Original Paper
Landslides (2012) 9:75–91
DOI 10.1007/s10346-011-0286-4
Received: 4 March 2010
Accepted: 12 July 2011
Published online: 28 July 2011
© Springer-Verlag 2011
Marc-André Brideau I Matthieu Sturzenegger I Doug Stead I Michel Jaboyedoff I Martin
Lawrence I Nicholas Roberts I Brent Ward I Thomas Millard I John Clague
Stability analysis of the 2007 Chehalis lake landslide
based on long-range terrestrial photogrammetry
and airborne LiDAR data
Abstract On December 4th 2007, a 3-Mm3 landslide occurred
along the northwestern shore of Chehalis Lake. The initiation
zone is located at the intersection of the main valley slope and the
northern sidewall of a prominent gully. The slope failure caused a
displacement wave that ran up to 38m on the opposite shore of
the lake. The landslide is temporally associated with a rain-onsnow meteorological event which is thought to have triggered it.
This paper describes the Chehalis Lake landslide and presents a
comparison of discontinuity orientation datasets obtained using
three techniques: field measurements, terrestrial photogrammetric
3D models and an airborne LiDAR digital elevation model to
describe the orientation and characteristics of the five discontinuity sets present. The discontinuity orientation data are used to
perform kinematic, surface wedge limit equilibrium and threedimensional distinct element analyses. The kinematic and surface
wedge analyses suggest that the location of the slope failure
(intersection of the valley slope and a gully wall) has facilitated
the development of the unstable rock mass which initiated as a
planar sliding failure. Results from the three-dimensional distinct
element analyses suggest that the presence, orientation and high
persistence of a discontinuity set dipping obliquely to the slope
were critical to the development of the landslide and led to a
failure mechanism dominated by planar sliding. The threedimensional distinct element modelling also suggests that the
presence of a steeply dipping discontinuity set striking perpendicular to the slope and associated with a fault exerted a
significant control on the volume and extent of the failed rock
mass but not on the overall stability of the slope.
falling as snow was followed by warmer temperature and rainfall
on December 3 and 4, 2007. Rain-on-snow events increase the
“effective intensity” of a precipitation occurrence by combining
the current rainfall with the water temporarily stored as snow on
the slope. Runoff and infiltration are thought to have increased
pore water pressures in the fractured rock mass and reduced the
effective friction angle along discontinuity surfaces. The failed
mass, with a volume in the order of 3 Mm3, entered Chehalis Lake,
resulting in a displacement wave that destroyed three local
campgrounds and removed the vegetation on the opposite shore
of the lake to a maximum height of 38 m (Fig. 6). A preliminary
assessment of the landslide conducted by BGC Engineering Inc.
(2008), at the request of the BC Provincial Emergency Program,
highlighted that the failure surface was controlled by pervasive
planar structures which in places consisted of a fault with a fewcentimetres-thick gouge. Fresh tension cracks were observed up
to 21 m behind the headscarp (BGC Engineering Inc. 2008). The
damage caused by the landslide-generated displacement wave
associated with the Chehalis Lake landslide has previously been
briefly described by Geertsema et al. (2009) and Stephenson and
Rabinovich (2009). The main objective of the research presented
in this paper was to characterise the geology, rock mass quality,
tectonic structures and the discontinuity sets present at the
Chehalis Lake landslide. This information was then used as input
in slope stability analyses to evaluate potential failure mechanisms for the event. A secondary objective was to integrate the
information of discontinuities in the initiation zone based on
several data collection techniques and to discuss the applicability
and limitations of each technique.
Keywords Terrestrial photogrammetry . LiDAR . Distinct
element . Limit equilibrium
Introduction
The Chehalis Lake landslide occurred on December 4, 2007 on the
northwest shore of Chehalis Lake in southwest British Columbia
(Fig. 1). The rock slope failure and debris avalanche occurred on
the southeast-facing slope of Mount Orrock in the Cohoe Creek
drainage (Fig. 2). The headscarp is located at an elevation of
∼840 m asl (above sea level) while the lake shoreline is at ∼240 m
asl. The headscarp is 210 m in width and the landslide travelled a
horizontal distance of 950 m before entering the lake. The
locations and cross-sections pre- and post-landslide are presented
in Figs. 3 and 4, respectively. The slope failure was associated with
an intense rain-on-snow event as illustrated by the meteorological
data collected at the Agassiz weather station located approximately 30 km southeast of Chehalis Lake (Fig. 5). A cold period
between November 30 and December 2, 2007 with precipitation
Relevant previous landslide studies
Meteorological conditions are regarded as one of the most
common trigger for landslides (Sidle and Ochiai 2004; Highland
and Bobrowsky 2008). Rain-on-snow event is a particular
meteorological condition that has been discussed by Jakob and
Weatherly (2003) as a contributing or triggering factor for debris
flows in southwestern British Columbia. Similar patterns of
pronounced cold and warm temperatures cycles coupled with
precipitation prior to rock slope failure initiation have been noted
in British Columbia (e.g. rockfall along transportation corridors:
Peckover and Kerr 1977; Hope Slide: Mathews and McTaggart
1978, East Gate Landslide: Brideau et al. 2006; Clanwilliam
Landslide: Brideau et al. 2008) and elsewhere around the world
(e.g. Bjerrum and Jorstad 1968; Dunlop and Hutchinson 2009).
There have been two other documented instances of
landslides entering lakes and generating displacement waves
Landslides 9 & (2012)
75
Original Paper
Fig. 1 Location and regional
geology map of the Chehalis Lake
landslide (geology from Massey et al.
2005; faults from Thompson 1975)
British
Columbia
N
Landslide
Study
Area
Chehalis
Lake
Legend
Road
Campground
River
Lake
Geology
Quartz diorite
Andesitic volcanics
Fault
in British Columbia. In 1946, a rock slope failure on Mount
Colonel Foster on Vancouver Island was triggered by an
earthquake and entered the lake at its base (Evans 1989). The
maximum run-up height of the displacement wave was
measured at 51 m and was still evident in the mid-1980s. The
second event occurred in 1998 at Troitsa Lake, in central
British Columbia, where the surface material associated with a
large fan delta failed into the lake, generating a 20-m-high
Fig. 2 Overview of the Chehalis Lake
landslide showing geomorphic
features referred to in slope stability
analyses. Slope angle along runout
segment 1 was between 36° and 40°
and between 30° and 32° along
segment 2
N49.40o
1km
W122.00o
displacement wave that unsettled the lake for a period of 4 h
(Schwab 1999). Landslides entering fiords and creating displacement waves have also been investigated in British
Columbia (e.g. Bornhold and Harper 1998; van Zeyl 2009).
Similar landslide-generated waves in lakes or reservoirs have
been described elsewhere in the world (e.g. Plafker and
Eyzaguirre 1978; Wang et al. 2004; Genevois and Ghirotti
2005; Grimstad 2005; Panizzo, et al. 2005).
Cohoe
Creek
840m.a.s.l.
North
gully
wall
South
gully
wall
Rock
slide
Valley
slope
1
Debris
avalanche
440m.a.s.l.
