Failure Behaviour of Two Landslides in Silt and Clay

FAILURE BEHAVIOUR OF TWO LANDSLIDES IN SILT AND CLAY

by

LARA ALEXANDRIA FLETCHER

B.A.Sc, The University of British Columbia, 1994

A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF

MASTER OF APPLIED SCIENCE

in

THE FACULTY OF GRADUATE STUDIES

(Department of Earth and Ocean Sciences) (Geological Engineering Program)

We accept this thesis as conforming to the required standard

THE UNIVERSITY OF BRITISH COLUMBIA

September, 2000

© Lara Alexandria Fletcher, 2000 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission.

Department

The University of British Columbia Vancouver, Canada

Date

DE-6 (2/88) Abstract

Two large landslides in overconsolidated glacio-lacustrine clay and silt deposits of British Columbia river valleys have been examined: the Attachie landslide, Peace River and the Slesse Park landslide, Chilliwack River. Both cases are quite similar in their main aspects. However, their failure behaviour was markedly different. One exhibited intermittent ductile deformations typical of compound landslides in stiff clay. The other developed suddenly into a catastrophic, extremely rapid flow-slide of 6.4 Mm3, following tens of years of slow, probably episodic movement. A comparison of the two cases is presented, with the purpose of explaining the difference in their failure behaviour. Comparison includes physiography, ground water characteristics, stratigraphy, material properties of the dominant units, and geometry. The most significant difference between the two cases, apart from scale (volume) is the presence at Attachie of a considerable content of non-plastic, cemented, unsaturated silt units.

Three possible mechanisms for brittle flow slide formation are proposed. The first, undrained brittleness, is the usual cause of flow slides in what are generally described as cohesionless, weakly bound or cemented, high porosity materials (Hutchinson, 1992). At small strains the soil structure collapses resulting in a sudden reduction in shear strength and allowing a flow slide to develop. The second is a process that leads to what can be described as "macroscopic" brittleness. In short, a substantial portion of the previously intact soil volume is transformed into a mass of blocks of intact clay and till separated by joints filled with a disturbed, loosely deposited matrix. Liquefaction of the loose material then occurs, exerting heavy fluid pressure inside the cracks. The third was proposed by Hutchinson (1987) to explain the 1963 Vaiont Slide in Italy. The factors which lead to this behaviour include: 1) a compound (markedly non- circular) shape of the bounding slip surface, so that the slide forms initially a kinematically inadmissible mechanism; 2) a strong contrast between a low shear strength, non-brittle bounding slip surface and a high strength, brittle slide mass. These conditions may produce a stored deficit in the overall factor of safety on the boundary shear surfaces, which can be suddenly released by a brittle failure along internal shears in the slide mass. Results of stability analyses suggest that a combination of the latter two mechanisms likely led to the development of the flow slide. TABLE OF CONTENTS Abstract -11 List of Tables iv List of Figures v Acknowledgements yi

1.0 Introduction 1 1.1 Objectives 3

2.0 Literature Review 4 2.1 Types of landslides in clay and silt 4 2.1.1 Rotational landslide 4 2.1.2 Translational slides ...... 6 2.1.3 Compound slides 7 2.1.4 Sudden spreading failures 9 2.1.5 Earthflows U 2.1.6 Flow slides 12 2.1.7 Complex slope movements 15 2.2 Landslides in glacio-lacustrine silt and clay in British Columbia and northern Alberta ..17

3.0 Methodology 19 3.1 Existing data 19 3.2 Fieldwork • 21 3.3 Laboratory Analysis • 22

4.0 Slesse Park landslide 24 4.1 Regional Setting 24 4.1.1 Physiography 24 4.1.2 Climate 24 4.1.3 Bedrock Geology 25 4.1.4 Quaternary history 25

4t1.5 Surficial geology 28 4.2 Site Description 30 4.2.1 Detailed stratigraphy 30 4.2.2 Groundwater characteristics 36 4.2.3 Landslide geometry and geomorphology 38 4.2.4 Soil properties 47 4.3 Failure behaviour 51 4.3.1 Main instability 51 4.3.2 Secondary movements 55 4.3.3 Trigger : • 58 4.3.4 Landslide dam 61 4.4 Slope stability analysis 62 4.4.1 Selection of strength parameters .-62 4.4.2 Sensitivity to shear strength 65 4.4.3 Sensitivity to pore pressure distribution 65

iii 5.0 Attachie landslide 67 5.1 Regional Setting 67 5.1.1 Physiography 67 5.1.2 Climate 69 5.1.3 Bedrock Geology 70 5.1.4 Quaternary history 70 5.1.5 Surficial Geology • 71 5.2 Site description 73 5.2.1 Detailed stratigraphy 73 5.2.2 Ground Water Conditions 76 5.2.3 Landslide geometry and geomorphology 83 5.2.4 Soil properties ; 95 5.3 Failure Behaviour 101 5.3.1 1973 Flowslide 101 5.3.2 Trigger 103 5.3.3 Landslide dam 106 5.4 Failure Analysis 106 5.4.1 Limit equilibrium analysis of initial widespread sliding 107 5.4.2 Dynamic analysis of flow slide 111 5.4.3 Flow slide mechanism 113

6.0 Comparison of two land slides 119

7.0 Conclusion 121

References 125

APPENDIX I Slesse landslide 132

APPENDIX II Attachie landslide 152

iv LIST OF FIGURES 1.1. Site location map = ...2 2.1. (a) Single and (b) Multiple rotational slip 6 2.2. Cross section of the Montagneuse River landslide, Alberta 9 2.3. Sudden spreading failure 10 2.4 A commonly occurring geometry of undrained loading in earthflows 12 2.5. Retrogressive, multiple rotational failure in sensitive clay 17 4.1.1 Glaciation of the Chilliwack Valley: (a) ca. 16 000 years BP; (b) ca. 11 500 years BP.... 27 4.1.2. Cross section revealing surficial geology at Slesse Park landslide 28 4.2.1. Detailed stratigraphy, Slesse Park landslide 31 4.2.2. Landslide dimensions 38 4.2.3. Exposure of toe of rupture surface near top of steep bluff face 41 4.2.4. Slickensided rupture surface 42 4.2.5. Horst feature indicating predominantly translational sliding at midslope 42 4.2.6. Section 1, Slesse Park landslide. Location shown on Figure Al 43 4.2.7. Section 2, Slesse Park landslide. Location shown on Figure Al .....44 4.2.8. Section 3/4, Slesse Park landslide. Location shown on Figure Al 45 4.2.9. Section 5, Slesse Park landslide. Location shown on Figure Al 46 4.2.10. Plasticity Chart, Slesse Park landslide 47 4.3.2. Slesse Park landslide: 1997 aerial photograph and overlay showing principal

geomorphological features :. 54 4.3.3. Earthflowof 1991, Slesse Park landslide 57 4.3.4. Earthflowof 1997, Slesse Park landslide 57 4.3.5. Intensity-frequency-duration plot, Chilliwack River valley 59 4.3.6. Chilliwack River Valley, Monthly precipitation (period of record 1984-1999) 60 4.3.7. Chilliwack River Valley, Monthly snowfall 60 4.4.1. Configuration of (a) ground surface and (b) sliding surface used in limit equilibrium analysis 63 4.4.2. Sensitivity of factor of safety to variations of the residual shear strength along the basal rupture surface 66 4.4.3. Sensitivity of factor of factor of safety to variations of the residual shear strength along the cross-bedding sliding surface 66 4.4.4. Sensitivity to variation of the depth of piezometric surface 67 5.1.1. Attachie landslide (photo courtesy of S.G. Evans) 68 5.1.2. Cross-section of the Attachie landslide showing stratigraphy and the 1973 sliding surface 72 5.1.3. Photograph of slide debris deposited at valley bottom taken within days of the 1973 event 76 5.2.1. 1970 Aerial photograph (BC7279;70) showing major and minor scarps in disturbed, pre-1973 failure area 79 5.2.2. 1973 Aerial photograph (BC5529;75) taken 40 hours after failure 80 5.2.3. Overlay of 1973 aerial photograph (BC5529;75) showing principal geomorphological features 81 5.2.4. Attachie landslide plan (original drawing B.C. Hydro 1016-C14-D1214). 82 5.2.5. Attachie landslide Section 1. See Figure 5.2.4. for location. See oversize 84 5.2.6. Attachie landslide Section 2. See Figure 5.2.4. for location. See oversize 85 v 5.2.7. Attachie landslide Section 3. See Figure 5.2.4. for location. See oversize 86 5.2.8. Exposure of toe of rupture surface in clay overlying basal gravels 87 5.2.9. Sharp ridges and narrow, discontinuous graben at foot of the main scarp 87 5.2.10. Plasticity chart of materials sampled from Attachie landslide 96 5.2.11. Consolidated, undrained triaxial compression tests of preglacial lacustrine clay 100 5.2.12. Consolidated, undrained triaxial compression tests of preglacial lacustrine clay 101 5.3.1. Monthly precipitation, Fort St. John Airport, May 1972 to May 1973 104 5.3.2. Monthly rainfall, Fort St. John Airport, May 1972 to May 1973 105 5.4.1. Configuration of the (a) pre-slide ground surface and (b) rupture surface of the Attachie landslide 109 5.4.2. Two dimensional model of Attachie landslide lower stage 110 5.4.3. Output from DAN analysis of Attachie landslide Ill 5.4.4. A simplified, approximate analysis of the balance of forces operating at the initiation of the flow slide 112 Al Slesse Park Landslide Geomorpho-fogical Map 172

vi LIST OF TABLES

4.2.1. Depth to groundwater in Piezometer 1.. 37 4.2.2. Landslide dimensions 37 4.2.3. Summary of classification test results, Slesse Park landslide 48 4.2.4. Unconfined compressive strength estimated from pocket penetrometer. 49 4.2.5. Shear strength estimated from shear vane in Auger Hole 2 49 4.2.6. Summary of classification test results adjacent to the main sliding surface 51 4.4.1. Summary of soil properties used in stability analysis .....64 5.2.1. Attachie landslide dimensions. Numbers refer to Figure 4.2.2 83 5.2.2. Summary of classification test results, Attachie landslide 95 5.2.2. Summary of classification test results adjacent to the main sliding surface 97 5.4.1. Summary of soil properties used in CLARA 108 5.4.2. Summary of soil properties used in Sarma limit equilibrium analysis 118 6.1. Comparison of key parameters, Attachie and Slesse Park landslide 120

vii ACKNOWLEDGEMENTS

My deepest gratitude extends to my advisor, Dr. Oldrich Hungr for kind encouragement and support not only in my pursuit of higher education, but also in my pursuit of adventure in the mountains. I wish to thank the members of my committee for their time and valuable suggestions; Dr. J. Howie and Dr. J. Fannin of the UBC Department of Civil Engineering and Dr. S. Evans from the Geological Survey of Canada. Dr. Evans also provided financial support and documentation regarding the Attachie landslide.

I appreciate the generosity of my friends who volunteered to cover themselves with mud, scratches and insect bites to assist me in the field: Dana Ayotte, Julia Matsubara, Guy Edwards, Matthias Jacob, and Andrew Watson. Thank you to Andrew Watson for his undergraduate thesis on the Slesse Park landslide and his drafting skills.

There are many others to whom I am grateful for their willingness to assist and educate me. Mr. R. Enegren patiently allowed repeated access to files. Mr. B. Gerath of Thurber also allowed access to files and enthusiastically accompanied me in the field. It was an honour to also have had the company and input of Dr. J. Clague of the SFU Department of Earth Sciences. I wish to thank Mr. B. Thomson of the Ministry of Environment, Lands, and Parks for his shared enthusiasm for a day in the field and for his knowledge of the Slesse Park landslide.

I thank the'faculty and staff of the Department of Earth and Ocean Sciences for their efforts and assistance on my behalf. Mr. R. Wust offered his overextended time to assist me in the laboratory, and interpret the results of x-ray diffraction. Mr. A. Toma of the Mining Division Research Unit allowed use of their plotter.

Finally, I wish to acknowledge the support, friendship, and constant inspiration I received from Dana Ayotte and Rina Freed. Without them, this endeavor would have been an altogether different and less enjoyable experience. 1.0 Introduction

In valleys across British Columbia and Alberta glacio-lacustrine sediments have been deposited in glacial lakes formed by the blockage of natural drainage during the advance and recession of glaciers. Subsequent glacial over-riding and post-glacial erosion of overburden sediments have overconsolidated the laminated silt and clay sediments. These sediments are subject to widespread slope instability where processes of oversteepening, stress release and valley rebound associated with fluvial degradation have occurred. The presence of weak layers of clay contribute to the instability of these slopes.

This thesis examines two landslides of this kind: The Attachie landslide on the Peace River in

Northeastern B.C. and the Slesse Park landslide on the Chilliwack River of Southwestern B.C.

Figure 1 shows the locations of these slides. The two cases are quite similar in their main aspects.

Both occur in heavily overconsolidated, laminated clay and silt, which was once over-ridden by glacial ice. They have compound profiles, consisting of a weak shallow-angled basal layer and a steep back scarp. The surface of rupture in both cases daylights high on the bank above a river and is underlain by pervious strata of sand and gravel. Both involve volumes measured in millions of m3.

There is, however, one major difference between the two cases. The Slesse Park slide exhibits behaviour typical of slides in overconsolidated clay (e.g. Skempton and Hutchinson, 1969): slow

to rapid, intermittent displacements controlled by temporal changes in pore water pressure,

development of scarps and sag depressions, localized earth flow movements and piecemeal

delivery of material to the crest of the toe scarp, where it rapidly slides into the river below. The

volume of the main instability is approximately 2Mm3. Figure 1.1. Site location map. Source of Peace River map: Evans et al. (1996), source of Chilliwack River map: Clague et al. (1988).

2 The Attachie slide behaved similarly until the spring of 1973, accumulating some tens of metres of displacements since at least 1952. Then, on May 26, 1973, a major flow slide was triggered. The failing mass of approximately 12 Mm3 started moving suddenly, 6.4Mm3 formed a flow slide which descended a 50 m bedrock scarp and travelled nearly 1 km to its distal limit on the opposite side of the Peace River floodplain. The debris generated a displacement wave that downed trees on the opposite bank and blocked the flow of the Peace River for about 10 hours (Evans et al.,

1997). Local inhabitants reported that noises generated by the slide lasted over a period of only

10 minutes (Thurber, 1973). This type of extremely rapid flow slide failure is not the usual mode of movement in insensitive, overconsolidated glacio-lacustrine sediments.

1.1 Objectives

One of the primary objectives of this research was to summarize all existing data and reports, and together with the results of continued investigation, provide a complete case history of each landslide. This in itself is considered to be a valuable contribution to the relatively sparse literature available regarding landslides in overconsolidated glacio-lacustrine silt and clay.

A further objective is to explain the apparently anomalous behaviour of the Attachie slide. It is

considered that a detailed comparison with the Slesse slide, which did not exhibit a similar flow

slide phase, may be helpful in bringing out factors responsible for it. In so doing, it may also be

determined whether or not the potential exists for the Slesse Park landslide to develop into a flow

slide similar to that of the Attachie.

