geo-referenced 3D model of the Juan de Fuca slab and associ- 3024–3036, 2022. doi: 10.1785/0220220141.
ated seismicity. US Department of the Interior, US Geological Fagereng, Å., Diener, J. F., Meneghini, F., Harris, C., and Kvadsheim,
Survey, 2011. doi: 10.3133/ds633. A. Quartz vein formation by local dehydration embrittlement
Blakely, R. J., Wells, R. E., Weaver, C. S., and Johnson, S. Y. along the deep, tremorgenic subduction thrust interface. Geol-
Location, structure, and seismicity of the Seattle fault zone, ogy, 46(1):67–70, 2018. doi: 10.1130/G39649.1.
Washington: Evidence from aeromagnetic anomalies, geologic Fairchild, L. H. and Cowan, D. S. Structure, petrology, and tec-
mapping, and seismic-reflection data. Geological Society of tonic history of the Leech River complex northwest of Victoria,
America Bulletin, 114(2):169–177, 2002. doi: 10.1130/0016- Vancouver Island. Canadian Journal of Earth Sciences, 19(9):
7606(2002)114<0169:lsasot>2.0.co;2. 1817–1835, 1982. doi: 10.1139/e82-161.
Bostock, M. G., Christensen, N. I., and Peacock, S. M. Seis- Feng, L., Newman, A. V., Protti, M., Gonzalez, V., Jiang, Y., and
micity in Cascadia. Lithos, 332:55–66, 2019. doi: Dixon, T. H. Active deformation near the Nicoya Peninsula,
10.1016/j.lithos.2019.02.019. northwestern Costa Rica, between 1996 and 2010: Interseismic
Brocher, T. M., Wells, R. E., Lamb, A. P., and Weaver, C. S. Evidence megathrust coupling. Journal of Geophysical Research: Solid
for distributed clockwise rotation of the crust in the northwest- Earth, 117(B6), 2012. doi: 10.1029/2012jb009230.
ern United States from fault geometries and focal mechanisms. Finley, T., Morell, K., Leonard, L., Regalla, C., Johnston, S. T., and
Tectonics, 36(5):787–818, 2017. doi: 10.1002/2016tc004223. Zhang, W. Ongoing oroclinal bending in the Cascadia forearc
Burgette, R. J., Weldon II, R. J., and Schmidt, D. A. Interseis- and its relation to concave-outboard plate margin geometry.
mic uplift rates for western Oregon and along-strike variation Geology, 47(2):155–158, 2019. doi: 10.1130/g45473.1.
in locking on the Cascadia subduction zone. Journal of Geo- Graham, A. Geometry, kinematics, and Quaternary activity of
physical Research: Solid Earth, 114(B1), 2009. doi: https://- the brittle Leech River fault zone, southern Vancouver Island,
doi.org/10.1029/2008JB005679. British Columbia, Canada. Master’s thesis, University of Victo-
Chen, Z., Schellart, W. P., Strak, V., and Duarte, J. C. Does ria, 2018.
subduction-induced mantle flow drive backarc extension? Greene, H. G., Barrie, J. V., and Todd, B. J. The Skipjack Island fault