2
Figure 7a
240m.a.s.l.
250m
76
Landslides 9 & (2012)
N
Fig. 3 Overview map with
orthophoto showing the location of
cross-sections, field stations and area
covered by the terrestrial
photogrammetry and COLTOP-3D
analyses
N
A
o
cr
cr
o
ss
-s
ec
tio
n2
B
tio
ec
-s
ss
B’
n1
A’
W122.00o
Legend
Field measurements of
discontinuities
Location from which terrestrial
photogrammetry acquired
N49.46o
Field geological mapping
Area from which discontinuity data
extracted using phogrammetry
Area from which discontinuity data
extracted from airborne-LiDAR
using COLTOP-3D
Road
Topographic contour
(10m interval)
300
150
0
Geology and rock mass description
Chehalis Lake is located along the contact of late Jurassic island
arc rocks of the Harrison Lake Formation and middle Jurassic to
early Cretaceous rocks of the Coast Plutonic Complex (Mahoney
et al. 1995). The area east of Chehalis Lake was mapped by
Monger (1970) and Arthur (1986) while the areas north and south
of the Chehalis Lake landslide were mapped by Thompson (1975)
and Mahoney et al. (1995), respectively. Monger (1970) and Arthur
(1986) observed that most of the Harrison Lake Formation
consists of intermediate to felsic volcanic flows and pyroclastic
rocks, while Thompson (1975) did not observe pyroclastic rocks in
the southern portion of Chehalis Lake. Arthur (1986) noted that
faults were difficult to distinguish in the field (due to the lack of
bedding structures and the degree of hydrothermal alteration) but
that several lineaments trending northwest–southeast were visible
300 Meters
on aerial photographs. Thompson (1975) and Mahoney et al.
(1995) included northwest–southeast trending faults in their
geological maps. Thompson (1975) also noted the presence of a
north-northeast (010o) trending fault (Fig. 1).
Fieldwork conducted as part of this project found that the
failure is located near an irregular intrusive contact between
andesite volcanics and quartz diorite units (Fig. 1). The failed
material is predominantly composed of quartz diorite. Xenoliths of volcanic material varying in size from a few
centimetres to several metres occur in the quartz diorite on
the northern edge of the landslide (Fig. 7a). Rock mass
descriptions and discontinuity orientation measurements were
obtained at six stations (Fig. 3). The rock mass quality was
assessed using the Geological Strength Index (GSI) chart
presented in Marinos and Hoek (2000) which is based on field
Landslides 9 & (2012)
77
Original Paper
a 1000
Cross-section 1
Legend
Pre-failure topography
(10m contour TRIM)
Post-failure topography
(1m LiDAR DEM)
Elevation (m.a.s.l.)
800
600
400
A’
200
0
400
600
800
1000m
Fig. 6 Damage to trees caused by the displacement wave generated by the
landslide entering Chehalis Lake. Photograph taken at the location of the former
campground at the northern end of the lake (view toward north and landslide 1
km to the left of the photograph)
Cross-section 2
760
720
680
640
B’
600
560
0
100
200
300
400
500m
Fig. 4 Longitudinal and transverse cross-sections of the pre- and post-failure
topography (see Fig. 3 for the location of cross-sections). Note that the pre-failure
topography has a lower resolution than the post-failure topography
observations of the rock mass structure and the condition of
the discontinuity surfaces. The application and limitations of
the GSI to rock slope characterization have been discussed in
Marinos et al. (2005) and Brown (2008). They discussed the
importance of the scale of observation, the need for a good threedimensional outcrop and how excavation techniques can reduce the
Fig. 5 Meteorological conditions
leading up to the rock slope failure
as recorded at the Agassiz weather
station (located approximately 30 km
from Chehalis Lake). Data from
Environment Canada (2009)
surface rock mass quality. The average GSI range obtained for the
rock mass at the Chehalis Lake landslide was 50–60. This
corresponds to a very blocky (four discontinuity sets or more) rock
mass with good (rough) surface conditions (Fig. 7a). Field estimates
of the intact rock strength of both the quartz diorite and andesite
suggested a strong to very strong or R4 to R5 range (∼100 MPa)
according to the strength classes defined in Hoek and Brown (1997).
The rock mass had a fresh (WI to WII) weathering state with some
localized zones of slight to moderate weathering (WII to WIII)
conditions (ISRM 1978).
Data collection
Discontinuity orientation data were collected at the Chehalis Lake
landslide using three different techniques. Upon increasing the
scale of observation, they are: traditional field data collection near
70
14
12
10
Precipitation (water equivalent)
Maximum temperature
Minimum temperature
60
Temperature (oC)
8
50
Date of the
Chehalis Lake
landslide
6
4
2
0
28/11/07
-2
40
30
29/11/07
30/11/07
01/12/07
02/12/07
03/12/07
04/12/07
05/12/07
06/12/07
20
-4
10
-6
-8
78
Landslides 9 & (2012)
0
Total water equivalent precipitation (mm)
Elevation (m.a.s.l.)
b
200
Fig. 7 Discontinuity sets in the rock
mass at the Chehalis Lake landslide.
a Bedrock exposed below the
initiation zone, b the initiation zone
(view toward northwest)
a
DS5
DS1
DS4
DS2
DS3
1m
b
Valley
slope
North
gully
wall
DS3
DS5
DS2
DS4
DS1
50 m
the landslide, terrestrial photogrammetry of the headscarp area
and colour-shaded relief maps from a 1-m resolution airborne
Light Detection and Ranging (LiDAR) of the area surrounding the
landslide. The location and extent over which each of the
techniques was applied are summarized in Fig. 3. The application
of remote sensing techniques was particularly suitable to this case
study due to the steep and inaccessible rock face at the headscarp
and because the slope is still considered to be unstable.
Field assessment
Detailed engineering and structural geology mapping was undertaken in the lower part of the Chehalis Lake landslide (Fig. 3). The
rock mass at each station was described according to the categories
and terminology described in ISRM (1978). At each station, the
authors evaluated the rock mass for the presence of discontinuity
sets with similar orientations. Once a discontinuity set was
identified, several measurements of its dip and dip direction were
collected using a Clar-type compass. A similar approach is described
in Mathis (1988) for cell mapping. The advantage of the field
assessment method is that, in addition to getting information about
discontinuity orientation, it can easily provide information about
discontinuity infill and roughness and about the rock lithology and
weathering grade. In turn, the disadvantages include limited access
to outcrops and potential danger to the person collecting the data
due to the unstable nature of the slope.
Landslides 9 & (2012)
79
Original Paper
Fig. 8 Detail of discontinuities
identified in the initiation zone of the
landslide based on the threedimensional model created from
terrestrial photogrammetry (view
toward west)
DS5
DS4
DS3
DS2
DS1
50m
Terrestrial photogrammetry
Sturzenegger and Stead (2009a, b) and Sturzenegger et al. (2009)
have demonstrated the applicability of long-range terrestrial
photogrammetry to measure the orientation of discontinuities in
large natural rock slopes and open pits. Their methodology was
applied to the Chehalis Lake landslide by acquiring digital
photographs using a Canon EOS 30D digital camera with 200mm focal length lens and processing the images obtained using
Fig. 9 Discontinuities identified in
the area surrounding the landslide
based on the airborne LiDAR DEM
coloured according to orientation
using COLTOP-3D
the 3DM CalibCam/Analyst software (Adam Technology 2009).