3 2.0 Literature Review

2.1 Types of landslides in clay and silt

The body of literature available on the subject of landslides in clay is extensive and several

comprehensive reviews classifying types of slope movement have been published (Skempton and

Hutchinson, 1969; Varnes, 1978; Hutchinson, 1988; Cruden and Varnes, 1996).

The following types of mass movement have been selected as pertaining to the Attachie or the

Slesse landslides; 1) rotational slides, 2) translational slides 3) compound slides, 4) sudden

spreading failures, 5) earthflows, 6) flowslides, and 7) complex slope movements. A brief

description of each and exemplary case histories are provided.

2.1.1 Rotational landslide

Rotational failure generally occurs in a relatively uniform fill or natural stratum, especially clay or

shale, under either drained or undrained conditions. Rupture takes place along an identifiable slip

surface that approximates a cylindrical or listric surface. Movement thus imparts a degree of

backward rotation or tilt to the slipping mass which is accompanied by sinking at its rear and

heaving at its toe. Figure 2.1a illustrates a single rotational failure.

"In natural slopes, failures which appear to have been of circular rotational type have

occurred in cliffs of over-consolidated, fissured London Clay at Warden Point, Kent

(Hutchinson, 1968) and in over-consolidated, intact clay till at Selset, Yorkshire

(Skempton and Brown, 1961)." (Skempton and Hutchinson, 1969)

4 Terzaghi and Peck (1948) suggest that slides in varved clay deposits may also be approximately circular providing that the porewater pressure in the silt layers is inconsequential.

Movement rates of slumps can vary by several orders of magnitude. According to (Buma and

Van Asch, 1996), rotational rock slides can move at speeds varying between a few centimetres a year to several metres per month, while soil slumps can attain velocities of up to three metres per second.

If the main mass of the slide moves down very far, or the initial slipped mass is eroded away, the result may be a new failure at the crown of the slide. Retrogression of single rotational slips thus results in the formation of two or more backward rotated slipped blocks. Hutchinson (1988) suggests that "multiple rotational landslips are generally of large scale and appear to be characteristic of situations where the strata are sub-horizontal and consist of a relatively thick

layer of stiff, fissured clay or shale, underlain by a more competent stratum and overlain by a

substantial capping stratum of well-jointed but otherwise strong rock." Multiple rotational slips

are absent or rare in cliffs formed entirely of stiff, fissured clays. In the absence of a competent

overlying stratum the steep scarp at the head of initial slip degrades rapidly, thus the level of stress

does not again rise high enough to bring about a further deep-seated failure. When such

circumstances exist, retrogression may continue until a stable slope of very low gradient is

attained. A multiple rotational failure is illustrated diagrammatically in Figure 2.1b. Bishop's

Simplified Method was developed, and is routinely used, for circular sliding surfaces (Bishop,

1955).

5 An example of this type is found near Quesnel, B.C. (Evans, 1982). Active slope movements are taking place in the upper part of a 150 m thick sequence of varved Quaternary glacio-lacustrine clays and silt. Retrogression at the main scarp takes place by multiple rotational slumping and produces backward tilted slump blocks which move downslope and undergo progressive

disintegration and remolding. The rate of scarp retreat was found to vary between 6 and 12

m/year. Rate of movement of the displaced slide mass has not been recorded.

Figure 2.1. (a) Single and (b) Multiple rotational slip (from Buma and Van Asch, 1996)

2.1.2 Translational slides

In translational sliding, the mass progresses along a more or less planar surface that is largely

controlled by surfaces of weakness within the structure (Skempton and Hutchinson, 1969).

In some cases translational failure may occur as a result of liquefaction of the basal slip surface.

The quick clay landslide at Furre, Norway was broadly a translational failure known as a flake

slide which failed on a thin, gently inclined layer of quick clay subject to liquefaction (Skempton

and Hutchinson, 1969). Hutchinson describes the landslide:

"The central part of the slide, forming nearly half of the total area, slid out as an almost

intact flake approximately at right angles to the river. During sliding, this flake sank and

tilted backwards, so that several small pools were impounded at its rear. The rear

6 boundary of the slide is marked by a steep scarp of maximum height 8m. Below this

scarp, for about 350m in the central part of the slide, the sliding surface consists of a layer

of very quick clay, 10cm or less thick, and inclined between 6° and 9°. Although the

actual failure in the ground spread in a matter of seconds, the subsequent forward

movements of the slide masses took appreciably longer. The central part of the slide

would appear to have come to rest after about 1 to 1.5 minutes."

The landslide on the Toulnustouc River, Quebec, was also described by Conlon (1966) as a flakeslide initiated along a liquefied sand parting. The large landslide associated with the Alaska earthquake of 1964 also resulted from liquefaction of silt and sand sized material (Seed and

Wilson, 1967).

2.1.3 Compound slides

Compound slides are intermediate between rotational and translational failures. The slip surface

consists of a steep, curved or planar rearward part intersecting a flatter sole (Hutchinson, 1988).

In contrast to circular and planar failures, in which movement can occur with relatively limited

internal distortion, compound slides are locked in place as a result of their slip surface geometry

and can move only when the slip mass is released by the development of internal displacements

and shears. Hutchinson makes a major distinction concerning the brittleness of the internal

failure.

"Where the brittleness is low to medium, the speed of failure is generally moderate. The

principal internal dislocations consist of one or more shear surfaces towards the rear of the

slide mass, inclined steeply into the slope and springing from the area where the slip

surface has the smallest radius of curvature. The rear part of the slide sinks down between

7 these internal shears and its rear scarp, often with little or no backward rotation, to form a

graben, and the middle part of the slide moves forward translationally. Such slides may

be divided into those with listric or bi-planar slip surfaces. In the former type, there are

usually two or more internal rearward slip surfaces: in the less common, latter type only

one is required."

Compound slides in which the slide mass has high brittleness, the speed of failure can be extremely rapid. The catastrophic failure of 1963 at Vaoint, Italy (Hendron and Patton, 1985) provides an example of this latter type with speeds of 20 to 30 m/s estimated. Though not consisting of a clay and silt slide mass, the rupture surface is believed to have exploited a clay seam.

The 1939 Montagneuse River landslide, Alberta (Cruden et al, 1997), is an example of a compound slide in which the surface of rupture followed clays in pre-glacial valley sediments deposited in the buried channel of the Peace River. The volume of the slide was 76 Mm3 and was reported to have lasted a minute. Displacements were limited to approximately 200 m. A reconstruction of the rupture surface is shown in Figure 2.2. Velocity estimates were not reported, though the landslide has been classified as very rapid. As the major part of the landslide is translational, the procedures of Morgenstern and Price (1965) were used for the analysis.

8 Figure 2.2. Cross section of the Montagneuse River landslide, Alberta illustrating compound rupture surface (from Cruden et al., 1997).

Hypothetical surfaces of rupture r-400 400 -i 350 350 1000 1200 1400 0 200 400 600 BOO metres

2.1.4 Sudden spreading failures

Sudden spreading failures constitute a category of translational failure in which the dominant mode of movement is lateral extension accommodated by shear or tensile fractures (Varnes,

1978). Movement is very rapid to extremely rapid and over low slope gradients (Hutchinson,

1988). They are characterized by the succession of graben and horst structures, the horst

structures frequently taking the form of sharp ridges. Odenstad (1951) developed a model of this

retrogression process, based on field observations in Sweden. The process is illustrated in Fig.

2.3. The backscarp retreats by subsidence of a wedge-shaped block bounded by normal faults.

This subsidence drives the front of the slide mass forward. As it subsides the lateral thrust against

the new backscarp may become too small to hold the backscarp in place. In that case a new

wedge will subside, pushing forward the previous backwall as a ridge-capped prism, and the

retrogression process will continue. Movement may involve fracturing and extension of coherent

material owing to liquefaction or plastic flow of subjacent material. The coherent upper units may

subside, translate, rotate, or disintegrate, or they may liquefy and flow. The mechanism of failure

can involve elements not only of rotation and translation but also of flow; hence, lateral spreading

failures may be properly regarded as complex, a classification which will be discussed

subsequently.

9 Figure 2.3. Sudden spreading failure (from Mitchell, 1978).

REGRESSION RATE

The South Nation River landslide of May 1971 provides an example of retrogressive spreading failure (Mitchell, 1978). Mitchell describes the morphology of the South Nation River as follows:

"Over most of the crater, a succession of crescent shaped, relatively intact and grass

covered blocks of surficial sand are separated by intruded silts and fine sands. The intact

blocks have been translated horizontally away from the backscarp and have dropped

vertically between 8 and 15m. In general the grassed blocks had extremely little surface

tilt, but the locally tilted portions had surface inclinations ranging from 30° backtilt

(dipping toward the backscarp) to 5° foretilt, with no definable pattern. The lower

portion of the crater spoil and the river spoil was fairly well mixed, but not completely

remoulded. Near the mouth of the crater a few intact pinnacles of grey silty clay stood as

evidence of rotational sliding. Broken pinnacles of clay and large blocks of silty and sandy

materials surrounded by remoulded silty clay filled the river channel to a depth of 8 to

10m. Many mature trees stood upright in this spoil, still rooted in the original sandy

10 topsoil. The total surface area of the land washed away by the slide was approximately

268 000 m2 and the average depth to the failure surface was 22.5m."

This type of failure is perhaps the most common in soft, sensitive soils which typically have a softer stratum underlying a stiff, stronger cap (Cruden and Varnes, 1996). However, Terzaghi and Peck (1948) noted that failures of this type occur in varved clays with high pore-water pressures in the silt or sand layers. Lateral spreading was reported to have occurred in fresh water varved clays at Wawbewawa and at Beattie Mine, Quebec (Mitchell, 1978).

2.1.5 Earthflows

Hungr et al. (in review) define an earth flow as a rapid or slower, intermittent flow of plastic, clayey earth. "Earth flows are commonly tongue or teardrop shape. They have rounded, bulging toes and sinusoidal longitudinal profiles, concave upward near the head of the earth flow and convex upward near the toe" Keefer and Johnson (1983). Keefer and Johnson limit movement to velocities of a few meters or less per day that persists for several days, months, or years whereas rapidly moving bodies of liquefied material is classified by Varnes' (1978) as a "rapid earthflow" or a flow slide (Hungr et al, in review) as described in subsequent paragraphs. The fastest recorded rates in "surging" earth flows are of the order of 0. lm/s (Hutchinson et al., 1974). Their

ability to flow on very flat angles can be explained by the highly disturbed state of the cohesive

material, resulting in swelling to a high water content and a low strength. As described by

Hutchinson and Bhandari, (1971) movement of an earth flow occurs by the mechanism of

undrained loading of the headward part of the slide by debris discharged from steeper slopes to

the rear. The development of forward thrust by the undrained loading of the rear part of the

mudslide, where the basal slip surface is inclined fairly steeply downwards in shown diagrammatically in Figure 2.4. Such types of movement are also referred to as earthslides, mudflows or mudslides.

Figure 2.4 A commonly occurring geometry of undrained loading in earthflows (from Hutchinson and Bhandari, 1971). Discharge of debris producing rapid loading at rear of sliding mass where basal slip surface inclined fairly steeply downwards Profile of lower Piezometric levels In freshly part of loaded area sliding mass

Pieiometric levels' Slip Forward thrust from loaded In situ material on slip surface surface area limited by passive of low permeability resistance at any section

The Thistle Slide, Utah (Duncan et al., 1986), is an example of an earth flow involving silty,

sandy, or gravelly clays of medium plasticity (average LL = 40; PI = 18). The landslide mass is

about 1800m long with an elevation difference of 305m from the crest of the main scarp to the

toe of the slide debris. In April, 1983, reactivation of an earthflow complex resulted in

displacements in excess of 150m over a slope of about 10°. Rates of movement of up to 1.8 m/hr

were measured.

2.1.6 Flow slides

There remains much ambiguity in the definition of the terms used to describe the mechanism by

which landslides exhibit rapid, brittle failure resulting in flow-like mass movement. Hutchinson

(1992) provides the following definition:

12 "Flow slides are characterised by a suddenness of failure, following some disturbance, and

a rapid and extensive run-out, commonly over very gentle or horizontal ground, which

comes to rest abruptly. The materials involved are generally cohesionless, weakly bound

or cemented, and are invariably of high porosity. These properties give them a metastable

structure. When, in response to some disturbance, a collapse of this structure is initiated,

the overburden load is suddenly transferred onto the pore-fluid, with the consequent

generation of high excess pore-pressure. As a result, a sudden reduction in effective

normal stress and hence in shear strength occurs which gives the failing material, briefly, a

semi-fluid character and allows a flow slide to develop."

Flow slides occur in loose, cohesionless materials; in lightly cemented, high porosity silt; and high porosity, weak rocks. The distance traveled is related to the brittleness index, though it is also a function of the geometry of the slipping mass, the topography of the ground in front of it, and the scale of the slide.

A disastrous flow slide which resulted in the loss of 144 lives occurred in colliery waste at

Aberfan, Wales, 1966 (Bishop, 1973). A rotational slip, probably triggered by artesian pore pressure at the base and toe, turned into a flow slide. The flow slide travelled down a 12.5° slope

for a distance of about 490 m to a Junior School which it largely destroyed. A drop in the

value of <}>' along the slip surface from a peak of 39.5° to a residual of 18° yields a Brittleness

Index, Ib, of 61/ The in situ moisture content of much of the tip lay in the range of 16-18%.

Density measurements of the intact remainder of the tip were very low (1536 and 1499 kg/m3 at

depths of 16 and 28 m below the crest) when compared with the density of the flow slide material

13 near the village (1717 and 1851 kg/m3 in shallow pits). Thus, evidence was provided that the

material was capable of undergoing a very large decrease in volume on being subjected to shear

displacements.

The 1880 landslide on the Thompson River, near Ashcroft, B.C. (Evans, 1984) involved

approximately 15 X 106 m3 of normally consolidated Early Holocene or late Pleistocene (post-

deglaciation) glaciolacustrine sediments. The sediments are characterized by varying and

laminations. Each rhythmite consists of a thick silt layer on top of a thin clay band. The silt bands

grade from one inch at the top of the section to 6 m thick at the base (Fulton, 1965). Although

the cause of the landslide is speculative, it was suggested that it was due to the irrigation of the

bench in the area of the failure (Stanton, 1898). When the silt became saturated to such an extent

that it could not sustain its own weight the whole mass suddenly dropped almost vertically and

flowed out, damming the Thompson River for approximately 44 hours.

Some authors have used the term "rapid earthflows" for flow slides in fine-grained silt, clay, and

clayey sand (Geertsema and Schwab, 1997; Mitchell, 1978; and Mitchell and Markell, 1973).

These flows form a complete gradation with slides involving failure by lateral spreading, but they

involve not only liquefaction of the subjacent material but also retrogressive failure and

liquefaction of the entire slide mass (Varnes, 1978). If significant disintegration of the failed mass

has occurred it may be difficult to identify the predominant mechanism based on post failure

morphology without the aid of eyewitness accounts.