Earth and Planetary Science Letters, 441:200–210, 2016. doi: zone: An active transcurrent structure within the upper plate of
10.1016/j.epsl.2016.02.027. the Cascadia subduction complex. Sedimentary Geology, 378:
Condit, C. B., Guevara, V. E., Delph, J. R., and French, M. E. Slab de- 61–79, 2018. doi: 10.1016/j.sedgeo.2018.05.005.
hydration in warm subduction zones at depths of episodic slip Groome, W. G., Thorkelson, D. J., Friedman, R. M., Mortensen, J. K.,
and tremor. Earth and Planetary Science Letters, 552:116601, Massey, N. W. D., Marshall, D. D., and Layer, P. W. Magmatic and
2020. doi: 10.1016/j.epsl.2020.116601. tectonic history of the Leech River Complex, Vancouver Island,
Cortés-Aranda, J., González, J., Molina, D., Astudillo-Sotomayor, British Columbia: Evidence for ridge-trench intersection and ac-
L., Tassara, A., Miller, M., Álvarez Amado, F., González, R., and Ba- cretion of the Crescent Terrane. Special Papers-Geological Soci-
hamondes, D. Linking Upper-Plate Fault Reactivation With the ety of America, pages 327–354, 2003. doi: 10.1130/0-8137-2371-
Megathrust Earthquake Cycle: The Case of the Northern Chile x.327.
Outer Forearc (19°S–23°S). Tectonics, 41(11):e2021TC006956, Harrichhausen, N., Morell, K. D., Regalla, C., Bennett, S. E.,
2022. doi: https://doi.org/10.1029/2021TC006956. Leonard, L. J., Lynch, E. M., and Nissen, E. Paleoseismic Trench-
Crouch, S. and Starfield, A. M. Boundary Element Methods ing Reveals Late Quaternary Kinematics of the Leech River Fault:
in Solid Mechanics: With Applications in Rock Mechanics and Implications for Forearc Strain Accumulation in Northern Cas-
Geological Engineering. Allen & Unwin, 1983. https:// cadia. Bulletin of the Seismological Society of America, 111(2):
books.google.com.tr/books?id=l-6uQgAACAAJ. 1110–1138, 2021. doi: 10.1785/0120200204.
Delano, J. E., Amos, C. B., Loveless, J. P., Rittenour, T. M., Sher- Harrichhausen, N., Finley, T., Morell, K. D., Regalla, C., Bennett, S.
rod, B. L., and Lynch, E. M. Influence of the megathrust earth- E. K., Leonard, L. J., Nissen, E., McLeod, E., Lynch, E. M., Sa-
quake cycle on upper-plate deformation in the Cascadia fore- lomon, G., and Sethanant, I. Discovery of an Active Forearc Fault
arc of Washington State, USA. Geology, 45(11):1051–1054, 2017. in an Urban Region: Holocene Rupture on the XEOLXELEK-Elk
doi: 10.1130/g39070.1. Lake Fault, Victoria, British Columbia, Canada. Tectonics, 42
DeMets, C., Gordon, R. G., and Argus, D. F. Geologically current (12), 2023. doi: https://doi.org/10.1029/2023TC008170.
plate motions. Geophysical Journal International, 181(1):1–80, Harrichhausen, N., Morell, K. D., and Regalla, C. Forearc faults