Three-dimensional low-resolution models were created for the
whole slope and higher-resolution ones for areas of interest in the
headscarp area (Fig. 3). Registration of the photogrammetric
models was achieved using the airborne LiDAR data to provide
control points. The ground resolution of the three-dimensional
models was calculated to be 38 cm (see Sturzenegger and Stead
(2009a) for details). The discontinuity orientations were obtained
N
N
DS4
W
DS1
E
DS2
S
DS3
DS5
Road
100 m
80
Landslides 9 & (2012)
Lake
by fitting circles to the discontinuity surfaces identified in the 3D
models of the headscarp (Fig. 8). The main advantages of the
terrestrial photogrammetry technique are that information about
the discontinuity orientation and persistence can be acquired with
no need to directly access the slope and it is possible to obtain
information from outcrops that cannot be physically accessed
while its limitation include occlusion and image resolution bias
(Sturzenegger and Stead 2009a, b).
COLTOP-3D
COLTOP-3D is a software package that uses high-resolution DEMs
to create colour-shaded relief maps (Jaboyedoff et al. 2004). The
pole of each DEM cell is assigned a unique colour based on its dip
and dip direction orientation. This facilitates the visual determination of the orientation of planar features, such as long
persistence discontinuities, fault scarps or landslide scarps
(Fig. 9). This approach has previously been applied successfully
to stability assessments of large rock slopes (e.g. Derron et al.
2005; Jaboyedoff et al. 2007; Pedrazzini et al. 2008; Jaboyedoff et
al. 2009; Metzger et al. 2009; Oppikofer 2009; Oppikofer et al.
2009). The advantages and limitations of this technique are
similar to those of the terrestrial photogrammetry listed in the
previous section.
Data interpretation
Field measurements
The orientation details of the discontinuity sets identified from
the field measurements are summarized in Table 1 and Fig. 10a.
Table 1 Summary of discontinuity
sets mean orientation for each of
the datasets acquired at the Chehalis Lake landslide
Five discontinuity sets were identified (Fig 7). Based on its
orientation and pervasiveness, discontinuity set (DS) 1 represents
surfaces that could have formed as sheeting joints (e.g. Hencher et
al. 2011), and together with DS2 they form the main failure surface
of the Chehalis Lake landslide. Based on its long persistence and
orientation approximately parallel to the north gully wall, DS3
appears to be associated with a fault that has the same orientation
as the lineaments observed by Arthur (1986) in aerial photographs. The surface characteristics of each discontinuity set are
presented in Table 2. The trace length of the DS5 was estimated to
be very low (<1 m); this is attributed to its dip into the slope,
creating overhanging surfaces that prevent the formation from
extensive exposures.
Terrestrial photogrammetry
The stereonet of the poles to the fitted circles extracted from the
f=200 mm lens models is presented in Fig. 10b. The spatial
distribution of the discontinuities identified in the 3D models
created from the terrestrial photogrammetry images is very
similar to the discontinuities identified based on field measurements. The same five discontinuity sets can be identified (Table 1;
Fig. 10b). Discontinuity sets 1 and 2 were extracted from the large
concentration of discontinuities identified in the photogrammetry
models based on the field observations. The biggest difference in
the orientation of the discontinuity sets is in the dip of DS1 which
is steeper in the photogrammetry models when compared with
the field measurements (Table 1). The merging of DS1 and DS2 in
the photogrammetry models is potentially due to three factors.
Firstly, because the scale of observation in the photogrammetry is
Dip/dip direction (o) N=number of measurements K=Fisher dispersion coefficient
Discontinuity set
Field measurements
Terrestrial photogrammetry
COLTOP-3D
DS1
28/150
43/149
38/198
N=17
N=68
N=8 surfaces
DS2
DS3
DS4
DS5
DS6
K=48
K=17
N=285 poles to DEM cell
59/135
61/146
67/149
N=13
N=37
N=8 surfaces
K=21
K=55
N=165 poles to DEM cell
67/062
77/074
51/052
N=12
N=26
N=8 surfaces
K=20
K=24
N=210 poles to DEM cell
74/200
76/222
64/246
N=16
N=19
N=8 surfaces
K=21
K=44
N=286 poles to DEM cell
43/292
55/307
12/297
N=19
N=15
N=8 surfaces
K=23
K=20
N=83 poles to DEM cell
NA
NA
20/101
N=1 surface
N=226 poles to DEM cell
Landslides 9 & (2012)
81
Original Paper
N
a
DS4
DS2
DS1
DS5
DS3
W
E
DS5
DS4
DS3
DS2
DS1
N=102 poles
to discontinuities
S
Field measurements
N
b
DS2
DS4
DS1
DS3
W
DS5
E
DS4
DS3
DS2
DS5
DS1
N = 202 poles
to fitted circles
Photogrammetry
line of sight
S
Terrestrial
photogrammetry
smaller (i.e. covers a larger area) than the field observations, DS1
and DS2 cannot be resolved as separate discontinuity sets. This
effect was highlighted in Sturzenegger and Stead (2009a).
Secondly, because the line of sight of the camera (trend/plunge
of ∼30/140°) when acquiring the digital photographs was similar
to the orientation of DS1 (mean dip/dip direction of ∼28/150° in
field measurements), it could have resulted in an orientation bias
(occlusion), as discussed in Sturzenegger and Stead (2009b) and
Lato et al. (2010), that would “miss” the shallower members of
DS1 (Fig. 10b). A third potential reason for the merging of DS1 and
DS2 is related to the contouring routine and the increased sample
size (field measurements: N=102; photogrammetry; f=200 mm;
N=202). This effect of the sample size has been discussed by
Stauffer (1966).
COTOP-3D
The airborne LiDAR data for the vicinity of the Chehalis Lake
landslide was used to create a DEM with a resolution of 1 m. This
DEM was used to identify 41 surfaces using COLTOP-3D which
represented a total of 1,389 DEM cells (i.e. each individual surface
is composed of multiple DEM cells; Fig. 10c). The spatial
distribution of the surfaces compares well with the discontinuity
data obtained from the field measurements and the photogrammetry models (Table 1). The dip direction of the DS1 from
COLTOP-3D is slightly more southward than was measured from
the other datasets. Discontinuity set five also has a noticeably
shallower dip in the COLTOP-3D results. This is attributed to a
directional bias as DS5 dips into the slope and is underrepresented in the airborne LiDAR data of the slope face on
which the failure occurred. A sixth discontinuity set was observed
in the stereonet from the COLTOP-3D results but not in the field
measurements or photogrammetric datasets. Using the orthophoto of the area to assess the origin of this surface, it was
tentatively attributed to the presence of small colluvial cones on
the slope and was disregarded in further slope stability analyses.