All sequences of sliding were actually present in the catastrophic landslide at Rissa, Norway in

1978 (Karlsrud et al., 1984). The sliding initiated by a minor rotational failure in earth fill on the

14 shore of Lake Botnen. Retrogressive sliding then took place over a 40 minute period and

subsequently stopped in that area. Shortly thereafter, however, a huge flake type slide occurred

in an adjacent area. There was also some sinking and lateral spreading along the west side of the

slide bowl. The slide finally covered an area of 3.3km2 and involved about 6 million m3 of clay.

In British Columbia, flow slides in both glacio-marine and glacio-lacustrine sediments have been

documented. At the Sharktooth and Inklin landslides, B.C., Geertsema (1998) attributes

pyramidal prisms to portions of ridges that translated along a nearly horizontal failure plane and

resemble prisms in other translational flow slides (Evans and Brooks, 1994; Geertsema and

Schwab, 1997).

2.1.7 Complex slope movements

Slope movements most often involve a combination of two or more types of movements. These

are termed complex slope movements. There are many possible combinations and gradations

from one type to the other defies rigorous, discrete classification. In fact, in any of the above

mentioned types of slope movement, the propagation of a failure surface and the development of

a failure scarp may occur while the landslide mass remains intact. As the slide mass moves

downslope it deforms and/or breaks up into independent blocks. As deformation and

disintegration continue, and especially as water content or velocity or both increase, the broken or

disrupted slide mass may change into a flow.

Due to their relatively common occurrence, Hutchinson (1988) classifies landslides breaking

down into mudslides or flows at the toe into two categories: 1) slump-earthflows and 2) multiple

rotational sensitive-clay slides.

15 1) The term slump-earthflow is used to describe slope movement by which the already broken and disrupted toe of the landslide breaks down further, by the processes of spreading, Assuring and softening, to form a mudslide or earthflow.

2) Multiple rotational failures are commonly initiated by a single rotational failure that may be analyzed as a drained condition, as illustrated by Figure 2.5. Once an initial slip has developed, the shear strength along the rupture surface is decreased to residual (ultimate). Further slipping and disturbance will reduce the residual strength to the remoulded strength, which may be only a few percent of the undisturbed strength in quick clays and saturated silts (Meyerhof, 1957).

Thus, depending on the sensitivity, the failed material will have the consistency of a slurry or consist of an intact slump block sliding along a surface layer of negligible shearing strength. In either case, the failed material offers practically no resistance to movement so that the material moves away. Thus, hardly any counterbalance or stability is provided to the oversteepened, exposed bank (Mitchell, 1978). "Under this condition the main scarp will normally be unstable with respect to long term stability. Removal of lateral support due to the initial failure would generate negative excess pore water pressures in the backscarp, however, and the factor of safety

against immediate retrogression should be estimated using a pore water pressure factor, ru, less

than that established for the initial slide. Occasionally this reduction in ru may be sufficient to prevent retrogression, at least temporarily. If retrogressive rotational slides do develop each successive retrogression must overcome a greater average shearing resistance due to the changing initial stress conditions as the backscarp becomes further removed from the initial slope. At some stage in this process, further slope failure should be analyzed as an undrained (<|)=0) phenomena."

(Mitchell and Markell, 1973).

16 Figure 2.5. Retrogressive, multiple rotational failure in sensitive clay (from Mitchell and Markell, 1974).

STIFF BASE

Retrogressive rotational flowslides may be distinguished by the characteristic 'flow bowl' and the

backward tilted linear ridges transverse to the flow direction (ribbed morphology). As the slump

blocks move down slope and undergo progressive disintegration and remolding residual pinnacles

of clay may remain in the partly remoulded spoil. Further down the slide, it may no longer be

possible to identify any remnant intact material. The spoil may flow out over very gentle to

horizontal ground, attesting to the very weak, fluid like consistency. The Nicolet flowslide,

Quebec, is an example of a retrogressive, multiple rotational failure (Beland, 1956)

2.2 Landslides in glacio-lacustrine silt and clay in British Columbia and northern Alberta.

Glaciolacustrine sediments, deposited in glacial lakes formed by the blockage of natural drainage

during the advance and retreat of glaciers of the Fraser glaciation, are found in the Interior

Plateau, in the Rocky Mountain Trench north of Cranbrook, in numerous valleys in the major

mountain ranges, and the Interior Plains of northeastern British Columbia and northwestern

Alberta (Evans, 1982). Thick sequences occur at elevations below 100m asl. in interior valleys

across the province (Eyles and Clague, 1991).

17 According to Stumpf et al (1998) hazard potentials in interior valleys are generally moderate to high in glacio-lacustrine sediments having: (a) a major silt component (Fulton 1965) or (b) mica concentrated along distinct bedding planes (Quigley 1976). Failures are most common at sites with high subsurface pore water pressure, commonly associated with perched water tables or artesian conditions (Evans, 1982). Glacio-lacustrine sediments with a major clay component are also subject to failure but generally exhibit slower movement.

Few landslides involving glaciolacustrine sediments underlying a till capping have been documented in British Columbia though several have occurred in the Interior Plains of adjacent northwest Alberta. Cruden et al. (1997) presented an analysis of the 1939 Montagneuse River landslide, a rapid landslide of 76Mm3 involving high plasticity preglacial channel deposits overlain by till. Nasmith (1964) described landslides on the Meikle River in which the seat of failure lies in heavily over-consolidated, stiff, fissured lacustrine silt and clay, and appears to be of multiple rotational type. Cruden et al. (1993) presented an analysis of Alberta's largest historic landslide, the 1990 Rycroft landslide (est. vol. 40 Mm3) on the Saddle River. The compound rupture surface was described as dipping steeply through postglacial lake clay and till, curving abruptly to pass on a subhorizontal surface through preglacial clay. No rates of displacement were indicated.

Lu et al., (1998) reviewed 5 major landslides that have occurred on tributaries of the Peace River,

Alberta. These landslides are reactivated, retrogressive, multiple, rapid translational earth slides.

Surfaces of rupture in four cases follow bedding in preglacial lake clays.

In other areas of B.C. a landslide in post-glacial lake clay was documented by Thomson and

Mekechuk (1982) on the Fraser River, near McBride. Re-initiation of this landslide was

18 considered to be due to an increase in pore pressure within the slide mass. Ingress of water to the slide was influenced by the stratigraphy, which allowed for vertical infiltration through an alluvial fan buried by lake sediments. Water is then transported along seams of relatively pervious silty material within the lake sediments contributing to increased, even artesian, pore pressures. Evans

(1982) makes a distinction of those glacio-lacustrine sediments dominated by silt varves, and those dominated by clay varves. He noted large failures of the slump-earthflow variety occurring in the Southern Interior glacio-lacustrine sediments typically dominated by silt varves. The more plastic nature of materials that typify the northern and central interior of British Columbia tends to give rise to multiple retrogressive slides with a number of multiple retrogressive flow slides being present. As noted above, flow slides in glacio-lacustrine sediments have also been reported in north western B.C. (Geertsema and Schwab, 1997).

3.0 Methodology

3.1 Existing data

The Attachie slide occurred within the reservoir of B.C. Hydro's proposed Site C hydroelectric power development, approximately 38 km upstream of the proposed Site C dam site. In response to the Attachie flow slide and concern for another catastrophic failure of the proposed Site C reservoir slopes, BCHydro conducted extensive field investigations of the slopes adjacent to the proposed reservoir area. Investigations were carried out between 1973 and 1982 during which time 8 boreholes were drilled. Of these 8 holes, 6 were instrumented with a total of 12 standpipe

piezometers and 2 pneumatics and 2 were instrumented with inclinometers. In addition, a system

of extensometers and survey monuments was installed in various locations across the slope. Data

from various instrumentation provide information on the stratigraphy, groundwater conditions,

19 location of the failure surface, and rate of movement of the slope immediately above, as well as

downstream of the 1973 slide. The data is attached in Appendix II.

Thurber Consultants were contracted to perform lab analyses to classify and characterize the

material properties of the predominant units. Classification tests included Atterberg limits and

grain size analysis. Material properties were determined through the use of drained direct shear

tests, and a consolidated undrained triaxial test. Two reports were prepared by Thurber

Consultants which included a physical description of the slide, a reconstruction of the slide events,

an explanation of the failure mechanism and a slope stability analysis.

Aerial photographs of the slide were taken on May 28 within 40h of its occurrence. These,

together with aerial photographs taken of the Attachie slopes in 1967, enabled large scale maps

showing topography before and after the 1973 slide to be prepared.

The only published documentation of the Attachie slide is a summary paper written by Evans et al.

(1996). However, all B.C. Hydro internal reports, and Thurber Consultants reports were made

available, a list of which is included in the References.

Other published information useful to the interpretation of the physical setting, geology and

glacial history of the area was a paper published by Mathews (1978) for the Geological Survey of

Canada.

Earth flows in March 1991 and January 1997 at the Slesse Park landslide caused temporary

blockages of the Chilliwack River. The slide activity sparked concern amongst the Department of

Fisheries and Oceans and area citizens about fine-grained (silt and clay) sediment which issues

20 from the slide area and causes degradation of fish habitat. Landsliding and blockage of the River also raised potential public safety issues. To address these concerns a geotechnical investigation of the Slesse landslide was conducted by Thurber Engineering. However, the investigation was limited and no subsurface investigation was done. The report prepared by Thurber Consultants

(1997) described the landslide, the results of laboratory testing of 8 samples and a slope stability analysis. At this time a 1:2500 scale topographic map of the study area was prepared from March

1997 aerial photos.

A Masters thesis completed by Saunders (1985), followed by papers published by Saunders et al

(1987) and Clague et al. (1988) interpreted the glacial history of the area and the late Quaternary

geology. Detailed mapping of the stratigraphic succession and morphology of the slide was

completed in an undergraduate thesis by Watson (1999).

The present study began by reviewing the above information as well as relevant engineering and

geological publications. Air photos of both slides and climate records were examined.

3.2 Fieldwork

At Slesse, detailed geomorphological field mapping was carried out to supplement previous work.

Mapping was undertaken with a compass, clinometer, hipchain and measuring tape. Transects

were taken from the scarp to the toe, parallel with the direction of slide movement. Features such

as surficial materials, vegetation type, orientation of trees, seepage, scarps, tension cracks,

detailed morphology of the ground surface and other features were recorded and mapped at a

scale of 1:2500 (Figure Al). At Attachie, transects were also taken to record important

21 geomorphological information. However, as the slide occurred in 1973, most of the morphological details have not been preserved.

Hand excavation was carried out to obtain samples which had not been significantly dried or wetted by direct surface exposure. Careful excavation using shovel, ice ax, and wire cutter allowed isolation of an intact block which was then wrapped in plastic for transport. The block samples were waxed for preservation of natural water content if testing was not to commence immediately.

At the Slesse Park landslide, laborious drilling using a hand auger was required to install two piezometers near the center of the slide mass at the edge of the flow bowl created by Central

Creek. Piezometers consisted of 1 Vi" PVC piping with slots cut into the bottom 30cm. No casing was installed. The full depth of the piping was sealed by natural squeezing of the test hole in remoulded clay. To test the in-situ shear strength of the remoulded clay a shear vane was designed by Dr. Hungr and constructed in the machine shop. The dimensions of the rectangular vane are 21mm wide and 120mm long. Torque was measured with a spring scale. Undrained compressive strength was estimated in the field with a standard pocket penetrometer manufactured by ELE International.

3.3 Laboratory Analysis

A full range of index testing was carried out on samples collected in the field. Atterberg limits were determined using a liquid limit device according to Lambe's Soil Testing for Engineers

(1991). Specific gravity was determined using a Multivolume Pycnometer 1305 manufactured by

22 Micromeritics Instrument Corp. Bulk density was calculated as the mass over the volume of water displaced by a submerged block sample.

Particle size distribution was determined using the Sedigraph 5100 X-ray system manufactured by

Micromeritics Instrument Corporation. The clay mineralogy was determined by X-ray diffraction using a Siemens D5000 Diffractometer.

3.4 Slope Stability Analysis

Three-dimensional models of the rupture surfaces of both landslides were reconstructed using a

detailed kinematic analysis of deformation features such as scarps, grabens, sag ponds and rotated

blocks surveyed on the ground surface. In the case of Attachie, this was augmented by borehole

data. The pre-slide slope surface was based on topographic maps prepared from airphotos.

Approximate digital elevation models (DEM's) of the rupture surface of both slides were

constructed from this analysis. A range of slope stability analyses was carried out, with the aim of

explaining post-failure behaviour.

Limit equilibrium analysis was completed using CLARA (Hungr et al., 1989) , a three-

dimensional extension of the Bishop's simplified method; a 2D version of CLARA using both

Bishop's and Spencer's method and Hoek's (1987) modification of the Sarma method.

The flow slide phase of the Attachie landslide was modeled using DAN (Hungr, 1995). DAN is a

numerical model of unsteady flow, developed exclusively for the analysis of rapid landslides.

23 4.0 Slesse Park landslide

4.1 Regional Setting

4.1.1 Physiography

Chilliwack River occupies a broad, west-trending valley displaying a classic glacially eroded cross section. The peaks are steep and rugged while at lower elevations the profile has been modified by thick deposits of valley fill. Local relief may reach 2000m (Saunders, 1985).

The Slesse Park landslide area is located on the outside of a meander bend along the north bank of the Chilliwack River (Figure Al). The slide area is bordered by Noonie Creek to the west and by what is referred to as Plantation Creek to the east. Central, West and East Twin Creeks flow through the centre of the unstable area, feeding water directly to the slide. Creeks above the slide are confined in very steep gullies. The head of each gully, except for the east fork of Central

Creek, is located below Army Road. There is a distinct, sloping plateau above the gully heads south of Army Road which is the south edge of the Ryder Lake Upland.

There is almost complete forest cover consisting of hemlock, fir, alder and cedar with weak understorey species although the area has been logged in the past, prior to 1940. The terrain is irregular, often hummocky and very typical of extensive slope instability with steep backscarps, flat boggy areas and distortion of trees.

4.1.2 Climate

The climate of the Chilliwack Valley is typical of Coastal BC with heavy precipitation during the winter and warm, moist summers. Snow packs develop in the alpine, and occasionally snow may accumulate as low as valley bottom though it does not remain for extended periods. The

24 biogeoclimatic zone is classified as Coastal Western Hemlock (B.C. Ministry of Forests, 1988) with an average annual precipitation of 1340 mm.

4.1.3 Bedrock Geology

Monger (1994) refers to the bedrock geology of the study area as belonging to the Chilliwack terrane, one of four terranes that can be identified in the Northwest Cascades System. It is comprised of "Devonian to Permian arc volcanics, clastic rocks and limestones of the Chilliwack

Group, stratigraphically overlain by Upper Triassic and Lower Jurassic fine grained elastics of the

Cultus Formation and Upper Jurrassic pelites." The rocks exhibit complex folding and faulting.