in northern Cascadia do not accommodate elastic strain driven and fault structure near the Leech River Fault Zone, southern
by the megathrust seismic cycle: techreport (1.2), 2024. doi: Vancouver Island. Journal of Geophysical Research: Solid Earth,
10.5281/zenodo.11104836. 123(4):2841–2855, 2018a. doi: 10.1002/2017jb015021.
Hasegawa, A., Yoshida, K., Asano, Y., Okada, T., Iinuma, T., and Ito, Li, S., Wang, K., Wang, Y., Jiang, Y., and Dosso, S. E. Geodetically in-
Y. Change in stress field after the 2011 great Tohoku-Oki earth- ferred locking state of the Cascadia megathrust based on a vis-
quake. Earth and Planetary Science Letters, 355-356:231–243, coelastic Earth model. Journal of Geophysical Research: Solid
2012. Earth, 123(9):8056–8072, 2018b. doi: 10.1029/2018jb015620.
Hayward, N. and Calvert, A. J. Seismic reflection and tomographic Littel, G. F., Bostock, M. G., Schaeffer, A., and Roecker, S.
velocity model constraints on the evolution of the Tofino forearc Microplate Evolution in the Queen Charlotte Triple Junc-
basin, British Columbia. Geophysical Journal International, 168 tion & Explorer Region: New Insights From Microseismic-
(2):634–646, 2007. doi: 10.1111/j.1365-246x.2006.03209.x. ity. Tectonics, 42(6):e2022TC007494, 2023. doi: https://-
Hyndman, R. D. and Wang, K. Thermal constraints on the zone of doi.org/10.1029/2022TC007494.
major thrust earthquake failure: The Cascadia subduction zone. Loveless, J. jploveless/tribem: tribem 1.2 (v1.2). Zenodo., 2021.
Journal of Geophysical Research: Solid Earth, 98(B2):2039–2060, doi: https://doi.org/10.5281/zenodo.5735649.
1993. doi: 10.1029/92jb02279. Loveless, J. P., Allmendinger, R. W., Pritchard, M. E., and González,
Hyndman, R. D. and Wang, K. The rupture zone of Cascadia G. Normal and reverse faulting driven by the subduction zone
great earthquakes from current deformation and the thermal earthquake cycle in the northern Chilean fore arc. Tectonics, 29
regime. Journal of Geophysical Research: Solid Earth, 100(B11): (2), 2010. doi: 10.1029/2009tc002465.
22133–22154, 1995. doi: 10.1029/95jb01970. MacLeod, N. S., Tiffin, D. L., Snavely Jr, P. D., and Currie, R. G. Ge-
Hyndman, R. D., Yamano, M., and Oleskevich, D. A. The seismo- ologic interpretation of magnetic and gravity anomalies in the
genic zone of subduction thrust faults. Island Arc, 6(3):244–260, Strait of Juan de Fuca, US–Canada. Canadian Journal of Earth
1997. doi: 10.1111/j.1440-1738.1997.tb00175.x. Sciences, 14(2):223–238, 1977. doi: 10.1139/e77-024.
Johnson, S. Y., Potter, C. J., and Armentrout, J. M. Origin Malatesta, L. C., Bruhat, L., Finnegan, N. J., and Olive, J.-A. L. Co-
and evolution of the Seattle fault and Seattle basin, Wash- location of the downdip end of seismic coupling and the conti-
ington. Geology, 22(1):71–74, 1994. doi: 10.1130/0091- nental shelf break. Journal of Geophysical Research: Solid Earth,
7613(1994)022<0071:oaeots>2.3.co;2. page e2020JB019589, 2018. doi: 10.31223/osf.io/uwzbr.
Johnson, S. Y., Potter, C. J., Miller, J. J., Armentrout, J. M., Finn, C., Marshall, J. S., Fisher, D. M., and Gardner, T. W. Central Costa
and Weaver, C. S. The southern Whidbey Island fault: an active Rica deformed belt: Kinematics of diffuse faulting across the
structure in the Puget Lowland, Washington. Geological Society western Panama block. Tectonics, 19(3):468–492, 2000. doi:
of America Bulletin, 108(3):334–354, 1996. doi: 10.1130/0016- 10.1029/1999tc001136.
7606(1996)108<0334:tswifa>2.3.co;2. Massey, N. W. D., MacIntyre, D. G., Desjardins, P. J., and Cooney,
Johnson, S. Y., Dadisman, S. V., Childs, J. R., and Stanley, W. D. R. T. Digital geology map of British Columbia. BC Ministry of