Slope stability analyses
N
c
DS2
DS2
DS1
DS5
DS5
DS4
DS3
DS6
W
DS3
E
DS5
DS5
DS4
DS4
DS2
DS2
DS6
DS6
DS1
DS1
S
N=1389 poles
to DEM cells
COLTOP-3D
Fig. 10 Stereonets of the poles to discontinuity surfaces identified in a field
measurements, b terrestrial photogrammetry and c COLTOP-3D. Lower
hemisphere equal angle stereonets
82
Landslides 9 & (2012)
Kinematic analysis
A kinematic analysis is a rock slope stability test for identifying
simple structurally controlled failure modes such as planar
sliding, wedge failure and toppling. It takes into consideration
discontinuity orientation, slope orientation and the friction angle
along the discontinuity surfaces. The stereographic techniques for
the kinematic analysis of these simple failure modes are described
in Richards et al. (1978). To account for the variation of the
topography in the immediate vicinity of the Chehalis Lake
landslide, two slope orientations were considered in the kinematic
analysis (valley slope at 35/115°, dip/dip direction; north gully wall
at 40/185°). The friction angle along all discontinuities was
assumed to be 30° based on the planar/undulating and rough
discontinuity surface conditions. Table 3 and Fig. 11 summarize
the results of the kinematic analyses performed using the three
discontinuity datasets. Overall, the north gully wall slope
orientation appears to be less stable than the valley slope. Sliding
along DS1 out of the north gully wall is feasible in all datasets and
marginally feasible out of the valley slope (Fig. 11a). Taking into
consideration the natural variability of the discontinuity set
orientations, wedge failures should be considered to be margin-
Table 2 Summary of the discontinuity surface characteristics
observed during the field
investigation
Discontinuity set
Large (metre scale)
roughness
DS1
Planar
Small (centimetre scale) roughness
Trace
length (m)
Rough some smooth in quartz diorite
1–3
Rough some slickensided in andesite
DS2
Planar and undulating
Rough some smooth in quartz diorite
3–10
Rough some slickensided in andesite
DS3
Planar
Rough some smooth in quartz diorite
1–3 and 3–10
Rough some slickensided in andesite
DS4
Planar
Rough some smooth in quartz diorite
1–3 and 3–10
Rough some slickensided in andesite
DS5
Undulating
Rough some smooth in quartz diorite
<1
Rough some slickensided in andesite
ally feasible out of the north gully wall but less so out of the valley
slope (Fig. 11b). Only random discontinuities were observed to fall
within the toppling envelope along both the north gully wall and
the valley slope (Fig. 11c).
Surface wedge analysis
Swedge is a limit equilibrium code (Rocscience 2006) that was
used to evaluate the stability of surface wedges in the rock slope
faces at the Chehalis Lake landslide. The combination analysis in
Swedge uses a list of discontinuities specified by the user (e.g.
discontinuity survey data) to calculate the factor of safety for each
valid wedge intersection for a given slope face. To account for the
topography in the immediate surroundings of the Chehalis Lake
landslide, the same two slope orientations (valley slope at 35/115°
dip/dip direction; north gully wall at 40/185°) considered in the
kinematic analyses were used in the combination analyses. The
analysis was performed using all three discontinuity orientation
datasets for both slope orientations. The results, summarized in
Table 4 and Fig. 12, strongly suggest that the north gully wall
orientation is more susceptible to wedge failures for all discontinuity orientation datasets. With the exception of the analysis
results for the north gully wall based on field measurements, all
datasets have consistent percentages of valid wedge failures
(Table 4). The discrepancy with the field measurements might
be related to the smaller number of discontinuity orientations in
that particular dataset compared to the other datasets.
Overall, the results from the surface wedge analysis are
consistent with those of the kinematic. Both methods suggest that
the north gully wall is more prone to failure. The advantage of the
surface wedge combination analysis is that it is capable of using
the full dataset to account for the natural variability in
discontinuity orientations at the site.
Three-dimensional distinct element modelling
The three-dimensional distinct element code 3DEC (Itasca 2008)
was used to model the interaction between the jointed rock mass
and the topography at the Chehalis Lake landslide. The 3DEC code
represents the rock mass as a collection of three-dimensional blocks
under static or dynamic loading. The strength of the material making
up the blocks and bounding discontinuities are specified by the user.
Large displacements and rotation along the discontinuities bounding
the blocks are permitted. More information about 3DEC can be
found in Cundall (1988) and Hart et al. (1988).
In this paper, 3DEC was used to evaluate the influence of:
(1) the mean discontinuity set orientations derived from the
various techniques, (2) individual discontinuities and (3)
variations in assumed friction angle along the discontinuities.
Rigid block conditions were assumed for the analyses pre-
Table 3 Summary of the kinematic analyses performed on the discontinuity datasets acquired at the Chehalis Lake landslide
Failure mechanism
Valley slope
North gully wall
Field measurements
Terrestrial photogrammetry
COLTOP-3D
Sliding
Feasible on DS1
Marginal on DS1
No
Wedge
Marginal on DS1/DS4
No
No
Toppling
Marginal on random
discontinuities
Marginal on random
discontinuities
Marginal on DS4
Sliding
Feasible on DS1
Feasible on DS1
Feasible on DS1
Wedge
Marginal along DS1/DS3
Marginal along DS1/DS3
and DS1/DS4
Feasible along DS1/DS4
and marginal along DS1/DS2
Toppling
Marginal on random
discontinuities
Marginal on random discontinuities
No
Landslides 9 & (2012)
83
Original Paper
Legend
pole to discontinuity
intersection of two
discontinuity sets
N
N
N
a
DS2
Daylight DS4
envelope
DS2
Valley
DS1
E
W
W
Daylight
envelopes
North
gully
DS3
DS3
DS1
DS4
Friction
cone
DS5
30o friction
cone
North
gully
DS2
DS4
DS1
Valley
E
W
DS5
Daylight
envelopes DS6
Friction
DS5 cone
Valley
DS3
E
North
gully
S
S
S
sliding
sliding
sliding
N
N
N
b
30o friction
circle
Valley
DS5
DS5
W
North
gully
DS3
DS2
DS5
DS1
E
DS4
W
DS3
DS3
DS4
DS2
North
gully
E
DS4
W
E
DS6
DS1
North
gully
DS2
DS1
Valley
Valley
S
S
S
wedge
wedge
wedge
N
N
N
c
Valley
Slip
limit
W
North
Slip gully
limit
Toppling
envelope
S
toppling
Field measurements
N = 102
E
Valley
W
E
W
E
Valley
Slip
limit
North
gully
Slip
limit
Toppling
envelope
Toppling
envelope
Toppling
envelope
S
toppling
Terrestrial
photogrammetry
N=202
North
gully
Slip
limit
Toppling
envelope
Slip
limit
Toppling
envelope
S
toppling
COLTOP-3D
N=1389 poles to
DEM cells
Fig. 11 Kinematic analysis conducted on the discontinuities a sliding, b wedge failure and c toppling failure mechanism for the field measurements, terrestrial
photogrammetry model and COLTOP-3D. Lower hemisphere equal angle stereonets
sented (i.e. the Chehalis Lake landslide is assumed to be a
structurally controlled rock slope failure). The rock density and
84
Landslides 9 & (2012)
discontinuity properties used in the 3DEC models are presented in Table 5 while a summary of the results obtained in
Table 4 Summary of the surface wedge combination analyses performed on the discontinuity measurements obtained with the various data collection methods used
the Chehalis Lake landslide
Field measurements
Valley slope
Number of combinations
5,151
Number of valid wedges
North gully wall
20,301
224
576
COLTOP-3D
963,966
31,378
Number of stable wedges
206 (92%)
564 (98%)
2,9579 (94%)
Number of failed wedges
18 (8%)
12 (2%)
1,799 (6%)
Number of combinations
Number of valid wedges
5.151
20,301
195
1,442
963,966
64,161
Number of stable wedges
166 (85%)
875 (61%)
41,279 (64%)
Number of failed wedges
29 (15%)
567 (39%)
22,882 (36%)
a Analysis for valley slope
100%
Unstable
Stable
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Field
Terrestrial photogrammetry
Photogrammetry
COLTOP-3D
Discontinuity orientation dataset
b Analysis for north gully wall
100%
Unstable
Stable
90%
80%
70%
the various models considered in this study is presented in
Table 6. A simplified geometry consisting of three planes
(valley slope at 35/115° dip/dip direction, north gully wall at 40/
185° and south gully wall at 40/080°), labelled in Figs. 2 and
13a, was used to approximate the average orientation of each
slope.