4.1.4 Quaternary history

Clague et al. (1988) have summarized the Quaternary glacial history concisely. "During the

Fraser Glaciation, Chilliwack valley was affected both by local, north and west-flowing valley glaciers and by a lobe of the Cordilleran Ice Sheet that flowed south and south-west across Fraser

Lowland from the Coast Mountains (Clague and Luternauer 1982; Hicock et al. 1982a; Saunders

et al. 1987). In this region, an early phase of Late Wisconsinan glaciation (ca. 19 000 - 25 000

years BP) was followed by glacier recession (ca. 18 000 - 19 000 years BP) and then by the

climactic, or Vashon, advance of the Cordilleran Ice Sheet (Hicock and Armstrong 1981; Hicock

et al 1982a, 1982b). During the Vashon Stade, a piedmont lobe filled Fraser Lowland and the

Strait of Georgia and flowed south into Puget Lowland, Washington. The Puget lobe reached its

limit near 47°N latitude about 14 000 - 14 500 years BP and began to retreat soon thereafter

(Rigg and Gould 1957; Mullineaux et al. 1965; Hicock and Armstrong 1985). Chilliwack valley

is more than 50 km inside the Late Wisconsinan glacial limit and thus presumably was covered by

ice 14 000 years ago."

25 It is believed that the glacio-lacustrine deposits at the site accumulated ca. 16 000 BP as a lobe of the Cordilleran Ice Sheet advanced south and south -west across Fraser Lowland blocking drainage of the Chilliwack River valley. Evidence to support this hypothesis comes from three radiocarbon dates of wood collected from within the glacio-lacustrine sequence of 15 950 ± 110,

16 000 ±180 and 16 100 ±150 years BP. These dates mark the latest ice free conditions in

British Columbia and thus more closely delimit the Fraser glacial climax. They also suggest that there was a major buildup of ice in this region between 16 000 and 14 000 - 14 500 years BP.

(Clague et al. 1988)

During the latter deglacial portion of the Fraser glaciation, ice in the eastern Fraser Lowland again blocked the mouth of the Chilliwack valley at a time when the mid reaches of the valley were ice free. "A glacial lake formed between the ice dam at the mouth of the valley and an outwash delta formed in the upper reaches of the valley (ca. 11,700 BP). Several relatively minor advances of

Fraser Lowland ice separated by brief periods of recession occurred ca. 11,500 and 11,200 years

BP and were responsible for another series of glacial lake conditions within the valley" (Saunders et al. 1987). Exposures of glacio-lacustrine sediments at similar elevation 2-3km away are overlain by till dated at 11 500 - 11 200 years BP.

By approximately 11,150 years BP ice disappeared from the lower reaches of the Chilliwack

Valley and subsequent drainage of glacial lakes completed. The ice damming of the Chilliwack valley ceased approximately 11,000 BP with the completion of deglaciation of the Fraser Lowland

(Saunders et al. 1987).

26 Figure 4.1.1 Glaciation of the Chilliwack Valley: (a) ca. 16 000 years BP; (b) ca. 11 500 years BP (modified from Saunders et al. 1987)

Following retreat of the valley ice, vigorous paraglacial conditions ensued. It was proposed by

Saunders (1985) that many of the alluvial fans in the area probably originated at this time.

Landslide activity in the middle Holocene is responsible for depositing localized valley fill and

temporary damming of the Chilliwack River (Saunders 1985).

27 4.1.5 Surficial geology

A thick, complex sequence of late Wisconsinan sediments fills the Chilliwack river valley at low elevation. Gullying, mass wasting, and fluvial erosion act to expose the stratigraphy which is summarized in Figure 4.1.2. Saunders (1985) suggests that the Slesse landslide (referred to as

Alison Pool in his thesis) is covered by a glaciolacustrine sequence that extends along the length of the Chilliwack river between Slesse and Vedder Crossing. Exposures of glacio-lacustrine sediments at a similar elevation 2-3 km downstream were overlain by till dated asl5610±130

BP believed to be an extension of the Ryder upland surface. Till was not found in the vicinity of the Slesse Park landslide.

Figure 4.1.2. Cross section revealing surficial geology at Slesse Park landslide

250 250

Cobble/gravel 230 I i Laminated clay and silt 230 t--:'.i'?s-j Sand mmm Coiiuvium 210 210

190 e u 170

E 170 150

150 130 20 40 60 80 100 120 140 160 180 200 220 240 260 280 . 300 130 meters

The lowest unit exposed in the vicinity of the Slesse landslide consists of well-rounded cobbles

and pebbles supported in a coarse sand matrix. It is horizontally stratified with minor cross bedding. It is interpreted to be glacio-fluvial in origin.

28 Overlying the gravel are fine grained glacio-lacustrine sediments. The contact is sharp and can be surveyed almost continuously for 500 m along the river at El. 158 m. The glacio-lacustrine sequence consists of laminated silt and clay interbedded with lenses or layers of fine sand partings of maximum 5 m thickness. In places the sediments are typical of varved structure. The silt and clay materials are very stiff and overconsolidated with low permeability whereas the sand layers are zones of seepage. There are discontinuous exposures of glacial lake material as high as 260 m, 112m above the river floodplain. However, it is unlikely that the deposition of a continuous valley floor occurred to this elevation. It is more likely that the Chilliwack Valley glacier lay along the center of the lower valley when sedimentation was active, thus the lacustrine sediments deposited high on the valley walls were ice marginal. Thus there may be kame deposits, mass wasting, or deposits affected by turbid conditions.

Glaciolacustrine sediments dominate the stratigraphy, accounting for approximately 91% of the source volume, while the remaining 9% is the granular surface veneer. Laminated silts and clay account for approximately 78% of the lacustrine sequence and the sandy facies 22%. The sands are interpreted as minor delta or channel deposits.

A discontinuous veneer of dense sandy gravel covers the landslide features formed in the glaciolacustrine sediments. It is estimated that this unit constitutes 9% of the source volume.

The faint wavy bedding and stratification found in some locations suggests that the material may have been glacial outwash deposited following deglaciation. In other locations, the lack of sorting

suggests that these materials result from mass wasting of the slopes above the landslide area.

29 4.2 Site Description

4.2.1 Detailed stratigraphy

Due to difficulties of site access, no drilling has been undertaken. The surficial geology has therefore been constructed by careful mapping of exposures revealed by landslide activity and alluvial erosion. Locations of these exposures, each assigned a section number, are shown on

Figure Al of the Appendix. A summary of the quaternary stratigraphy interpreted from these exposures along Central Creek gully and the main back scarp (Watson, 1999 and Clague et al.

1988) is shown in Figure 4.2.1. The stratigraphy is described from the lowest exposed unit upwards.

Unit 1: consists of well-rounded cobbles and pebbles with an average diameter of 10 cm and maximum diameter of 30cm, supported in a coarse sand matrix. It is horizontally stratified with minor cross bedding. The coarseness, poor sorting, and lack of well defined bedding suggests a flood deposit, perhaps glacial outwash. Based on the near vertical face within these materials and limited erosion even where continuing seepage flows discharge down the face, these lower granular deposits appear to be very dense and overconsolidated with high permeability though relatively dry. Though a colluvial apron now covers most of this lowest exposed unit, photographs document the cobbles and pebbles extending continuously across the 500m of bluffs.

Unit 2: rests conformably upon Unit 1. The contact at El. 158 m can be traced east/west for

500m, dipping 1° to the south. Unit 2 consists of laminated silt and clay interbedded with layers or lenses of fine sand partings. This Unit is interpreted to be of glacio-lacustrine origin.

Dropstones 0.2-3cm in diameter are randomly scattered. The layers are typical of varved deposition. The materials are very stiff and overconsolidated with low permeability.

30 Figure 4.2.1. Detailed stratigraphy, Slesse Park landslide

1-T

31 Unit 3: is a well sorted fine and medium sand with a minor silt component. The sand is bedded, with ripples and small scale cross bedding. Zones of massive sand are also present. Beds dip gently toward the west. The unit is 6 m thick, from El. 167 to 173 m, and exhibits a general coarsening upward sequence. The top lm contains mainly coarse sand with occasional pebbles; most of the pebbles are concentrated at or near the upper contact. The sands are dense and non- cemented. They exhibit high permeability and erosion by piping processes commonly occurs at the bottom of this unit, along the contact with the underlying low-permeability clays. The upward- coarsening, dipping structure and the close association with lacustrine sediments suggest Unit 3 was deposited as a small delta.

Units 1-3 are located within the steep bluffs along the river and the steep gully walls eroded by

Central Creek (Sections 1-3). A nearly continuous sequence is exposed from river level at El.

148 m to El. 220 m. Units 1-3 are interpreted to be relatively undisturbed as suggested by their continuity and by the near-horizontal bedding.

Unit 4: is exposed between El. 173 and 207 m in Central Creek gully, and in limited exposures at the same elevation. Unit 4 is another glacio-lacustrine sequence of laminated silts and clay similar to Unit 2. The main failure plane is located within this unit at approximately El. 178 within the

Central Creek gully, 5m above the base of Unit 4 (Figure 4.2.3). The rupture surface exploits a

clay lamination 11mm thick. Clay at the failure plane is homogeneous, light grey, stiff, and highly

plastic. A 16mm sand lens overlies the clay lamination. The sand is very fine, brown, dense and

contains sheared blobs of clay. Above the failure plane, a zone several metres thick has been

remoulded to varying degrees as evidenced by lack of laminations and sheared blobs of clay.

Above this zone the clay and silt is again laminated with bedding dipping 4° to 20° into slope

32 suggesting varying degrees of backward rotation. Secondary flame and fold structures are exhibited at El. 193. Unit 4 is believed to extend laterally across the full width of the slide at approximately the same elevation. At Section 7, Unit 4 is exposed in the back scarp of a recent earth flow. Here it is a massive clay with some minor laminations dipping 4° to the south-west.

At Section 10, Unit 4 is exposed as a stiff, laminated clay with some stones in the back scarp of a minor mudflow located approximately 20 m east of the western lateral scarp at El. 225 m.

Overlying the clay is deltaic sand, with bedding dipping 25° into the slope (Unit 5).

At Section 13, stiff, massive silt is exposed in the back scarp of a minor slump located between

170 - 180 m. This is the approximate elevation of the failure plane. However, the exposure is west of the western lateral scarp beyond which the main failure surface does not continue. It is possible that the material is sheared lacustrine silts of Unit 4 and this section may be the back

scarp to a secondary slide that appears in the 1940 airphotos at this location. The silt material is

overlain by dense gravel fining upwards to a sandy gravel similar to Unit 7. This is the western

most location Unit 4 was observed.

Unit 4 was again exposed at. Section 19, east of the eastern lateral margin. At El. 185 m, a stiff,

horizontally laminated clay was observed. 1 cm lenses of clayey silt are separated by 2mm silt and

fine sand laminations. This exposure marks the eastern most location laminated silt and clay has

been observed.

Unit 5: The glaciolacustrine sequence exhibits a coarsening upward with the upper 2m grading

from silt to uniform, fine grained sand, referred to as Unit 5. Unit 5 is similar to Unit 3 though

33 large rip-up clasts of laminated clay were observed at the top of this unit. Unit 5 is generally associated with and overlying clay sediments.

Unit 6 is found as a discontinuous veneer over most of the landslide area. Unit 7 has a very dense sandy matrix, with some silt, which supports subangular clasts of gravel (modal 4 cm) and some boulders (maximum 80cm). There is no sorting or structure except in some locations where there appears to be faint, wavy bedding. Bedding suggests that this unit may be outwash sands and gravels deposited following deglaciation. The sediments are dense to very dense.

Units 4, 5 and 6 are exposed within the western lateral scarp at El.200 m (Section 12). In the sandy gravel unit, Unit 6, a silty sand matrix supports subangular-subrounded clasts 20cm average diameter. The clasts constitute 60% by volume. Some are composed of laminated clay. The sandy gravel is dense and moist. A very fine sand lense (Unit 5), 10-15cm thick lies between the sandy gravel and the underlying clay (Unit 4). The sand is rust colored and is a zone of seepage.

Units 4, 5 and 6 are located in the headscarp between Central Creek and West Twin (section 14).

The exposure consists of a diamicton with large cobbles and pebbles, maximum diameter 15 cm,

set in a silty clay matrix. Clasts are approximately 20% by volume. The fabric is poorly sorted

and massive. This material is believed to be a silty clay facies of glaciolacustrine sediments with

dropstones (Unit 4). Overlying the diamicton is a 5m thick sand unit (Unit 5). The texture of

this unit is variable along the scarp, ranging from sandy clay to a medium sand. The sand unit has

many seepage erosion pipes, some as large as 0.75m diameter. Overlying the sand is another

diamicton similar to Unit 7. The silty clay matrix supports a higher percentage of cobbles (40%)

with average diameter 20cm, and no pebbles. Some of the clasts are composed of laminated clay.

34 In this location, the backscarp is approximately 10 m high. Unit 6 is approximately 3m thick and is covered by 2.5 m of sandy gravel outwash (Unit 7). The lower 4.5 m is covered by a colluvial

apron.

Unit 7: At the intersection of Central Creek and the head scarp (Section 5), a contact between

the glaciolacustrine sediments and an overlying diamicton, originally believed to be till, is

observed. The diamicton is very hard and composed of a low plasticity clay matrix supporting

subangular to subrounded clasts with average diameter 25cm. Some appear to be rip-up clasts of

laminated clay with maximum diameter 6 cm. A radiocarbon date of 8 700 ± 100 years old

indicates that the diamicton is likely colluvium deposited by a debris flow.

Units 8 and 9 are exposed along the backscarp, at Section 5, in a coarsening upward sequence of

glaciolacustrine sediments. Unit 8 is a laminated to massive clayey silt, exhibiting flame and load

structures, grading to dense, massive, sandy silt with some stones, flecks of charcoal and rip up

clasts of clay. Unit 8 in turn grades to a fine sand (Unit 9). The thickness of the sand varies from

l-3m. The glaciolacustrine sediments are dissected by a steep, unconformable erosional contact

with sandy gravel. It appears that a channel was incised into the lacustrine sediments which was

then filled with sandy gravel (Unit 7). Units 8 and 9 are again exposed due to recent sloughing in

the bank of West Twin Creek was exposed at El. 250m (Section 16). This section is similar to

that exposed at Section 6 and is at approximately the same elevation. Massive clay (Unit 8) is

bound by a conformable contact with an overlying fine sand (Unit 9).

Units 8 and 9 appear to extend continuously between El. 230 to El. 260 as small exposures were

encountered at El. 250 and clay overlain by sand was exposed again at El. 260m in the banks of

35 West Twin Creek (Section 17). In section 17, the sand is bedded, dipping approximately 15° to the south. Clay is also encountered to an elevation of 260 in the western portion of the main scarp at Section 7, suggesting that this Unit of clay may extend continuously between sections 7 and 17, but are concealed by glacial outwash or colluvium. Another possibility is that the clay/silt sediments are discontinuous, and the area subject to sliding is controlled by the presence of these low strength materials.

4.2.2 Groundwater characteristics

It has been noted that slope movement is triggered by events of heavy precipitation and/or snowmelt and subsequent groundwater conditions. Due to the lack of instrumentation monitoring pore pressures, it is difficult to make quantitative descriptions of ground water conditions within and adjacent to the slide.