Active tectonics of the Seattle fault and central Puget Sound, Energy and Mines, Geofile, 7:2005, 2005.
Washington Implications for earthquake hazards. Geologi- Mazzotti, S., Dragert, H., Hyndman, R. D., Miller, M. M., and Henton,
cal Society of America Bulletin, 111(7):1042–1053, 1999. doi: J. A. GPS deformation in a region of high crustal seismicity: N.
10.1130/0016-7606(1999)111<1042:atotsf>2.3.co;2. Cascadia forearc. Earth and Planetary Science Letters, 198(1):
Johnson, S. Y., Dadisman, S. V., Mosher, D. C., Blakely, R. J., and 41–48, 2002. doi: 10.1016/s0012-821x(02)00520-4.
Childs, J. R. Active tectonics of the Devils Mountain fault and Mazzotti, S., Dragert, H., Henton, J., Schmidt, M., Hyndman, R.,
related structures, northern Puget Lowland and eastern Strait of James, T., Lu, Y., and Craymer, M. Current tectonics of north-
Juan de Fuca region, Pacific Northwest. U.S. Geological Survey ern Cascadia from a decade of GPS measurements. Journal
Professional Paper, (1643), 2001. doi: 10.3133/pp1643. of Geophysical Research: Solid Earth, 108(B12), 2003. doi:
Johnson, S. Y., Blakely, R. J., Stephenson, W. J., Dadisman, S. V., 10.1029/2003jb002653.
and Fisher, M. A. Active shortening of the Cascadia forearc and Mazzotti, S., Leonard, L. J., Cassidy, J. F., Rogers, G. C., and
implications for seismic hazards of the Puget Lowland. Tecton- Halchuk, S. Seismic hazard in western Canada from GPS strain
ics, 23(1):27 pages, 2004a. doi: 10.1029/2003tc001507. rates versus earthquake catalog. Journal of Geophysical Re-
Johnson, S. Y., Nelson, A. R., Personius, S. F., Wells, R. E., Kelsey, search: Solid Earth, 116(B12), 2011. doi: 10.1029/2011jb008213.
H. M., Sherrod, B. L., Okumura, K., Koehler, R., Witter, R. C., McCaffrey, R. Slip vectors and stretching of the Sumatran fore
Bradley, L.-A., et al. Evidence for late Holocene earthquakes arc. Geology, 19(9):881–884, 1991. doi: 10.1130/0091-
on the Utsalady Point fault, northern Puget lowland, Washing- 7613(1991)019<0881:svasot>2.3.co;2.
ton. Bulletin of the Seismological Society of America, 94(6):
McCaffrey, R. Oblique plate convergence, slip vectors, and forearc
2299–2316, 2004b. doi: 10.1785/0120040050.
deformation. Journal of Geophysical Research: Solid Earth, 97
Khazaradze, G., Qamar, A., and Dragert, H. Tectonic deformation (B6):8905–8915, 1992. doi: 10.1029/92jb00483.
in western Washington from continuous GPS measurements.
McCaffrey, R. Time-dependent inversion of three-component
Geophysical Research Letters, 26(20):3153–3156, 1999. doi:
continuous GPS for steady and transient sources in northern
10.1029/1999gl010458.
Cascadia. Geophysical Research Letters, 36(7), 2009. doi:
Kimura, G. Oblique subduction and collision: Forearc tectonics of 10.1029/2008gl036784.
the Kuril arc. Geology, 14(5):404–407, 1986. doi: 10.1130/0091-
McCaffrey, R., Long, M. D., Goldfinger, C., Zwick, P. C., Nabelek,
7613(1986)14<404:osacft>2.0.co;2.
J. L., Johnson, C. K., and Smith, C. Rotation and plate lock-
Leonard, L. J., Hyndman, R. D., and Mazzotti, S. Coseismic subsi- ing at the southern Cascadia subduction zone. Geophysi-
dence in the 1700 great Cascadia earthquake: Coastal estimates cal Research Letters, 27(19):3117–3120, October 2000. doi:
versus elastic dislocation models. Geological Society of America 10.1029/2000gl011768.
Bulletin, 116(5-6):655–670, 2004. doi: 10.1130/b25369.1.
McCaffrey, R., Qamar, A. I., King, R. W., Wells, R., Khazaradze, G.,
Li, G., Liu, Y., Regalla, C., and Morell, K. D. Seismicity relocation Williams, C. A., Stevens, C. W., Vollick, J. J., and Zwick, P. C.