Discontinuity set orientations
Figure 13 compares the calculated total displacement, contoured
in metre, obtained after 30,000 numerical calculation steps in
models with the a mean discontinuity set orientation based on the
datasets acquired at the Chehalis Lake landslide. The threedimensional distinct element model associated with the field
measurement data resulted in a large slope failure on the valley
slope and the intersection between the valley slope and the north
gully wall (Fig. 13a). The volume of the unstable mass (arbitrarily
defined as the cumulative volume of blocks having a total
displacement >1 m; Table 6) obtained in this scenario is
approximately four times larger (11 Mm3) than the calculated
volume based on aerial photograph interpretation (3 Mm3). The
3DEC model using the mean discontinuity set orientations
obtained from terrestrial photogrammetry resulted in a small
volume (0.17 Mm3) slope failure at the intersection of the valley
slope and north gully wall (Table 6; Fig. 13b). The COLTOP-3D
dataset did not result in a significant slope failure (Fig.13c); only
very small blocks displayed any significant calculated displacement (Table 6).
60%
50%
40%
30%
20%
10%
0%
Field
Photogrammetry
COLTOP-3D
Discontinuity orientation dataset
Fig. 12 Cumulative bar graphs summarizing the percentage of stable and
unstable wedges obtained for each dataset along the a valley slope and b north
gully wall
Influence of individual discontinuity set
The influence of each discontinuity set on the slope stability
conditions was assessed by creating five three-dimensional
distinct element models that included only four discontinuity
sets, with each of the five discontinuity sets being omitted in turn
(Fig. 14). This series of models was built using the mean
discontinuity set orientations obtained from the field measurements dataset since it is the one which resulted in the largest
simulated slope instability. Where DS1 was omitted (i.e. the model
assumed that only DS2, DS3, DS4 and DS5 were present), all three
slopes were stable with only a few very small blocks moving more
than 1 m (Fig. 14a; Table 6). This result highlights the critical role
played by DS1 in the failure of the Chehalis Lake landslide. The
models that did not include DS2 (Fig. 14b), DS4 (Fig. 14d) and DS5
Landslides 9 & (2012)
85
Original Paper
Table 5 Rock density and discontinuity properties used in the three-dimensional
distinct element models
Material property
Density (kg/m3)
2,700
Discontinuity properties
Shear stiffness (GPa/m)
1
Normal stiffness (GPa/m)
5
o
Friction angle ( )
25 or 30
Cohesion (MPa)
0
Tensile strength (MPa)
0
(Fig. 14e) all resulted in similar maximum calculated total
displacements and unstable volume masses (53 to 71 m and 7.3
to 8.0 Mm3; Table 6). Visually, the unstable areas in these three
models (Fig. 14b, d, e) are similar to the equivalent model with
five discontinuity sets (unstable volume of 11 Mm3; Fig. 13a),
but the models with only four discontinuity sets had smaller
volumes with an average of ∼7.5 Mm3. Finally, the model that
did not include DS3 (Fig. 14c) resulted in a 1.7-Mm3 rock slope
failure along the intersection of the valley slope and the north
gully wall (Table 6). DS3 therefore plays an important role in
the development of an unstable rock mass at the Chehalis Lake
landslide.
Assumed friction angle
The previous 3DEC analyses were performed assuming a friction
angle of 25°. The sensitivity of the slope stability conditions to the
friction angle value was assessed by investigating a model with all
five discontinuity sets using the mean orientation from the field
measurements dataset, 40 m discontinuity spacing and an
assumed 30° friction angle. The model with the higher friction
angle value led to stable slope conditions with only small blocks
moving more than 1 m (Table 6).
Table 6 Summary of maximum
total displacements and cumulative
volumes of unstable blocks (blocks
with total displacement greater
than 1 m) obtained in threedimensional distinct element
models
86
Landslides 9 & (2012)
Discussion
Discontinuity orientation datasets
The three discontinuity orientation datasets acquired at the
Chehalis Lake landslide all revealed similar distribution patterns
(Fig. 10), although the mean orientation of the specific discontinuity sets varied (Table 1). This variation in orientation led to
some differences in the basic slope stability analyses (kinematic
and surface wedge), but the overall expected rock slope behaviour
based on the various discontinuity datasets was similar. In the
three-dimensional simulations, the orientation variations have a
larger effect. Three-dimensional distinct element models based on
the field measurements resulted in a large slope failure, while the
one using the terrestrial photogrammetry dataset resulted in
smaller slope failures. The model using the COLTOP-3D datasets
resulted in stable slope conditions. These analyses highlight the
fact that the stability of the blocks created by repeating
discontinuity sets and the topography is sensitive to the evaluation of the mean discontinuity set orientation.