Given the temperate coastal climate, and that the slide area is located near the steep valley bottom, it is likely that the ground water table is relatively high and subsurface soils remain wet all year. This is evidenced by seepage in the face above the river even during the dry, summer months. The lacustrine units act as an impervious cover over the granular facies beneath the rupture surface. Therefore, perched water tables exist and the near surface soils are usually saturated by local infiltration, as observed in the shallow piezometer installed in slide debris near the toe. The 6 m deep piezometer was located at the edge of the Central Creek gully at approximately El. 186m. During augering for piezometer installation, water filled the hole to within 10cm of the ground surface within one hour of installation. Subsequent water levels are

shown in Table 4.2.1.

36 Ta jle 4.2.1. Depth to groundwater in Piezometer 1 Date March, 1999 April, 1999 July, 1999 Sept., 1999 installation date Depth to water (m) 0.10 3.35 3.18 3.43

Due to hazards of working in the vicinity of Central Creek gully when the site is wet, no piezometric data has been recorded during winter months. It is assumed that during the winter perched water tables are at ground surface as evidenced by the abundance of ponded water.

Because the basal gravels are overlain by low permeability sediments and probably do not have a good connection to any recharge area they presumably act to drain the slopes. Seepage interception by horizontal sand layers has been evidenced by piping erosion, though it is not known how continuous these units are. If the sand and gravel layers were fully saturated it would be expected that more vigorous seepage erosion along the toe scarp face would be evidenced.

The accumulation of colluvium at the base of the slope may contribute to pore pressure build-up in the slope and instability of the colluvial material.

Groundwater may further contribute to slope failure in that softening may occur adjacent to the failure plane due to the ingress of water through a relatively permeable 1.5cm thick sand lens overlying the failure plane. Increased rate of infiltration along tension cracks and grabens also results in increased instability.

37 4.2.3 Landslide geometry and geomorphology

Definitions of landslide dimensions follow those utilized by the IAEG Commission on Landslides

(1990). A geomorphological map is included in the Appendix as Figure ALL Cross sections are

shown as Figures 4.2.6 through 4.2.9.

Figure 4.2.2. Landslide dimensions: Numbers refer to dimensions defined in Table 4.2.2. (IAEG Commission on Landslides (1990)).

Table 4.2.2. Landslide dimensions

Number Name Dimension (m) 1 Width of displaced mass, Wa 580

2 Width of surface of rupture, Wr 580 3 Length of displaced mass, Ld 240

4 Length of surface of rupture, Lr 160

5 Depth of displaced mass, Dd 24

6 Depth of surface of rupture, Dr 22 7 Total length, L 250

38 The riverbank at this location consists of steep, 40m high bluffs. The surface of the failing mass slopes upward at about 16° from the top of the bluff to a steep, 1 to 21m high main scarp. The elevation difference, H, between the toe of the displaced mass at river level and the top of the main scarp is about 80m to a maximum at the western lateral margin of 100m. The elevation difference between the crest of the main scarp and the toe of the main failure surface is 53m at the center of the slope, to a maximum of 80 m at the western portion of the slope.

The western lateral scarp is well defined and, based on photogrametry, does not appear to have changed since at least 1940. Retrogression has reached a maximum here with the back scarp

jogging to the north to reach elevation 280m (Lr=193m, La=275m). As evidenced in airphotos of

1952, shortly prior to that year, the western portion of the slide area, between the main back scarp and the bluff, appears to have moved by several meters. The ground surface of this disturbed portion of the slide is subject to continuing slumps and small mud flows. This area can be distinguished in the field and air photographs by the relatively immature deciduous trees and dense hydrophilic vegetation.

A continuous main scarp defines the upper limit of the landslide. It consists of steep, freshly exposed soil scarps that vary in height from 2 to 21m. The eastern boundary of the slide is poorly defined. It appears that recent movement has occurred as far east as 125m beyond Plantation

Creek as evidenced by tension cracks, distressed vegetation and ponded water. A north trending scarp located at Section 18 may define the eastern lateral margin of the slide, in which case the maximum width of the slide is 760m. If the eastern lateral margin is taken to be the boundary of the 1997 rapid earth flow the width of the landslide is 580m. The active slide area encompasses approximately 11.6 ha.

39 A reasonably detailed reconstruction of the configuration of the rupture surface and the geometry of the internal deformations (Figure 4.1.2) can be made by the study of the slide morphology.

This effort was aided considerably by the steep main scarp and an excellent exposure of the toe of the rupture surface near the top of the steep bluff face. The toe exposure suggests the main body of the slide appears to be moving forward towards the river over a distinct failure surface located within the laminated silts and clay of Unit 4 (approximate El. 178m). This sliding surface is knife- edged and can be traced for almost 100 m along the river escarpment (Fig. 4.2.3). It is an extremely smooth slickensided discontinuity dipping 1° to the west (Fig. 4.2.4). Within the steep gully walls of Central Creek the failure surface extends 40m into the slope.

The main scarp suggests the rupture surface dips at 55° through the glacial lake sediments, then curves along a intermediate planar element dipping 15° to the south to join the subhorizontal surface under the main part of the slide. The presence of this second segment of the rupture surface is evidenced by the relative lack of back-tilting and the distribution of internal shears in the upper part of the slide, which indicates that the slide is not very deep in that region.

The lack of rotation of the main body is further indicated by a horst and graben feature below the main scarp, upright tree cover, and a prominent horst observed at approximately El. 215 (Figure

4.2.5). Laminations within the failure prism of this ridge dip 5° into the slope. Similar prisms are evidenced in flowslides in sensitive clay (Geertsema and Schwab, 1997) where translational failure dominates. As well as reconstructing the rupture surface over which the slide mass moves, prominent scarps within the slide area were extended to intersect the basal rupture surface, dividing the slide mass into wedges.

40 Figure 4.2.3. Exposure of toe of rupture surface near top of steep bluff face.

There are relatively few constraints other than the slide morphology and bedding

inclinations, and other geometries may be possible. Figs. 4.2.6 through 4.2.9 suggest that

movement of intact, triangular shaped wedges may be responsible for the formation of grabens and prisms observed within the existing slope profile. Some backtilting may

have occurred as a result of internal deformation but the movement is predominantly translational with subsidence of an active wedge driving forward a passive distal block.

The displaced material above the sliding surface is remoulded to varying degrees whereas the material below remains intact and undisturbed. The estimated volume of displaced material isl .84 Mm3. In estimating this volume, it was assumed that the lateral scarps cut up steeply with the same slope as the back scarp.

41 Figure 4.2.4. Slickensided rupture surface. Figure 4.2.6. Section 1, Slesse Park landslide. Location shown on Figure Al. See oversize.

43 Figure 4.2.7. Section 2, Slesse Park landslide. Location shown on Figure Al. See oversize.

44 Figure 4.2.8. Section 3/4, Slesse Park landslide. Location shown on Figure Al. See oversize.

45 Figure 4.2.9. Section 5, Slesse Park landslide. Location shown on Figure Al. See oversize.

46 4.2.4 Soil properties

Classification tests including Atterberg Limits, natural moisture content, and grain size analysis were completed for the dominant units exposed by landslide movement. Emphasis was placed on classification and characterization of the glaciolacustrine sediments, being the unit principally involved in the slide movement and constituting 91% of the source volume. Results of natural moisture content and Atterberg limits from surface exposures are summarized in Table 4.2.3. The plasticity chart is shown in Figure 4.2.10.

47 Table 4.2.3. Summary of classification test results, Slesse Park landslide

Colluvium Property Glacio-lacustrine Diamicton n=15 n=l n=7 Natural water content (%) 28 12.9 21.5 {19.5-37.8} {13.2-26.8} Liquid limit (%) 47.7 22.5 28.1 {30-72} {23-33.8} Plastic limit (%) 23.5 15.7 18.8 {10.4-32} {16-20.9} Plasticity index (%) 24 6.8 9.3 {9.8-40} {6-13} Liquidity index (%) 0.18 -0.41 0.31 {-0.06-0.52} {-0.38-0.97} n=6

Clay content (%) 41.1 43.5 47.1 {26.3-65.7} n=l n=9 Silt content (%) 58.9 23.9 51.3 {34.1-73.6} n=l n=6 Sand content (%) 19.8 9.3 1.6 {7.7-31.9} n=l n=2 Activity 0.53 0.16 0.2 {0.17-0.83} n=l n=9 Bulk density (kg/m3) 1935 n=l n=number of samples tested

Bulk properties

The glacio-lacustrine sediments vary from inorganic, low to high plasticity clay. They are

interbedded with non plastic silt and fine sand laminations. The laminations result in a gradation in

plasticity and clay content vertically. This gradation is not apparent from the Atterberg limits and

grain size analysis due to the complete mixing associated with these tests.

The low liquidity indices of the intact clay (-0.58 - -0.03) suggest that the relatively undisturbed

landslide material is moderately to highly overconsolidated. Remoulded clay in the disturbed

parts of the exposures show liquidity indices higher than that of intact clay, probably as a result of

48 swelling. Material subject to earthflow is expected to have a liquidity index close to 1. However, at the time of sampling, the mudflow colluvium was drained and desiccated (LI = -0.38).

The stiffness of the clay is another parameter used to indicate that the clay is overconsolidated.

Below the shear plane, the laminated clay was impenetrable by both the shear vane and pocket penetrometer. Therefore, the unconfined compressive strength of the intact laminated silt and clay was estimated as exceeding 490kPa and is classified as hard.

Quantitative strength tests were conducted at Slesse Park on the remoulded clays near the edge of the gully walls in the flowbowl created by Central Creek. The shear strength of the clay was estimated by two different methods, a pocket penetrometer, and a shear vane. The following tables summarize these results:

Table 4.2.4. Unconfined compressive strength estimated from pocket penetrometer in Auger Holel. Depth (m) Unconfined Strength (kPa) 1 36.3 2 220.5 3 73.5 4 196.0 4.5 147.0 5 58.8 5.5 196.0 5.7 171.5 6 171.5-220.5 6 (undisturbed) >490

Table 4.2.5. Shear strength estimated from shear vane in Auger Hole 2 Depth (m) Peak Strength Residual Strength (kPa) (kPa) 1.3 52.0 10.4 1.4 41.6 10.4 2.0 62.4 10.4 2.67 62.4 20.8 2.75 (undisturbed) >208

49 There was no trend of increasing strength with depth observed in either of the auger holes.

Within the remoulded material the texture ranged from a soft, reworked clay to zones and lenticular shaped clasts of stiff clay. Using a pocket penetrometer, the estimate of the unconfined compressive strength may vary from 60 to 200 kPa within 2 cm in any direction depending on the degree of remoulding. The shear strength obtained from the shear vane is thought to be a better average.

The above results have to be interpreted with some caution. The laminations and sand inclusions

result in heterogeneity, which is partly destroyed by mixing during the tests. Also, some of the

near surface exposures have been affected by desiccation (hence improbably low LI values of

some remoulded samples. Nevertheless, the test results clearly show that the intact material is

heavily overconsolidated, that the bulk of the material is moderately to highly plastic and that the

sensitivity is moderate or low. The considerable variation of Liquidity Indices and shear strengths

of the remoulded material also reflects the presence of intact fragments, surrounded by a softer

matrix, typical of earth flow material (cf. Hutchinson, 1988).

Activities range from 0.17 to 0.83 with an average of 0.47, indicative of an inactive, kaolinite/illite

rich clay. An activity of 0.47 is typical of late-glacial clays derived largely by mechanical erosion

of non-argillaceous rocks by ice-sheets, and deposited in ice-dammed lakes (Skempton, 1952).

Material adjacent to the main sliding surface

The clay layer that contains the slickensided failure surface is uniform, light grey and

approximately 1cm thick. Of the materials tested, this clay shows the highest plasticity (PI=36%)

and the highest clay content (66%). Material sampled within 4cm below the slickensided surface has a lower plasticity (PI=17%) and clay content (26%). A fine sand 1.5cm thick overlies the failure surface. Material sampled from 0-4 cm above the sand (2-6cm above the failure surface) was remoulded. The plasticity of this material was between that of the undisturbed material

below the sheared clay surface and the clay material in which the main failure occurs.

Table 4.2.6. Summary of classification test results adjacent to the main sliding surface LI Location CC LL PI (%) (%) (%) (%) Clay from main weak surface 66 66 36 0.1 2-6cm above weak surface, remoulded clay 37.4 57 31 0.23 0-4cm below weak surface, laminated silt and clay 26 37 17 0.76

The mineralogy of the clay adjacent to the sliding surface was evaluated by X-ray diffraction. The

results suggest that the clay fraction (<2u,m) is primarily illite and kaolinite (ratio ~ 5:1) and

glacially ground rock flour (quartz and feldspar). Vermiculite and smectite are also present,

though no quantification was undertaken.

4.3 Failure behaviour

4.3.1 Main instability

The main scarp and the major deformational features of the slide are evident on the oldest

airphoto (1940 - B.C. 209:59,60). Measured displacements, fresh soil exposures, thrown trees,

and stressed vegetation including stretched roots are evidence to indicate recent to very recent

movement in the main scarp area over a distance of more than 500m. However, the ground

surface between the freshly disturbed back scarp and the slide toe is forested and shows relatively

few signs of fresh disturbance, suggesting that the main body of the slide is moving forward

largely as rigid blocks with limited internal deformation.

51 Measured displacements

The main body of the slide appears to move episodically in response to climatic conditions, especially snow melt and/or rain on snow events and associated groundwater conditions. Slope movements are numerically indicated by measuring distances between trees marked with metal reference markers. Stations are marked on Figure Al. These extensometers, installed in July

1997, have since recorded displacements of maximum 2.6m. Most of this movement, at a rate

approximately 0.4m/month, occurred during the winter of 1998-1999. Displacement values are

included as Table AI. 1 of the appendix.

Maximum displacements have occurred between the extensometers located directly above and

below the main scarp. Within the main body of the slide, displacements between the

extensometers do not exceed 0.8m. Displacement appears to be greater within the eastern half of

the slide, as quantified by the extensometers, and evidenced by fresh displacement along the main

scarp and numerous fresh tension cracks. Station W7 was destroyed some time between Dec.

1998 and March 1999 when the tree containing the survey tag fell as a consequence of a tension

crack opening within the root system.

Following the 1991 rapid failure a system of wire extensometers were installed to monitor slope

movement for one week during construction of a protective berm at the base of the slope. From

March 12-19, 1991 13.4cm were recorded within the main body of slide and 2.2 cm across the

main scarp.

52 Landslide progression

The type of failure exhibited by the slide in the past and the landslide progression is based on airphoto interpretation aided by the technique of photogrammetry. Refer to Figures AI. 1 to AI.4 of the appendix for the results of photogrametric mapping.

The oldest airphotos available, from 1940, show a well developed back and western lateral scarp in approximately the same location as they are today. It may be concluded that failure of the main body of the slide occurred prior to that time. At that time the unstable area extended further west. A lobe of colluvium appeared to have blocked a channel of the river, resulting in the river shifting to the south. The colluvium is vegetated, suggesting that the failure occurred several years prior to 1940.