Fault locking, block rotation and crustal deformation in the structural models of Cascadia, south Alaska, SW Japan, and
Pacific Northwest. Geophysical Journal International, 169(3): Chile. Journal of Geophysical Research: Solid Earth, 104(B7):
1315–1340, 2007. doi: 10.1111/j.1365-246x.2007.03371.x. 14965–14991, 1999. doi: 10.1029/1999jb900060.
McCaffrey, R., King, R. W., Payne, S. J., and Lancaster, M. Active Pacific Northwest Seismic Network. Seismicity Data from January
tectonics of northwestern US inferred from GPS-derived surface 1, 1970 to April 29, 2015 [techreport], 2022. https://pnsn.org/.
velocities. Journal of Geophysical Research: Solid Earth, 118(2): Personius, S. F., Briggs, R. W., Nelson, A. R., Schermer, E. R., Ma-
709–723, 2013. doi: 10.1029/2012jb009473. harrey, J. Z., Sherrod, B. L., Spaulding, S. A., and Bradley, L.-
McCrory, P. A., Blair, J. L., Oppenheimer, D. H., and Walter, S. R. A. Holocene earthquakes and right-lateral slip on the left-
Depth to the Juan de Fuca slab beneath the Cascadia subduc- lateral Darrington–Devils Mountain fault zone, northern Puget
tion margin; a 3-D model for sorting earthquakes. U.S. Geologi- Sound, Washington. Geosphere, 10(6):1482–1500, 2014. doi:
cal Survey. Data Series 91, 2006. https://pubs.usgs.gov/ds/91/. 10.1130/ges01067.1.
McKenzie, K. and Furlong, K. Isolating non-subduction-driven tec- Pezzopane, S. K. and Weldon, R. J. Tectonic role of active fault-
tonic processes in Cascadia. Geoscience Letters, 8(1):1–12, 2021. ing in central Oregon. Tectonics, 12(5):1140–1169, 1993. doi:
doi: 10.1186/s40562-021-00181-z. 10.1029/92tc02950.
Meade, B. J. Algorithms for the calculation of exact displace- Polenz, M., Wegmann, K. W., and Schasse, H. W. Geologic map
ments, strains, and stresses for triangular dislocation elements of the Elwha and Angeles Point 7.5-minute quadrangles, Clallam
in a uniform elastic half space. Computers & geosciences, 33(8): County, Washington. Washington State Division of Geology and
1064–1075, 2007. doi: 10.1016/j.cageo.2006.12.003. Earth Resources, 2004.
Meade, B. J. and Loveless, J. P. Block modeling with connected Quigley, M., Van Dissen, R., Litchfield, N., Villamor, P., Duffy, B.,
fault-network geometries and a linear elastic coupling estima- Barrell, D., Furlong, K., Stahl, T., Bilderback, E., and Noble,
tor in spherical coordinates. Bulletin of the Seismological Society D. Surface rupture during the 2010 Mw 7.1 Darfield (Canter-
of America, 99(6):3124–3139, 2009. doi: 10.1785/0120090088. bury) earthquake: Implications for fault rupture dynamics and
Miller, M. M., Johnson, D. J., Rubin, C. M., Dragert, H., Wang, K., seismic-hazard analysis. Geology, 40(1):55–58, 2012. doi:
Qamar, A., and Goldfinger, C. GPS-determination of along- 10.1130/G32528.1.
strike variation in Cascadia margin kinematics: Implications Regalla, C., Fisher, D. M., Kirby, E., Oakley, D., and Taylor, S. Slip
for relative plate motion, subduction zone coupling, and per- Inversion Along Inner Fore-Arc Faults, Eastern Tohoku, Japan.
manent deformation. Tectonics, 20(2):161–176, 2001. doi: Tectonics, 36(11):2647–2668, 2017. doi: 10.1002/2017tc004766.
10.1029/2000tc001224. Rippke, J. Forearc Stresses in the Northern Cascadia Subduction
Moore, J. C. and Saffer, D. Updip limit of the seismogenic zone Zone. Master’s thesis, University of Minnesota, 2020.
beneath the accretionary prism of southwest Japan: An ef- Ryder, I., Rietbrock, A., Kelson, K., Bürgmann, R., Floyd, M., Soc-
fect of diagenetic to low-grade metamorphic processes and in- quet, A., Vigny, C., and Carrizo, D. Large extensional after-
creasing effective stress. Geology, 29(2):183–186, 2001. doi: shocks in the continental forearc triggered by the 2010 Maule