The reasons for the variation in discontinuity set orientation
are threefold. Firstly, the various techniques sampled different
regions of the landslide and the adjacent area (i.e. geostatistical
variation of sample; Fig. 3). The field measurements were
collected at a few locations on the lakeshore and in the lower
part of the failure zone due to access restrictions. The terrestrial
photogrammetry was used to primarily map the headscarp
(initiation zone) of the Chehalis Lake landslide. The discontinuity
orientation data generated using COLTOP-3D were collected in
the area surrounding the landslide. As a result, the different
techniques sampled different (photogrammetry vs. field measurements) or several (COLTOP-3D) structural domains. Figure 15
conceptually illustrates that the dip angle of DS1 changes in the
vertical direction within the slope. In reality, the variation of DS1
is three-dimensional as both the dip and dip direction change
(stereonet in Fig. 15). To further demonstrate the spatial influence
of the dataset, COLTOP-3D was used to extract discontinuity
orientation from the point cloud created in the terrestrial
photogrammetry three-dimensional model. The results show that
Calculated
maximum
displacement
(m)
Cumulative volume
with displacement >1 m
(Mm3)
Dataset
Friction
angle (o)
DS not
included
Field
25
All included
57
Field
30
All included
13
Field
25
DS1
Field
25
DS2
53
8.0
Field
25
DS3
64
1.7
Field
25
DS4
57
7.4
71
3.9
11
0.00005
0.000003
Field
25
DS5
Photogrammetry
25
All included
3.3
0.17
7.3
COLTOP-3D
25
All included
7.9
0.00072
a Field measurements dataset, 40 m spacing
North gully
wall
(m)
South
gully wall
Valley
slope
b Terrestrial photogrammetry,
0
5
10
15
20
25
30
35
40
45
50
55
57
40 m spacing
(m)
0
0.25
0.50
0.75
1.00
1.25
1.50
1.75
2.00
2.25
2.50
2.75
3.00
3.25
3.29
c COLTOP-3D, 40 m spacing
(m)
0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
7.9
Fig. 13 Calculated total displacements, contoured in metre, for the threedimensional distinct element models using the mean orientation of the
discontinuity sets based on a field measurements, b terrestrial photogrammetry
and c COLTOP-3D datasets. A spacing of 40 m and friction angle of 25° was
assumed on all discontinuity sets
discontinuity sets DS1, DS2 and DS3 had a very similar orientation
of the mean values obtained using the 3DM Analyst (Adam
Technology 2009) software, while the dip direction of DS4 and the
dip angle of DS5 differ by approximately 20° (Table 7). The results
obtained when applying COLTOP-3D to the three-dimensional
model created in the terrestrial photogrammetry are more
consistent with the terrestrial photogrammetry dataset using
3DM Analyst (all of them sampling the headscarp area) than the
discontinuity extracted from the airborne LiDAR DEM (sampling
the area surrounding the landslide; Figs. 10 and 16; Tables 1 and
7).To fully assess the location/extent of structural domains, a
systematic field survey in the various sections of the landslides
and its surroundings would need to be undertaken.
The second factor that may account for the variability in
discontinuity set orientation is the relative accuracy of each of the
different data collection methods. Sturzenegger and Stead (2009b)
reported that variations in the order of ±4° for the dip and ±8° for
the dip direction can be expected between field measurements
and terrestrial photogrammetry discontinuity survey of the same
area. The third factor influencing the mean discontinuity set
orientation includes the biases associated with the different scales
and viewing angles at which the data were collected. For example,
the terrestrial photogrammetry was subject to occlusion (shadow
zone) due to camera (or laser scanner) perspective (Sturzenegger
and Stead 2009b; Lato et al. 2010) in the lower part of the failure
plane orientation (DS1), while the COLTOP-3D data were based on
the airborne LiDAR DEM which is based on data collected
vertically and cannot resolve discontinuities dipping into the
slope (e.g. DS5 at the study area). These three factors represent
sources of uncertainty associated with the data collection
methodology/processing and the structural model, all of which
transfer into the slope stability analyses (Read 2009).
Three-dimensional distinct element models
Based on the limited direct field assessment, the discontinuity
surfaces in the quartz diorite (Fig. 7a) were found to be planar/
undulating and rough while the discontinuities in the andesitic
rocks were smoother. These general surface conditions were
assumed to correspond to a friction angle of 30°, but the temporal
association of the failure with the rain-on-snow meteorological
event could reduce the mobilized effective friction angle. The
3DEC models demonstrated that a friction angle of 30° resulted in
generally stable slope conditions, whereas a friction angle of 25°
resulted in relatively large slope failures for a series of geometries
(Table 6). This friction angle threshold (25° to 30°) is consistent
with the dominant planar sliding failure mechanism along DS1
which has a mean dip of 28° based on field measurements. An
increase of pore water pressure associated with rain-on-snow
event could have lowered the effective friction angle below a
critical threshold.
The 3DEC model without discontinuity set DS3 (Fig. 14c)
led to a slope failure whose location and volume most closely
resemble the actual Chehalis Lake landslide, but since DS3
appears prominently in all the stereonet for all discontinuity
orientation datasets (Fig. 10), it cannot simply be discounted. A
review of outcrop-scale photographs at the northern edge of
the landslide (Fig. 7a) showed DS3 surfaces with limited
persistence, whereas photographs of the southern edge of the
slope failure showed DS3 surfaces with a larger persistence
(Fig. 8). These observations combined with the coincidence of
very-high-persistence DS3 with gullies (COLTOP-3D Fig. 9)
could represent the location of faults with limited persistence
Landslides 9 & (2012)
87
Original Paper
a DS1 not included
b DS2 not included
(m)
(m)
0
5
10
20
25
30
35
40
45
50
53
0
0.25
0.50
0.75
1.00
1.25
1.50
1.75
2.00
2.25
2.50
2.75
3.00
3.25
3.25
3.75
3.86
c DS3 not included
d DS4 not included
(m)
(m)
0
5
10
15
20
25
30
35
40
45
50
55
60
64
0
5
10
15
20
25
30
35
40
45
50
55
57
e DS5 not included
(m)
0
5
10
15
20
25
30
35
40
45
50
55
60
65
70
71
Fig. 14 Three-dimensional distinct element models investigating the influence of the individual discontinuity sets on the slope stability conditions. Models without a
DS1, b DS2, c DS3, d DS4, e DS5. Plots are of calculated total displacement contoured in metre. A spacing of 40 m and friction angle of 25° were assumed for all
discontinuity sets
DS3 in between them. The northwest strike of DS3 corresponds
with the northwest–southeast lineaments previously reported
by Arthur (1986) and the faults mapped by Thompson (1975)
88
Landslides 9 & (2012)
and Mahoney et al. (1995). A more detailed field investigation
should be combined with a comprehensive analysis of the
photogrammetric datasets to confirm this interpretation and to
N
Region sampled using
terrestrial photogrammetry
Fig. 15 Conceptual model illustrating
how the field measurements and the
photogrammetry datasets could
represent orientation data collected in
different structural domains
W
E
DATASET
S
Field measurements
Terrestrial photogrammetry
Region sampled
by field measurements
characterise the variation of the persistence of the discontinuity
sets over the landslide area.
Discontinuity set DS1 was shown to be critical to the
stability conditions of the modelled slopes. A variation of ±10°
in the dip direction and ±15° in the dip between the field
measurements and the terrestrial photogrammetry datasets
results in a variation of one order of magnitude in calculated
maximum total displacement and four orders of magnitude in
cumulative unstable volume in the 3DEC models (Fig. 13;
Table 6).