Shortly prior to 1952, the western portion of the slide area, between the main back scarp and the

bluff, moved by several meters, in what appears like a large earth flow tongue. The disturbed

surface of this feature, devoid of standing trees, covers nearly one-third of the slide area. Its

frontal portion appeared to have detached and flowed over the bluff, and reached the river

channel, but no major blockage can be seen on the photos.

A linear ridge feature that parallels the main scarp 40m downslope in the vicinity of Central Creek

can be observed in airphotos from 1979. This ridge appears to be approximately 150m long.

This feature can be observed in the field as a horst that indicates translational spreading (Figure

4.2.5).

53 54 Air photographs taken in January, 1997 document the earthflows which occurred in the colluvial apron at the base of the bluffs and resulted in partial blocking of the river. By 1997, a large flow bowl had developed at Central Creek, extending as far back as to intersect the horst feature described in photos from 1979. There is continued shallow sliding and slumping in the disturbed

1952 landslide area. Figure 4.3.2 shows the principal morphological features of the 1997 airphoto.

It was intended that photogrammetry might be a useful investigation tool to determine the magnitude of the "1952" failure described above. From transects taken through the 1952 landslide area, a profile of the pre (1940) and post (1952) failure ground surface could be

interpreted. From this, one may be able to determine whether the landslide was a deep seated

failure exploiting the main failure surface, or a series of more shallow slumps and flows.

However, due to the poor quality of the air photos from 1940 and 1952, distortion of the images

could not be eliminated by the stereoplotter and the transects were considered unreliable.

However, it may be concluded that retrogression of the main scarp has not occurred since at least

1940.

4.3.2 Secondary movements

The main instability at this site is a ductile compound sliding failure. The gradual forward

displacement of the main mass causes three types of very rapid secondary movements of limited

magnitude.

The first type, block falls and minor slumps along the crest, occur periodically each year and are

probably more frequent following a period of displacement in the main mass. Seepage erosion by

55 piping within the sand unit approximately 10m below the main weak surface undermines blocks of fractured silt and clay, increasing the frequency of failure. On March 3, 1991, an earthflow deposited 15 000 to 20 000 m3 of debris into the river channel. The debris appears to have detached from the crest of the oversteepened bluffs in the vicinity of Central Creek (Figure 4.3.3).

Subsequently, a protective berm was constructed to reduce erosion along the toe of the slide mass.

The debris from the secondary slides currently accumulates at the foot of the toe scarp, forming a

steep colluvial apron. The apron can build up to the full height of the bluff and become colonized

by vegetation. It consists of a mixture of blocks of clay and silt, woody debris, set in a matrix of

softened remoulded clay. The second secondary movement type was illustrated by the events of

January 1997, following a shift of the main mass. Blocks slumped from the toe of the slide mass

and thrust into the top of the colluvial apron. A large undrained failure occurred, moving much of

the apron across the dyke and briefly damming the river channel (Figure 4.3.4). The volume of

this slide was 50 000 m3. Its total displacement of 30 m resulted in a dam across the 20m wide

river channel which was quickly removed by erosion.

A smaller failure of similar type took place a few hours after the initial event. This landslide

involved a volume of approximately 4,000 m3 of colluvium. The total displacement of the slide

front was approximately 30m and the peak velocity was of the order of 3m/sec.

The third type of secondary movement is the flow- like displacement on the surface of the main

mass, which was observed on the 1952 airphotos. The scar is mapped on an overlay of the 1997

airphoto, Figure 4.3.2. The dynamic behaviour of this movement is unknown. Although it

56 Figure 4.3.3. Earthflow of 1991, Slesse Park landslide. covered the entire length of the main slide from main scarp to toe, it does not appear to have lowered the surface of the slope by more than a few metres and therefore suggests a shallow failure. There is also no evidence of it having delivered large quantities of debris to the floodplain. This evidence indicates that the movement may have been similar to an earth flow surge (e.g. Hutchinson, 1988) involving primarily remolded material and probably not faster than rapid (3m/sec).

4.3.3 Trigger

The initial cause for instability was likely the result of undercutting of the toe of the slope by river

erosion. However, as the base of the rupture surface is 30m above river level, it may be

concluded that river erosion does not directly trigger present slope movement.

Landslide movement is triggered by rainfall and/or snowmelt. As has been shown by studies by

Iverson and Major (1987) and (Church and Miles, 1987) the timing, duration, and speed of slope

movement does not often correlate directly with the timing and amount of rainfall.

A plot of the 1-30 day analysis of intensity versus duration of rainfall indicates that the antecedent

rainfall leading up to the January 23, 1997 rapid failure was not exceptional, having a return

period less than 2 years (Figure 4.3.5). Heavy rainfall occurred on January 18, three days prior to

the rapid failure, with accumulations of 49.5 mm. The 1-30 day analysis prior to this storm is also

plotted in Figure 4.3.5. The three day intensity has a return period of 2 years. As shown, the

cumulative precipitation during the month of January reached the maximum recorded at the

Chilliwack River hatchery over the period of record 1984-1999. Furthermore, November and

December, 1996 received maximum snowfall accumulations for the 15 year period of record

58 (Figure 4.3.7). Of this, a significant depth remained on the ground and was noted at the time of the January 1997 failure.

A similar pattern was observed prior to the March 3, 1991 failure. Though on the day of the

failure 27.4 mm of rain accumulated, the preceding 12 days had been relatively dry. Monthly

rainfall over the winter of 1990/91 (Oct. - March) was greater than average, November being the

wettest month on record since 1962 with accumulations of 497mm, 242% above the monthly

Figure 4.3.5. Intensity-frequency-duration plot, Chilliwack River valley.

100

Return period (years)

TJ E E,

in a cu +—

Duration (days)

59 Figure 4.3.6. Chilliwack River Valley, Monthly precipitation (period of record 1984-1999).

E Total EMean B Max

Feb-96 Mar-96 Apr-96 May-96 Jun-96 Jul-96 Aug-96 Sep-96 Oct-96 Nov-96 Dec-96 Jan-97

Month

3.7. Chilliwack River Valley, Monthly snowfall (period of record 1984-1999).

• Total EJMean • Max

Feb-96 Mar-96 Apr-96 May-96 Jun-96 Jul-96 Aug-96 Sep-96 Oct-96 Nov-96 Dec-96 Jan-97

Month

60 average. Rainfall accumulation for the month of February was also near maximum recorded , exceeding the monthly average by 182 % though inspection of the intensity-duration-frequency curve indicates that this 30 day duration was not exceptional. However, the maximum snowfall recorded since 1982 at the Chilliwack River hatchery was nearly 200% of the monthly average.

In summary, the January 1997 event occurred during heavy rain with snow on the ground. The winter of 1996-1997 was the wettest on record in the Vancouver area (Environment Canada).

November and December had seen unusually high snowfalls, with heavy rainfall accumulating during January. A similar pattern was observed prior to the March 3, 1991 failure. January 1991 saw moderately heavy snowfall, and though rainfall in February was not exceptional, accumulations were greater than the mean.

4.3.4 Landslide dam

On several occasions debris from minor landslides have caused partial and temporary blockage of the Chilliwack River. Historically, failure in similar material at an upstream location caused temporary river blockage (Thomson, 1999). The largest of the January 1997 rapid mass movements (50 000m3) was sufficiently mobile to cause temporary damming of the river (Figure

4.3.4). The landslide dam was breached quickly, without significant ponding or damaging

consequences. Continued failures of the same order of magnitude are considered likely.

Due to the potential for a more lasting blockage causing a flood surge upon release or

overtopping of the landslide dam, a report was prepared by Northwest Hydraulic Consultants Ltd.

(1997). They concluded that some type of bypass or diversion channel through the inside of the

61 large meander loop might mitigate a potential flood surge. The ground surface through the inside

of the meander loop is approximately 4 m above river level.

4.4 Slope stability analysis

The main slide block at Slesse Park evidently exists at the state of limit equilibrium. The objective

was to conduct a back-analysis of the stability of the existing slope to evaluate the available soil

strengths and the pore pressure distribution just prior to reactivation of the slide. The limit

equilibrium stability analysis is conducted using the program CLARA (Hungr et al, 1989) which is

capable of calculating three-dimensional stability. The configuration of the sliding surface is

shown in Figure 4.4.1 (b).

The analyses requires a reasonably accurate reconstruction of the rupture surface and conditions

existing at the time of failure. There are a number of assumptions inherent in the choice of the

phreatic surface and range of material strengths. Therefore, it is not possible, nor is it the intent,

to accurately model the conditions of the slope and obtain specific values. Rather, the analyses

are used to test assumptions regarding these parameters and to obtain reasonable estimates.

4.4.1 Selection of strength parameters

The soil properties initially used in the stability analysis are summarized in Table 4.4.1. The

materials are numbered from the bottom up as is the convention in the stability analysis program.

Given the heterogeneity of the materials, it is reasonable to assume different shear strength

parameters act along the rupture surface. For simplicity however, the strength is assumed to be

uniform along a given segment.

62 Figure 4.4.1 Configuration of (a) ground surface and (b) sliding surface used in limit equilibrium analysis. Given that large displacements of the main unstable mass have occurred, it is assumed that the weak surface is at its residual strength. The composition of the highly plastic clays that the weak surface exploits was discussed in Section 4.2.4. Skempton (1985) determined that the angles of residual shearing resistance of the three most commonly occurring clay minerals are; approximately 15° for kaolinite, 10° for illite or clay mica and 5° for montmorillonite. The presence of smectite may be responsible for a relatively low residual strength, whereas quartz and

feldspar have high values of friction angle of 14.5° was assigned to the weak surface parallel to bedding. This value is within the range suggested by Mesri and Capeda-

Diaz (1986) and Stark and Eid (1994) for a liquid limit of 66%.

While the weak, subhorizontal surface is believed to exploit a lamination of high plasticity clay, the segments of the failure surface dipping upslope at approximately 15° and 55° presumably cut

across clay, silt, and sand laminations. Therefore, a higher 'r of approximately 17° was assumed.

Table 4.4.1. Summary of soil properties used in stability analysis

Material number Soil Unit weight, y Friction angle, <|>'r (kN/m3) (degrees) Clay along bedding - 14.5 1 Clay and Silt across bedding 19 17 2 Sand and colluvium 18 33

The sand is described as a dense, angular, uniform sand. The colluvium ranges from very stiff

diamicton, with a clay fraction of 43% to a sandy gravel. The strength of this heterogeneous

deposit has been assigned an estimated value of

units to the shear resistance is negligible as the combined thickness constitutes only 9% of the

total. 4.4.2 Sensitivity to shear strength

The stability of the slope is very sensitive to the available shear resistance along the rupture surface as illustrated in Figures 4.4.2 and 4.4.3. The range of values presented is considered within reasonable limits based on existing correlations between liquid limit, composition and residual friction angles (Stark and Eid, 1994; Mesri and Capeda-Diaz, 1986). Back analysis also aids in limiting these values as too low a friction angle may require an unrealistically low piezometric surface and vice versa. A variation in the specific values of + 3° cannot be ruled out and the selected values should be considered an estimate only.

4.4.3 Sensitivity to pore pressure distribution

There is considerable uncertainty regarding the ground water flow regime and it should be noted

that the use of a single piezometric surface is an over simplification. For the material strengths

summarized in Table 4.4.1, a depth to piezometric surface of 13.5 m was required to maintain

stability. The equivalent pore pressure ratio required to maintain stability is ru = 0.19. A

piezometric surface 3 m below the ground surface, as measured in the piezometer located

adjacent to Central Creek, requires unrealistically high strengths to maintain stability. This

suggests the water encountered at this depth was a perched water table and that the piezometric

surface is lower than 3 m. Figure 4.4.4 illustrates the sensitivity of the factor of safety to changes

in the depth to the water table. The sensitivity of the stability of the slope to changes in the

piezometric surface in the slope supports the hypothesis that filling of cracks with water and

seasonal rise of groundwater levels may contribute to the trigger for slope movement.

65 Figure 4.4.2. Sensitivity of factor of safety to variations of the residual shear strength along the basal rupture surface.

1.10

^ 1.05

"5 © 1.00 1- o selecte d for stability analysis o ca I* 0.95

0.90 16 12 13 14 15 Residual friction angle (degrees)

Figure 4.4.3. Sensitivity of factor of factor of safety to variations of the residual shear strength along the cross-bedding sliding surface.

1.1

1.05

•a o u © Dilitv analysis "5 selected for sta 03 0.95

19 15 16 17 18 Residual friction angle (degrees)

66 Despite the variability of these parameters, the occurrence, continuation, and character of the instability is explained using a conventional failure mechanism and analysis, typical for a large

landslide in overconsolidated clay.

5.0 Attachie landslide

5.1 Regional Setting

5.1.1 Physiography

Attachie, located at the southwest of the Charlie Lake area, lies at the western edge of the Alberta

Plateau subprovince of the Great Plains (Bostock, 1948; Holland, 1964). The area is underlain by

gently dipping Cretaceous sandstones and shales which contribute to a subdued topography

including northwesterly trending belts of hilly and lowland terrain.

67 Figure 5.1.1. Attachie landslide (photo courtesy of S.G. Evans).

The platform extending up from the rims of the trenches to the base of the uplands has a low, continuous slope of about lA° to 2 °, although it bears a significant microrelief of mounds and swampy hollows (Mathews, 1978).

Peace River trench has been cut into the flat to gently undulating surface of the Interior Plains in postglacial time. The trench follows for the most part an interglacial valley of the Peace which contains a dissected, complex Pleistocene fill. Where the postglacial channel follows the interglacial valley, the floor of the present trench is cut into the bedrock floor of the interglacial

68 valley and is therefore lower. The toe and the basal shear plane of landslides in the Pleistocene sediments, therefore, lie above present river level. In the vicinity of Attachie , the trench averages

205 m in depth and varies between 2.5 and 3.0 km in width, from rim to rim (Evans et al. 1996).

Landslides have occurred in both the Cretaceous shales (e.g. Hardy, 1963) and the fine grained

Pleistocene sediments of the region. "The vast majority have developed in the lake deposits of the

last interglacial interval and incorporate the overlying till and late-glacial lacustrine beds. The

basal part of the lake deposits, close to underlying buried gravels, is a common locus for sliding.

The underlying gravel is generally well drained, and uplift pressures do not seem to initiate sliding.

Fissuring of the overlying beds following initial movement at depth, however, may permit ingress

of snowmelt and rainwater and thus may contribute to additional and perhaps much larger scale

movements" (Mathews, 1978).

5.1.2 Climate

The Peace River region of British Columbia is part of the Boreal White and Black Spruce

biogeoclimatic zone (B.C. Ministry of Forests, 1988). Rolling uplands and prairies allow Arctic

air to flow in unobstructed, resulting in cold winters. January temperatures average -19°C to -

11°C. Temperatures in summer tend to be warm, 10°C - 22°C, with humid air and occasional

light rain accompanied by frequent, intense thundershowers.

Annual precipitation in the region is unlike much of British Columbia, with peak monthly

precipitation in July and drier conditions in the winter months. Monthly and annual average

69 temperature and precipitation data is attached as Appendix II. The area is categorized as subhumid (Drinkwater et al., 1969) with an average annual precipitation of 466 mm.