10.1130/0091-7613(2001)029<0183:ulotsz>2.0.co;2. earthquake, Chile. Geophysical Journal International, 188(3):
Morell, K. D., Gardner, T. W., Fisher, D. M., Idleman, B. D., and 879–890, 2012. doi: 10.1111/j.1365-246x.2011.05321.x.
Zellner, H. M. Active thrusting, landscape evolution, and late Saux, J. P., Molitors Bergman, E. G., Evans, E. L., and Loveless,
Pleistocene sector collapse of Barú Volcano above the Cocos- J. P. The Role of Slow Slip Events in the Cascadia Sub-
Nazca slab tear, southern Central America. Bulletin, 125(7-8): duction Zone Earthquake Cycle. Journal of Geophysical Re-
1301–1318, 2013. doi: 10.1130/b30771.1. search: Solid Earth, 127(2):e2021JB022425, 2022. doi: https://-
Morell, K. D., Regalla, C., Leonard, L. J., Amos, C., and Levson, doi.org/10.1029/2021JB022425.
V. Quaternary Rupture of a Crustal Fault beneath Victoria, Savard, G., Bostock, M. G., and Christensen, N. I. Seismicity,
British Columbia, Canada. GSA Today, 27(3–4), 2017. doi: metamorphism, and fluid evolution across the Northern Cas-
10.1130/gsatg291a.1. cadia fore arc. Geochemistry, Geophysics, Geosystems, 19(6):
Morell, K. D., Regalla, C., Amos, C., Bennett, S., Leonard, L., Gra- 1881–1897, 2018. doi: 10.1029/2017gc007417.
ham, A., Reedy, T., Levson, V., and Telka, A. Holocene surface Savard, G., Bostock, M., Hutchinson, J., Kao, H., Christensen, N.,
rupture history of an active forearc fault redefines seismic haz- and Peacock, S. The northern terminus of Cascadia subduc-
ard in southwestern British Columbia, Canada. Geophysical Re- tion. Journal of Geophysical Research: Solid Earth, 125(6):
search Letters, 45(21):11–605, 2018. doi: 10.1029/2018gl078711. e2019JB018453, 2020. doi: 10.1029/2019jb018453.
Mouslopoulou, V., Nicol, A., Begg, J., Oncken, O., and Moreno, M. Schermer, E. R., Amos, C. B., Duckworth, W. C., Nelson, A. R., Ang-
Clusters of megaearthquakes on upper plate faults control the ster, S., Delano, J., and Sherrod, B. L. Postglacial Mw 7.0–7.5
Eastern Mediterranean hazard. Geophysical Research Letters, 42 earthquakes on the North Olympic Fault Zone, Washington. Bul-
(23):10–282, 2015. doi: 10.1002/2015gl066371. letin of the Seismological Society of America, 111(1):490–513,
Nelson, A. R., Johnson, S. Y., Kelsey, H. M., Wells, R. E., Sherrod, 2021. doi: 10.1785/0120200176.
B. L., Pezzopane, S. K., Bradley, L.-A., Koehler, R. D., and Buck- Schmalzle, G. M., McCaffrey, R., and Creager, K. C. Central Cascadia
nam, R. C. Late Holocene earthquakes on the Toe Jam Hill subduction zone creep. Geochemistry, Geophysics, Geosystems,
fault, Seattle fault zone, Bainbridge Island, Washington. Geo- 15(4):1515–1532, 2014. doi: 10.1002/2013gc005172.
logical Society of America Bulletin, 115(11):1388–1403, 2003. doi:
Sherrod, B. L., Blakely, R. J., Weaver, C. S., Kelsey, H. M., Barnett, E.,
10.1130/b25262.1.
Liberty, L., Meagher, K. L., and Pape, K. Finding concealed active
Nelson, A. R., Personius, S. F., Wells, R. E., Schermer, E. R., Bradley, faults: Extending the southern Whidbey Island fault across the
L.-A., Buck, J., and Reitman, N. Holocene earthquakes of Mag- Puget Lowland, Washington. Journal of Geophysical Research:
nitude 7 during westward escape of the Olympic Mountains, Solid Earth, 113(B5), 2008. doi: 10.1029/2007jb005060.
Washington. Bulletin of the Seismological Society of America, 107
Sieh, K. and Natawidjaja, D. Neotectonics of the Sumatran fault, In-
(5):2394–2415, 2017. doi: 10.1785/0120160323.
donesia. Journal of Geophysical Research: Solid Earth, 105(B12):
Oleskevich, D. A., Hyndman, R. D., and Wang, K. The updip and 28295–28326, 2000. doi: 10.1029/2000jb900120.
downdip limits to great subduction earthquakes: Thermal and