Conclusions
This paper presented the first detailed assessment of the failure
mechanism for the Chehalis Lake landslide. Field observations
indicate that the failure occurred at the intersection of the
valley slope and a prominent north gully wall. The rock mass
appears to have initially slid into the gully and was then
subsequently redirected along the gully toward the lake. The
Chehalis Lake landslide is also temporally associated with a
rain-on-snow meteorological event wherein high pore water
Table 7 Comparison between the mean orientation of the discontinuity set obtained from the terrestrial photogrammetry dataset using 3DM Analyst and COLTOP-3D software
Discontinuity set
Terrestrial
photogrammetry
using 3DM analyst
(dip/dip direction)
Terrestrial
photogrammetry point
cloud using COLTOP-3D
(dip/dip direction)
DS1
43/149
46/147
DS2
61/146
60/152
DS3
77/074
70/080
DS4
76/222
72/248
DS5
55/307
74/314
pressures could have provided the trigger for the rock slope
failure. Discontinuity orientations were collected at the base of
the landslide, in the headscarp and in the area next to the
landslide using field and remote sensing techniques. The three
discontinuity orientation datasets obtained all produced a
similar fracture pattern with a minor variation in the specific
orientation of the individual discontinuity sets. This variability
was attributed to the different sampling locations, the accuracy
of the different sampling techniques, the observation scale and
view orientation biases and data interpretation limitations
when processing different sample sizes. This highlighted the
importance of combining field and terrestrial photogrammetry
data to avoid sampling biases when defining the mean
orientation of discontinuity sets. Kinematic analyses using the
three datasets suggested that planar or wedge failures involving
DS1 were the likely failure mechanisms. Surface wedge analyses
suggested that the north gully wall was less stable than the
valley slope. Three-dimensional distinct element modelling
suggested that the Chehalis Lake landslide occurred due to
specific interactions between the discontinuity sets and the
topography. Based on these results, it appears that: (1) planar
sliding was the dominant failure mechanism; (2) the mobilized
effective friction angle along the discontinuities was less than
30° for the landslide to initiate, (3) the presence and
orientation of DS1 is critical for the development of the slope
failure and (4) DS3 has an important control on the volume
and extent of the unstable rock mass but less so on its overall
stability.
Acknowledgements
The authors would like to thank E. Fea and T. Sivak for their
assistance in the field. Funding for the fieldwork was provided
by BC Hydro and BC Ministry of Forests and Range. The
airborne LiDAR data were commissioned and provided by BC
Hydro. The authors would also like to acknowledge the editor
and reviewers for their constructive comments which improved
the paper.
Landslides 9 & (2012)
89
Original Paper
N
a
DS4
DS2
DS1
DS3
W
DS5
E
DS4
DS3
DS2
DS5
DS1
N = 202 poles
to fitted circles
Photogrammetry
line of sight
S
Terrestrial photogrammetry
using 3DM Analyst
N
b
DS2
DS4
DS1
DS3
W
E
DS5 DS4
DS3
DS2
DS5
DS1
S
Terrestrial photogrammetry
cloudpoint using COLTOP-3D
N = 8727 poles
to DEM cell
Photogrammetry
line of sight
Fig. 16 Stereonets comparing the results obtained from the terrestrial
photogrammetry dataset using the a 3DM-Analyst and b COLTOP-3D software
References
Adam Technology (2009) 3DM CalibCam and 3DM Analyst, version 2.3. Perth, Australia
Arthur AJ (1986) Stratigraphy along the west side of Harrison Lake, southwestern British
Columbia. Geol Survey Can Curr Res 86-1B:715–720
BGC Engineering Inc. (2008) Cohoe Creek landslide and landslide induced wave (seiche)
at Chehalis Lake. Report to BC Provincial Emergency Program, January 3, 2008. 31 pp
Bjerrum L, Jorstad F (1968) Stability of rock slopes in Norway. Norwegian Geotechnical
Institute, Report 79, 12 pp
Bornhold BD, Harper JR (1998) Engineering geology of the coastal and nearshore
Canadian Cordillera. In: DP Moore, O Hungr (eds), 8th International IAEG Congress.
Balkema, Rotterdam, pp 63–75
90
Landslides 9 & (2012)
Brideau M-A, Stead D, Couture R (2006) Structural and engineering geology of the East
Gate landslide, Purcell Mountains, British Columbia, Canada. Eng Geol 84(3–4):183–
206
Brideau M-A, Stead D, Couture R (2008) The 1999 Clanwilliam landslide: a preliminary
analysis of potential failure mechanisms. 4th Canadian Conference on Geohazards,
Quebec, pp 263–270
Brown ET (2008) Estimating the mechanical properties of rock masses. In: Potvin Y,
Carter J, Dyskin AV, Jeffrey R (eds) Southern Hemisphere International Rock
Mechanics Symposium. Perth, Australia, pp 3–22
Cundall PA (1988) Formulation of a three-dimensional distinct element model—part I:
a scheme to detect and represent contacts in a system composed of many
polyhedral blocks. Int J Rock Mech Min Sci Geomech Abstr 25(3):107–116
Derron M-H, Jaboyedoff M, Blikra LH (2005) Preliminary assessment of rockslide and
rockfall hazards using a DEM (Oppstahornet, Norway). Nat Hazards Earth Syst Sci
5:285–292
Dunlop S, Hutchinson DJ (2009) Rockslides in a changing climate: establishing
meteorological trigger thresholds in south western Norway. 62nd Canadian
Geotechnical Conference, Halifax, Canada, pp 1043–1050
Environment Canada (2009) National Climate Data and Information Archive. http://
www.climate.weatheroffice.gc.ca/; accessed September 10, 2009
Evans SG (1989) The 1946 Mount Colonel Foster rock avalanche and associated
displacement wave, Vancouver Island, British Columbia. Can Geotech J 26:447–452
Geertsema M, Highland L, Vaugeouis L (2009) Environmental impact of landslides. In: K
Sassa, P Canuti (eds) Landslides—disaster risk reduction. p 589–607
Genevois R, Ghirotti M (2005) The 1963 Vaiont landslide. Giornale de Geologia Applicata
1:41–52
Grimstad E (2005) The Loen rock slide—an analysis of the stability. In: Senneset K,
Flaate K, Larsen JO (eds) 11th international conference and fieldtrip on landslides.