5.1.3 Bedrock Geology

The study area is underlain by Cretaceous shales and sandstones, which in general dip gently to the northeast and east, interrupted in the west by a few flexures and local faults (Mathews, 1978).

The Dunvegan Formation, of Upper Cretaceous age, is widespread through northeastern British

Columbia with varying lithology from sandstones to carbonaceous shales and coal (Stott, 1982).

5.1.4 Quaternary history

The Laurentide Ice Sheet reached the Charlie Lake area, which covers the Attachie, on at least

three separate occasions during the Quaternary. Ice from the mountains to the west also reached

the area, with the last major mountain glaciation being essentially contemporaneous with the

Laurentide Glaciation. (Mathews, 1978)

The last interglacial interval is marked by the erosion and development of the Peace River trench,

a 6.5 to 8 km wide trench with its bedrock floor about 30m above the present river. In the

location of Attachie, the floor of this trench is mantled by a 15 to 18m deep layer of alluvial

gravels thought to have deposited under mildly aggrading conditions.

The erosional interval was terminated by an advance of Laurentide ice from the east which

dammed the Peace River. The subsequent lacustrine environment is believed to have existed

between 43 500 and 38 000 years B.P. during which time a lacustrine succession covered the

gravel-floored trenches. The underlying gravels, may therefore be considered as dating back

70 more than 43 500 radiocarbon years but probably are still mid-Wisconsinan in age. "The local occurrence of ice-rafted pebbles within silt high in the succession suggests the proximity of glacial ice. Accordingly, the fine grained unit is considered to mark the change from completely nonglacial conditions at the time of gravel deposition, to proglacial conditions heralding the latest advance of ice from the east" (Mathews, 1978).

The ice sheet continued to advance from the northeast eventually covering the Charlie Lake area,

depositing till over the lacustrine sediments. Till is considered to be of late Wisconsin age.

During the retreating stages of the last Laurentide ice sheet, obstruction of the regional easterly to

northeasterly drainage of this part of the Great Plains created a series of ice-dammed lakes within

which sediments accumulated. Late-glacial lake sediments date between 13 500 and 9960 BP

(Mathews, 1978).

5.1.5 Surficial Geology

The present Peace River trench in most places lies within the limits of an interglacial valley of the

Peace buried by a thick succession of Pleistocene sediments. The present river cut its postglacial

trench through the unconsolidated fill into the bedrock floor of the interglacial valley. The floor

of the buried Peace River trench is therefore about 30 m above present river level and slopes

easterly with about the same gradient as that of the present river (Mathews, 1978).

Figure 5.1.2 summarizes the stratigraphic succession at the Attachie landslide. The lowest unit of

the Pleistocene succession is a 23 to 30 m thick layer of clean, well sorted fluvial or glacio-fluvial

71 gravel (Unit 1). The top of the gravel unit is sharp and is located typically about 64 m above present river level. The gradient of this contact is nearly identical to that of the river over a length of 90km within the Charlie Lake map area (Mathews, 1978).

Figure 5.1.2. Cross-section of the Attachie landslide showing stratigraphy and the 1973 sliding surface. Pre-slide topography is prepared from 1967 aerial photographs (BC 5267;63,64)

0 ' 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 meters

Overlying the gravel are fine grained glacio-lacustrine sediments (Unit 2). The contact is sharp

and located at an elevation between 492 m and 499 m. The sequence fills the interglacial Peace

River Trench to between el. 603 and 612 m which gives a maximum thickness of 120 m. Within

the sequence silt generally dominates, but sand may occur at any level in units up to 30 m thick

(Unit 3). The glacio-lacustrine sequence is capped by till (Unit 4) of average thickness 12 m.

Late-glacial lacustrine deposits overlie the till (Unit 6). The thickness of these lake sediments

over the till substratum show marked variations from 0 to a mean of 30 m.

72 5.2 Site description

5.2.1 Detailed stratigraphy

The stratigraphy at the site is based on 8 boreholes drilled between 1979 and 1980, from field observations, and from correlation to the regional stratigraphy. The stratigraphy at Attachie follows the regional succession described by Mathews (1978) and summarized in Section 5.1.5 and Figure 5.1.2. Five holes were drilled in the upstream slide and three holes in the downstream slide. Refer to Figure 5.2.4 for locations. Borehole logs are attached as an Appendix II.

The channel of the Peace River is cut into Cretaceous shale which also forms the interglacial valley side that rises upslope beneath the Pleistocene sediments. The shale is horizontally bedded, with a surface sloping towards the river at angles of 0° to 15°.

The base of the Pleistocene succession is exposed in the Attachie slopes at approximately El. 475

m, 40 m above the present level of Peace River. Unit 1 is composed of fluvial, or glaciofluvial,

clean, coarse grained, well sorted gravel with minor sand. Pebbles and cobbles are similar in size

distribution (2.5 to 15 cm) and provenance to those of the present Peace River (Mathews, 1978).

Quartzite clasts dominate; well cemented sandstones, grey granitic rocks, and greenish to pink

volcanic rocks are common; red granite or gneiss from the Canadian Shield is very rare. The

average thickness of the gravel unit is 14 m at the base of the slope, thinning to 0-7 m above the

crest of the slope.

Gravels are overlain conformably by a sequence of interstratified silt and clay (Unit 2) at elevation

492-499m. Within the sequence silt dominates with laminations of clay and sand partings. Very

73 few of the fine sediments sampled had clay fractions exceeding 50%. Laminations are on a scale of 1-5 mm. Black, organic laminations 2-6 mm thick occur at 5 - 20 cm intervals. The clays are stiff to hard, well-consolidated and can be highly plastic and brittle. Clays part easily along fine sandy lamination. A maximum thickness of glacio-lacustrine sediments was encountered in borehole 63-2, where 82 m was penetrated. Two zones of fine grained, micaceous sand (Unit 3) were encountered between elevation 555 to 559 m and 519 to 533 m, constituting 22% of the fine sediments in the stratigraphic column. A 9 m thick zone of sand was encountered in borehole 63-

5 comprising 20% of the 46 m of fine sediments. A 6 m zone of sand was encountered in borehole

63-3 m, 110 m upslope from 63-5. It is not known whether this unit is continuous between these borehole locations.

"Surface exposures of the lacustrine sediments immediately downstream of the 1973 Attachie slide consist of interlayered fine clayey sands, laminated silts, and silty clays. Beds and

laminations in the exposures are frequently backtilted due to slumping. The backtilting is

accompanied by complex normal faulting which offsets laminations up to 25 cm within a block.

In addition, many clay bands within the sequence are slickensided parallel to bedding in a

downslope direction. Shelby tube samples exhibit local shearing and brecciation, softened

contacts, backtilted varves, and bedding distorted by folding. This deformation has been

interpreted to be predominantly the result of glaciotectonism and pre- and post-glacial slumping"

(Evans et al., 1996).

A prominent layer of very dense, low to zero plasticity laminated silt can be observed in the side

scarp between Elevations 530-542m. Combining this layer with other less extensive interbeds, it

74 is estimated that the glacio-lacustrine sequence contained approximately 31% of low-plasticity or non-plastic silt, in addition to 48% of clayey silt of moderate to high plasticity and 21% sand.

The glacio-lacustrine units are overlain by glacial till (Unit 4) deposited by Late Wisconsinan ice.

Till 23 to 38 m thick was encountered above the crest of the slide. The till is massive, clay rich, and characterized by scattered well rounded pebbles and boulders. Stiffness ranges from very stiff to hard. The till exhibits vertical jointing, and stands in steep slopes. This unit has an important role in the stability of the Attachie slopes since it forms a resistant cap to the Pleistocene succession (Evans et al., 1996).

Varved, normally consolidated silts and clays overlie the till at the top of the Attachie slopes. The thickness varied between 0 - 27 m. These late-glacial lacustrine sediments are characterized by

clay with scattered pebbles.

Slide debris encountered on the lower slopes and damming the river is a mixture of gravel, sand,

silt, and clay with some organic debris. It is generally grey-brown, moist, with slight to moderate

plasticity. Photos taken shortly after the 1973 flow slide show that the debris is surprisingly

blocky (Figure 5.1.3) and made up largely of boulder-sized angular blocks of clayey silt and till,

supported by a remoulded matrix. Many of the blocks had straight planar sides forming shapes

of cubes and pyramids, which suggests separation along joints.

75 Figure 5.1.3. Photograph of slide debris deposited at valley bottom taken within days of the 1973 event. Helicopter in top left of photograph provides scale.

5.2.2 Ground Water Conditions

Due to widely contrasting hydraulic conductivities of the multilayered units at Attachie, the groundwater flow regime is complex. "The basal gravel layer, though more permeable than other surficial units, does not have a good connection to any recharge area beyond the slope crest. Its main effect is therefore to drain the slide area toward the valley, rather than to conduct

groundwater to it." (Thurber, 1982). Furthermore, the basal gravels "are separated from the

overlying glacio-lacustrine clay and silt by a cemented sandstone layer, less than 10cm thick, that

forms an aquitard, preventing water flowing from laminated sediments to the basal gravels."

(Evans et al., 1996). The hydraulic conductivity varies considerably within the lacustrine

sediments of Units 3 to 5. Low permeability clays are interlayered with relatively high

76 permeability fine sands. Intermediate are the silt dominated layers. Seepage horizons and perched water tables are expected to exist.

Six of a total of eight holes were instrumented with a total of 12 standpipe piezometers and two pneumatics between July 1977 and July 1980. Monitoring of instrumentation was discontinued in

August, 1982 and resumed briefly in 1989. By 1989, all of the piezometers were either destroyed, or dry. Figures 5.2.5 to 5.2.7 indicate the locations and average recorded water levels

of the piezometers.

The following is a summary of their performance prior to destruction:

- 11 piezometers were dry or nearly so.

- 3 piezometers yielded a positive water level reading (63-2 P2, 63-6 P2, 63-8 PI).

The piezometers indicate that the groundwater levels are low. Positive water levels are

interpreted as perched water tables. Piezometric levels within glacially overridden strata fluctuate

between El. 537 at the base of the slope (63-2 P2) and 595 m.a.s.l. above the headscarp (63-8

PI). 63-6 P2 is interpreted to be a perched water table within the slide debris at the bottom of the

slope.

Although these field measurements are believed to adequately reflect the existing groundwater

conditions, it should be noted that more extreme temporary water conditions are possible.

Thurber (1982) made the following observations of groundwater conditions existing at the site:

77 "Immediately following the slide which occurred May, 1973, water was observed to be

abundant in the slide area including springs in the basal gravel beneath the slide. Much of the

disturbed silts were in a saturated state such that it was difficult to walk upon them."

A further example of rapidly rising piezometer levels was observed in 1975 in a slide area on the north bank near Taylor in similar materials (Maber and Stewart, 1976). In late spring a near record rainfall occurred and 5 days afterwards the piezometric levels in the basal gravel rose rapidly within 24 hours such that in one case a water level above ground level was recorded. The

event was followed by a ground movement of about 1.5 m which severed the piezometers. The

piezometers had been installed in late summer of the previous year and at that time had reportedly

encountered no or very low water levels (Thurber, 1982).

Thus the Attachie Slide area could be subject to temporary rapid increases in piezometric level

caused by a groundwater surge which may last for only a few days.

Exceptionally high piezometric pressure in the pore water of the gravel or the silt at the base of

the varved clay may have triggered movement of the slide, but would not be sufficient to explain

the disintegration and flow of the slide debris that took place within minutes.

78 Figure 5.2.1. 1970 Aerial photograph (BC7279;70) showing major and minor scarps in disturbed, pre-1973 failure area. Figure 5.2.2. 1973 Aerial photograph (BC5529;75) taken 40 hours after failure. Figure 5.2.3. Overlay of 1973 aerial photograph (BC5529;75) showing principal geomorphological features. Figure 5.2.4. Attachie landslide plan (original drawing B.C. Hydro 1016-C14-D1214).

Legend J Survey monument with apparent displacement direction (no magnitude implied) I Wireline extensometer • Borehole •" Tension crack

, , c _ SCALE: 100 0 100 201 -I—L ScarP 1:10 000 U i-i U I T

82 5.2.3 Landslide geometry and geomorphology

The toe of slide daylights approximately 61m above river level between El. 496 and 500 m. The ground surface from the base of the slide slopes with an average gradient of 15° to the top of the slope at El. 650 m. A 560 m long continuous main scarp defines the upper limit of the 1973 failure (Figure 5.2.4). It consists of a near vertical scarp that varies in height from 15 to 35m. At the center of the slope the elevation difference, H, between the toe of the main failure surface and the crest of the backscarp is 162 m. The pre-failure ground surface sloped at an average of 17°.

As a result of failure the top of the slope retrogressed 61-76 m and the average gradient was

reduced to 15°.

Movement of the slope had diffuse boundaries both upstream and downstream due to the

instability of neighbouring slopes. The lateral margins have been estimated from the topography

and mapping of airphotos and photographs taken shortly after the slide occurred (Figure 5.2.2).

Based on these estimates, the segment of the slope that failed was 760 m wide.

Table 5.2.1. Attachie landslide dimensions. Numbers refer to Figure 4.2.2. Number Name Dimension (m) 1 Width of displaced mass, Wa 760

2 Width of surface of rupture, Wr 760 3 Length of displaced mass, La 1495

4 Length of surface of rupture, Lr 522 5 Depth of displaced mass, Da 35

6 Depth of surface of rupture, Dr 75 7 Total length, L 1537

83 Figure 5.2.5. Attachie landslide Section 1. See Figure 5.2.4. for location. See oversize. Figure 5.2.6. Attachie landslide Section 2. See Figure 5.2.4. for location. See oversize. Figure 5.2.7. Attachie landslide Section 3. See Figure 5.2.4. for location. See oversize. Figure 5.2.8. Exposure of toe of rupture surface in clay overlying basal gravels.

87 Borehole logs, pinched piezometer tubes and observations in the field have been used to define the location of the failure surface (Figures 5.2.5 to 5.2.7). The failure surface daylights at the base of the slope between El. 496 and 500 m as a sharply defined, slickensided plane located in a plastic clay at the base of the fine grained glacially overridden sequence 30-45 cm above the contact with the underlying gravels (Figure 5.2.8). The basal gravels are undisturbed though some amount of gravel is incorporated into the lower 10 or 15 cm of the clay. Mathews' report

(1978) on the regional quaternary stratigraphy describes this contact as a common locus for

sliding.

Observations of borehole logging indicate that there are multiple failure planes and that which was

exploited by the upper stage of the 1973 failure was relatively shallow. For example, in borehole

63-5, the laminations go from 15° dip to horizontal at a depth of 35m below surface though no

evidence of shearing at this elevation was detected. There is however, shearing and folding 55m

below the surface, at the contact between the clay and underlying gravels.