Sitchler, J. C., Fisher, D. M., Gardner, T. W., and Protti, M. Con- Cascadia subduction tremor muted by crustal faults. Geology,
straints on inner forearc deformation from balanced cross sec- 45(6):515–518, 2017. doi: 10.1130/g38835.1.
tions, Fila Costeña thrust belt, Costa Rica. Tectonics, 26(6), 2007. Yoshioka, S., Wang, K., and Mazzotti, S. Interseismic lock-
doi: 10.1029/2006tc001949. ing of the plate interface in the northern Cascadia subduc-
Sternai, P., Jolivet, L., Menant, A., and Gerya, T. Driving the up- tion zone, inferred from inversion of GPS data. Earth
per plate surface deformation by slab rollback and mantle flow. and Planetary Science Letters, 231(3-4):239–247, 2005. doi:
Earth and Planetary Science Letters, 405:110–118, 2014. doi: 10.1016/j.epsl.2004.12.018.
10.1016/j.epsl.2014.08.023.
Stevens, V. and Avouac, J. On the relationship between strain rate The article Inner forearc faults in northern Cascadia do not
and seismicity in the India-Asia collision zone: Implications for accommodate elastic strain driven by the megathrust seis-
probabilistic seismic hazard. Geophysical Journal International, mic cycle © 2024 by Nicolas Harrichhausen is licensed under
2021. doi: 10.1093/gji/ggab098. CC BY 4.0.
Thomas, A. L. Poly 3D: a three-dimensional, polygonal element,
displacement discontinuity boundary element computer pro-
gram with applications to fractures, faults, and cavities in the
Earth’s crust. Master’s thesis, Stanford University, 1993.
Townend, J. and Zoback, M. D. Stress, strain, and mountain build-
ing in central Japan. Journal of Geophysical Research: Solid
Earth, 111(B3), 2006. doi: 10.26686/wgtn.13876118.
Townend, J., Sherburn, S., Arnold, R., Boese, C., and Woods,
L. Three-dimensional variations in present-day tectonic stress
along the Australia–Pacific plate boundary in New Zealand.
Earth and Planetary Science Letters, 353:47–59, 2012. doi:
10.1016/j.epsl.2012.08.003.
Van Rijsingen, E., Funiciello, F., Corbi, F., and Lallemand, S.
Rough subducting seafloor reduces interseismic coupling and
mega-earthquake occurrence: Insights from analogue mod-
els. Geophysical Research Letters, 46(6):3124–3132, 2019. doi:
10.1029/2018gl081272.
Vrolijk, P. On the mechanical role of smectite in subduction
zones. Geology, 18(8):703–707, 1990. doi: 10.1130/0091-
7613(1990)018<0703:otmros>2.3.co;2.
Wallace, L. M., Beavan, J., McCaffrey, R., and Darby, D. Subduction
zone coupling and tectonic block rotations in the North Island,
New Zealand. Journal of Geophysical Research: Solid Earth, 109
(B12), 2004. doi: 10.1029/2004jb003241.
Wang, K. Stress–strain ‘paradox’, plate coupling, and fore-
arc seismicity at the Cascadia and Nankai subduction zones.
Tectonophysics, 319(4):321–338, 2000. doi: 10.1016/s0040-
1951(99)00301-7.
Wang, K. and Bilek, S. L. Invited review paper: Fault creep caused
by subduction of rough seafloor relief. Tectonophysics, 610:
1–24, 2014. doi: 10.1016/j.tecto.2013.11.024.
Wang, K. and Tréhu, A. M. Invited review paper: Some outstanding
issues in the study of great megathrust earthquakes—The Cas-
cadia example. Journal of Geodynamics, 98:1–18, 2016. doi:
10.1016/j.jog.2016.03.010.
Wang, K., Wells, R., Mazzotti, S., Hyndman, R. D., and Sagiya, T. A re-
vised dislocation model of interseismic deformation of the Cas-
cadia subduction zone. Journal of Geophysical Research: Solid
Earth, 108(B1), 2003. doi: 10.1029/2001jb001227.
Wang, K., Hu, Y., Bevis, M., Kendrick, E., Smalley, R., Vargas, R. B.,
and Lauría, E. Crustal motion in the zone of the 1960 Chile earth-
quake: Detangling earthquake-cycle deformation and forearc-
sliver translation. Geochemistry, Geophysics, Geosystems, 8(10),
2007. doi: 10.1029/2007gc001721.
Wells, R. E. and McCaffrey, R. Steady rotation of the Cascade arc.
Geology, 41(9):1027–1030, 2013. doi: 10.1130/g34514.1.
Wells, R. E., Weaver, C. S., and Blakely, R. J. Fore-arc mi-
gration in Cascadia and its neotectonic significance.
Geology, 26(8):759–762, 1998. doi: 10.1130/0091-
7613(1998)026<0759:famica>2.3.co;2.
Wells, R. E., Blakely, R. J., Wech, A. G., McCrory, P. A., and Michael, A.

Harrichhausen 2024

Uploaded by

Copyright:

Available Formats

Harrichhausen 2024

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Harrichhausen 2024

Uploaded by

Copyright:

Available Formats

doi:10.26443/seismica.v2i4.

Inner forearc faults in northern Cascadia do not

1 Introduction is thought to be driven by interseismic coupling on the

1 SEISMICA | ISSN 2816-9387 | volume 2.4 | 2024

Our models test if the stress induced by either inter-

4 SEISMICA | volume 2.4 | 2024

which, along with oroclinal bending, may accommo-

8 SEISMICA | volume 2.4 | 2024

10 SEISMICA | volume 2.4 | 2024

7 Conclusions There are no competing interests pertaining to the au-

13 SEISMICA | volume 2.4 | 2024

14 SEISMICA | volume 2.4 | 2024

15 SEISMICA | volume 2.4 | 2024

16 SEISMICA | volume 2.4 | 2024

17 SEISMICA | volume 2.4 | 2024

You might also like