Taylor and Francis, Norway, pp 129–135
Hart RD, Cundall PA, Lemos JV (1988) Formulation of a three-dimensional distinct
element model—part II: mechanical calculations for motion and interaction of a
system composed of many polyhedral blocks. Int J Rock Mech Min Sci Geomech
Abstr 25(3):117–125
Hencher SR, Lee SG, Carter TG, Richards LR (2011) Sheeting joints: characterisation,
shear strength and engineering. Rock Mech Rock Eng 44(1):1–22
Highland LM, Bobrowsky P (2008) The landslide handbook—a guide to understanding
landslides. UGSG Circular 1325, 129 pp
Hoek E, Brown ET (1997) Practical estimates of rock mass strength. Int J Rock Mech Min
Sci 34(8):1165–1186
ISRM (1978) Suggested methods for the quantitative description of discontinuities in
rock masses. Int J Rock Mech Min Sci Geomech Abstr 15:319–368
Itasca (2008) 3DEC v. 4.1. Itasca Consulting Group, Minneapolis
Jaboyedoff M, Baillifard F, Couture R, Locat J, Locat P (2004) Toward preliminary hazard
assessment using DEM topographic analysis and simple mechanic modeling. In:
Lacerda WA, Ehrlich M, Fontoura AB, Sayo A (eds) Landslides evaluation and
stabilization. Balkema, Rotterdam, pp 191–197
Jaboyedoff M, Metzger R, Oppikofer T, Couture R, Derron M-H, Locat J, Turmel D (2007)
New insight techniques to analyze rock-slope relief using DEM and 3D-imaging cloud
points: COLTOP-3D. In: Eberhardt E, Stead D, Morrison T (eds) 1st Canada–US Rock
Mechanics Symposium. pp 61–68
Jaboyedoff M, Couture R, Locat P (2009) Structural analysis of Turtle Mountain (Alberta)
using digital elevation model: toward a progressive failure. Geomorphology 103
(1):5–16
Jakob M, Weatherly H (2003) A hydroclimatic threshold for landslide initiation on the
North Shore Mountains of Vancouver, British Columbia. Geomorphology 54:137–156
Lato MJ, Diederichs MS, Hutchinson DJ (2010) Bias correction for view-limited LiDAR
scanning of rock outcrops for structural characterization. Rock Mech Rock Eng 43
(5):615–628
Mahoney JB, Friedman RM, McKinley SD (1995) Evolution of a Middle Jurassic volcanic
arc: stratigraphic, isotopic, and geochemical characteristics of the Harrison Lake
Formation, southwestern British Columbia. Can J Earth Sci 32:1759–1776
Marinos P, Hoek E (2000) GSI: a geologically friendly tool for rock mass strength
estimation, GEOENG 2000. Technomic, Melbourn, pp 1422–1440
Marinos V, Marinos P, Hoek E (2005) The Geological Strength Index: applications and
limitations. Bull Eng Geol Environ 64:55–65
Massey NWD, MacIntyre DG, Desjardins PJ, Cooney RT (2005) Digital geology map of
British Columbia: tile NN10 Southwest British Columbia, BC Ministry of Energy and
Mines, Geofile 2005–3
Mathews WH, McTaggart KC (1978) Hope rockslides, British Columbia, Canada. In: B
Voight (ed) Rockslides and Avalanches. Elsevier, New York, pp 259–275
Mathis JI (1988) Development and verification of a three dimensional rock joint model.
Ph.D. thesis, University of Lulea, Sweden, 144 pp
Metzger R, Jaboyedoff M, Oppikofer T, Viero A, Galgaro A (2009) Coltop3D: a new
software for structural analysis with high resolution 3D point clouds and DEM,
frontiers + innovation. Canadian Society of Petroleum Geologists, Canadian Society
of Exploration Geophysicists and Canadian Well Logging Society Joint Convention,
Calgary, Alberta, pp 278–281
Monger JWH (1970) Hope map-area, West half (92 H W1/2), British Columbia, Paper
69–47. Geological Survey of Canada, Ottawa, Ontario, 75 p
Oppikofer T (2009) Detection, analysis and monitoring of slope movements by
high-resolution digital elevation models. PhD thesis, Université de Lausanne,
193p
Oppikofer T, Jaboyedoff M, Blikra L, Derron M-H, Metzger R (2009) Characterization and
monitoring of the Aknes rockslide using terrestrial laser scanning. Nat Hazards Earth
Syst Sci 9:1003–1019
Panizzo A, de Girolamo P, di Risio M, Maistri A, Petaccia A (2005) Great landslide events
in Italian artificial reservoirs. Nat Hazards Earth Syst Sci 5:733–740
Peckover FL, Kerr JWG (1977) Treatment and maintenance of rock slopes on
transportation routes. Can Geotech J 14(4):487–507
Pedrazzini A, Jaboyedoff M, Froese C, Langenberg W, Moreno F (2008) Structures and
failure mechanisms analysis of Turtle Mountain. 4th Canadian Conference on
Geohazards, Quebec, pp 349–356
Plafker G, Eyzaguirre VR (1978) Rock avalanche and wave at Chungar, Peru. In: B Voight
(ed) Rockslides and avalanches, vol. 2 Engineering Sites, pp 269–279
Read JRL (2009) Data uncertainty. In: JRL Read, PF Stacey (eds) Guideline for Open Pit
Slope Design. CRC, pp 213–220
Richards LR, Leg GMM, Whittle RA (1978) Appraisal of stability conditions in rock slopes.
In: Bell FG (ed) Foundation engineering in difficult ground. Newnes-Butterworths,
London, pp 449–512
Rocscience (2006) Swedge v5.0. Rocscience, Toronto
Schwab JW (1999) Tsunamis on Troitsa Lake British Columbia. Forest Service British
Columbia Extension Note #35, 4 pp
Sidle RC, Ochiai H (2004) Landslides: processes, prediction and land use. AGU Water
Resour Monogr 18:312
Stauffer MR (1966) An empirical–statistical study of three-dimensional fabric diagrams
as used in structural analysis. Can J Earth Sci 3:473–498
Stephenson FE, Rabinovich AB (2009) Tsunamis on the Pacific Coast of Canada recorded
in 1994–2007. Pure Appl Geophys 166:1777–210
Sturzenegger M, Stead D (2009a) Quantifying discontinuity orientation and persistence
on high mountain rock slopes and large landslides using terrestrial remote sensing
techniques. Nat Hazards Earth Syst Sci 9:267–287
Sturzenegger M, Stead D (2009b) Close-range terrestrial digital photogrammetry and
terrestrial laser scanning for discontinuity characterization on rock cuts. Eng Geol
106:163–182
Sturzenegger M, Stead D, Beveridge A, Lee S, Van As A (2009) Long-range terrestrial
digital photogrammetry for discontinuity characterization at Palabora open-pit mine.
In: Diederichs MS, Grasselli G (eds) Third Canada–US Rock Mechanics Symposium.
Paper 3984, Toronto, 10 pp
Thompson RW (1975) Petrologic study of lower amphibolite metamorphism of igneous
rock, Chehalis Lake area near Harrison Hot Springs, British Columbia. B.Sc. thesis,
University of British Columbia, Vancouver, 44 pp
van Zeyl D (2009) Evaluation of subaerial landslide hazards in Knight Inlet and Howe
Sound, British Columbia. M.Sc. Thesis, Simon Fraser University, 184 pp
Wang FW, Zhang YM, Huo ZT, Matsumoto T, Huang BL (2004) The July 14, 2003
Qianjiangping landslide, Three Gorges Reservoir, China. Landslides 1(2):157–162
M.-A. Brideau : M. Sturzenegger : D. Stead : N. J. Roberts : B. C. Ward :
J. J. Clague
Department of Earth Sciences,
Simon Fraser University,
Burnaby, Canada
M. Jaboyedoff
Institute of Geomatics and Risk Analysis,
Université de Lausanne,
Lausanne, Switzerland
M. Lawrence
BC Hydro,
Burnaby, Canada
T. H. Millard
British Columbia Ministry of Forests and Range,
Nanaimo, Canada
M.-A. Brideau ())
School of Environment,
University of Auckland,
Auckland, New Zealand
e-mail: m.brideau@auckland.ac.nz
Landslides 9 & (2012)
91