The 1973 slide area is separated by a mid-slope scarp with a distinct, but somewhat discontinuous

topographic expression on the slope surface, best identifiable on airphotos (Figures 5.2.2 and

5.2.3). This appears to extend beyond both lateral margins of the slide area. Subsurface

observations indicate that there is a corresponding step in the rupture surface (Figures 5.2.5 to

5.2.7). Most material depletion during the 1973 slide (up to 50m vertically) occurred below the

scarp, in the lower part of the slope ("lower stage"). A greater degree of disintegration of the

debris material also generally occurred here. One exception is a horizontally lying raft of forested

ground, 350 by 110 m in area, situated immediately above the toe scarp and underlain by only

88 about 20-3 Om thickness of debris. The location where this raft rests has been depleted by at least

10m vertically indicating that either the raft slid into its present location after much of the underlying soil left the slope or a soft layer was squeezed out downslope from beneath it. The original volume of this zone was primarily made of glacio-lacustrine sediments dominated by silt.

The upper part of the slope also had a bi-linear profile, but shallower and composed of a large

proportion of till ("upper stage"). As shown by drilling, this part of the landslide did not

completely disintegrate, but preserves large intact blocks back-tilted up to 15°. It also contained

shear planes situated beneath the rupture surface. A series of sharp ridges and a narrow,

discontinuous graben formed in till at the foot of the main scarp (Figure 5.2.9).

The total volume of the mass displaced by the 1973 flow slide is estimated to be 12.4 Mm3.

About 6.0 Mm3 remained on the slope while the other 6.4 Mm3 vacated the slope, descended the

50 m shale scarp and travelled 900 m across the flat floodplain of the Peace River. The deposit in

the river was uniformly spread out, covering an area of 83.5 ha with an average depth of 7.6m

(Evans et al., 1996). The total length of the landslide path, from the main scarp to the distal limit

of the deposit is 1.5 km with a fahrboschung of 7.7°.

The following field observations were contained in a report by Thurber (1981), compiled 8 years

after the occurrence of the landslide.

"The slide debris exhibits a unique manner of disintegration. The large-scale stratigraphic

make-up of the soil deposits is completely broken both in the material in the river and in that

89 which remained on the slope. No very large intact blocks have been left, except in the

"unfailed" material at the lateral margins. The character of the debris on the scar and on the valley bottom appears quite uniform, consisting mainly of boulder-size blocks of intact soil surrounded by a smooth, once probably relatively fluid, matrix. These intact blocks are conspicuous everywhere and give the deposit an appearance similar to that of rock fall debris.

Many of the blocks have well defined angular shapes with multiple perfect planar surfaces and sharp edges forming pyramids and cubes which suggests separation of the soil along pre• existing joints. The fact that these shapes could have been preserved for 8 years since the occurrence of the landslide suggests high degree of compaction of the parent material. Most of the blocks observed are formed of silty clay till with a significant proportion of gravel particles as well as large blocks of varved glacially-overridden sediments, some exhibiting heavily sheared fabric.

The debris surface between blocks and accumulations of blocks is quite smooth and contains flow features such as minor lobes. Its general character is similar to that of the earth flow

debris described in the previous paragraph. This matrix must have been relatively fluid during

the slide, judging from its appearance and from the overall smoothness of the deposit's

topography. Intact till blocks appear to have been rafted in this soft material and thus escaped

complete disintegration during the rapid displacement of over 1 km. Numerous small ponds

and wet soft areas on the deposit show that the interstices between the intact blocks are

completely filled by the matrix.

90 The proportion of intact blocks and remoulded material on the surface does not appear to

vary significantly along the fall line from the base of the main scarp to the toe of the landslide.

This is what mainly prompts the use of the term "flow slide" for this event.

Vertical and inclined open joints have been found in the till at several locations in the nearly

vertical head scarp approximately 20 m beneath the crest of the slope. Large gypsum crystals

(up to about 2 cm length) have also been found at two locations in the upper part of the scar.

These probably grew in open cracks.

The bulk of the debris on the slope appears to originate from the till stratum, which is the

second topmost member of the stratigraphic sequence of the main scarp. The material on the

valley floor has a relatively large proportion of blueish-grey varved clay, which probably

originates from the lower part of the slope."

After the 1973 slide, two separate potential slide masses remained adjacent to the Attachie. They have been designated as the upstream and the downstream slides. The upstream slide (Fig. 5.2.4) underlies the upper portion of the 1973 slide. The backscarp of the upstream slide was defined by a steep, 700m long tension crack located about 65m behind the 1973 backscarp. This crack is as much as 1 m wide with up to 3 m of vertical displacement. It is open in some locations to as great as 8 m of depth. The lateral boundaries of the upstream slide have been determined by the extent of this feature. The base of the upstream slide daylights at about El. 535 where disturbed debris is overthrusting undisturbed sediments and causing continual ravelling. The total volume of moving soil is about 11 x 106 m3 (BC Hydro, 1981).

91 The failure surface of the downstream slide (Fig. 5.2.4) is defined by a pinched-off piezometer in

DH63-2, shearing in inclinometer in DH63-7, and a tension crack, located immediately downslope

of DH63-3. The lateral boundaries of the slide were estimated from topography. The total

volume of moving soil in this area is about 10 Mm3.

Monitoring

The unstable masses described above have the potential for catastrophic failure in the same order

of magnitude as the original 1973 slide. Consequently, BCHydro has been monitoring movement

of the slope since 1978. The upstream slide is being monitored by inclinometer 63-4, located just

above the prominent tension crack defining its upper boundary and inclinometer 63-7, located

centrally within the downstream slide.

Monitoring of Attachie Slide was discontinued in June 1984 but recommenced in 1989. The

frequency of readings was twice per year in 1990 and once per year since 1990. Movement

trends at Attachie up to 1995 are summarized in BCHydro report MEP185 (1996) as follows:

Inclinometer 63-4 (Upstream Slide)

• No apparent movement has been recognized up to 1995.

Inclinometer 63-7 (Downstream Slide)

• Possible Surface Movement Zone - Occurring within clays at 1 to 5 m depth. The 1995 data

confirmed movement in this zone with a present rate of about 0.4 mm/year and total 92 movement since 1980 of about 19 mm. Movement appears to be somewhat erratic since

1992. A slowing trend is indicated in current data.

• Possible Movement Zone - Movement occurs within a silty clay unit at 25.6 to 26.2 m depth.

Fairly consistent movement in the order of 1.8 mm/year tracked between 1980 and 1989; no

apparent movement observed between 1989 and 1993; and movement has apparently resumed

as of 1993 at a rate of about 0.3 mm/year, however, may not be real as an uphill direction is

indicated. Total movement since 1980 has been about 5.2 mm.

• Basal Movement Zone - Confirmed movement zone at 71.9 to 73.8 m depth which

corresponds to a well defined shear zone in laminated silts and black clay logged in the drill

core. The inclinometer casing was pinched-off in this zone in November 1980. Over the

short period of recorded movement between installation in July 1980 and November 1980, the

movement rate was in the order of 156 mm/year.

In addition, a system of 6 extensometers installed across the tension crack defining the upper

boundary of the upstream slide monitored displacements from May 1980 to May 1984. Due to

frequent damage and recalibration of instruments it is not possible to quantify the total slope

displacement since measurements began. However, extension across the tension crack appeared

to occur at rate of approximately lcm/month with a maximum of llcm/month in June 1980. The

apparent trend for the interval recorded, is a gradual decrease in the rate of extension to a

minimum of 6mm/month between Oct. 1983 and April, 1984.

93 A summary of monitoring of survey monuments is illustrated in Figure 5.2.4. Data was collected by BC Hydro between 1980 and 1984 and supplemented by data collected by the Ministry of

Transportation and Highways between late 1983 and August 1989 (Richmond, 1990). Two modes of displacement appear to be occurring. The first is a relatively constant downhill creep which varies between 10 and 30 mm/year. The second consists of periods of substantially higher rates of movement as exemplified by that of M5 which moved 384 mm/month between April to

Oct. 1983. The MOTH measurements indicate that large movements also occurred between

March 1984 and October 1985 and between August and November 1987. Not all monuments

were affected by these very slow displacements which averaged about 15 mm/month (Richmond,

1990). Readings for 1988 and 1989 show slow creep at some monuments while others showed

slow movement; for instance MOTH Monument Bl, located at the crest of the downstream slide

area, had an indicated 260 mm displacement over a 19 month period. Movements of the

monuments located within the upstream slide mass trend northwest. Movement of the upper

slope of the downstream slide trend northwest whereas the lower slope trends north with almost

no westerly component. There appears to be no consistent seasonal variation of the creep

displacement nor is there a seasonal correlation to the initiation of larger movements.

It is probable that the segment of the slope which failed in 1973 exhibited a similar pattern of

slow, episodic slope movement prior to catastrophic failure.

94 5.2.4 Soil properties

Classification Tests

Classification tests including Atterberg Limits, natural moisture content, and grain size analysis were completed for the three units involved in the landslide; preglacial lake clay, till, and post• glacial lake clay. Tests to obtain material properties were conducted on the two units principally involved in the slide movement, that is, the till and preglacial lake clay. A summary of classification tests is given in Table 5.2.2. Details are contained in Appendix II.

Table 5.2.2: Summary of classification test results, Attachie landslide Property Till Pre-glacial lake sediments n=3 Plastic Silt n=38 n=7 Natural water content 24 31 {16-35} {20-37} - Liquid limit (%) 38 41 30 {33-44} {27-59} {NP-34} Plastic limit (%) 17 21 21 {8-22} {14-27} {NP-24} Plasticity index (%) 17 18 9 {17-22} {8-34} {NP-12} Liquidity index (%) 0.27 0.19 ~ {-0.01-0.57} {-3.18-0.59} Clay content (%) 31 46 16 {19-37} {28-68} {7-27} Silt content (%) 46 54 84 {25-63} {32-72} {73-91} Bulk density (kg/m3) 1947 {1846-2171} n=15 n=number of samples tested NP=non-plastic

Bulk properties

The preglacial lake deposits are generally classified as inorganic clay of low to medium plasticity

(CL) (Figure 5.2.10). Grain size analyses indicate that most samples described as clay are actually

95 clayey silt with an average clay fraction of 40. Although non plastic silts are known to be abundant, with the exception of two samples collected from a depth of 70m and 77m in borehole

63-2, few samples were successfully recovered from drill holes. However, in August 1998, three samples of hard, dry laminated silt (T3-1 - T3-3) were collected from a 12 m surface exposure at an El. of 530 m, 29 m above the top of the gravels, in the downstream side scarp. These samples represent a thick non-plastic to low plasticity silt to fine sand facies of the glacio-lacustrine sediments. Evidence of calcite and gypsum was found on the slope, which may act as cementing agents in the sediments.

Figure 5.2.10. Plasticity chart of materials sampled from Attachie landslide.

60

50

•

CH A-li ne

40 • MH • • • CL • • 20 £ —#-# fr- • Glacio acustrine clay fr om failure p ane • • • • Colluvi um * y • • • / A Till 10 CL CL-ML ML ML 0 10 20 30 40 50 60 70 80 90 100

Liquid Limit(%)

96 A relatively stiff consistency and light overconsolidation is indicated by the Liquidity Index, which ranges from -0.04 to 0.6. Activity is low to medium (0.5 to 1.0) indicating low swelling potential comparable to that of kaolinite.

The till is classified as a well graded, medium plastic silty clay (CL).

Material adjacent to the main sliding surface

An exposure of the main sliding surface was exposed in a gully at the toe of the slope. The clay

layer that contains the slickensided failure surface is a sheared marl unit, beige, and approximately

0.5-1 cm thick (Figure 5.2.8). This material is classified as low plasticity clay (PI=13) with a clay

content of 46%. Underlying the sliding surface, the material grades downwards from silt to fine

sand underlain by basal gravels.

Table 5.2.2: Summary of classification test results adjacent to the main sliding surface Location CC LL PI (%) (%) (%) Rupture surface exposed at toe 46 26 13

The mineralogy of the clay adjacent to the sliding surface exposed at the toe of the slope was

evaluated by X-ray diffraction. The results suggest that the clay fraction (<2u.m) is primarily illite

and kaolinite (ratio ~ 17:3) and glacially ground rock flour (quartz and feldspar). Vermiculite and

smectite are also present, though no quantification was undertaken.

97 Material properties of the preglacial lake sediments

Several samples of materials from the vicinity of the known sliding surfaces have been subjected to direct shear and triaxial testing by BCHydro. Tests were performed on intact blocks excavated from surface exposures, and from shelby tube samples recovered during drilling. Direction reversal direct shear tests were conducted parallel to bedding to determine peak and residual effective shear strength parameters. Residual tests were also carried out on remoulded samples in the same normal stress range. One consolidated, undrained triaxial compression test was carried out for undrained shear strength and to obtain an estimate of the pore pressure parameters. A summary of all laboratory testing follows, details are included in Table II. 1 of Appendix II.

In order for a flow slide to occur, the material must exhibit some degree of brittleness. One soil

property used to relate post-rupture behaviour is the brittleness index (Bishop, 1973). Drained

IB=(T'P~ t'r)/ t'p where x' is the resistance to shear for a given value of effective normal stress, and:

Undrained IB - [(cu)P - (Cu)r]/(Cu)P where c„ denotes the apparent cohesion for the case <|) = 0. For

each equation, the suffixes f and r relate to the peak and residual stress.

Results of direct shear tests show a moderate degree of drained brittleness. The brittleness index

ranges from 0.02 - 0.39. In a normal stress range from 500 to 3000 kPa, peak friction angles (fy'p)

range from 20.9° to 27.5° with cohesion of 0 to 230kPa. Residual friction angles (')r range from

16.7° to 25.5°.

98 Compression Index, obtained from the consolidation stages of three direct shear tests run on clay material from 60 to 70 m depth, range from 0.06 to 0.09, indicating some overconsolidation

(normally consolidated clay of the same plasticity would have a Compression Index of about 0.3).

(Terzaghi and Peck, 1948)

Three consolidated, undrained triaxial compression tests were run on samples from 46 m depth

with effective confining pressures of 500 - 2000 kPa. Results are shown in Figures 5.2.11 and

5.2.12. Assuming a zero cohesion intercept, peak friction angles ((j)'p) range from 31.1° - 32.0°,

consistent with values obtained from the direct shear tests. At confining stresses greater than

1250kPa, a cohesion intercept of 176kPa was measured.

The pore pressure response indicates light overconsolidation. This is shown by the pore pressure

parameter A ranging from 0.12 to 0.55 for effective confining pressures of 600-1400kPa,

corresponding to 30 to 70 m depth. According to a diagram published by Bishop and Bjerrum

(1960), this would correspond to an Overconsolidation Ratio (OCR) of approximately 2 to 3 at

the sampling depth. A plot of stress paths, Figure 5.2.12, also suggests that the material behaves

as slightly overconsolidated and does not exhibit an undrained brittleness (Lambe and Whitman,

1969). Morgenstern (1992) lists several limitations to laboratory testing of stiff, dilatant clays to

determine strength parameters including: moisture sensitivity, fissuring, difficulty of obtaining an

undisturbed sample, and pre-existing shear surfaces formed under a stress regime very different

from that acting today. Consequently, laboratory testing tends to indicate residual strengths

higher than those existing in-situ.

99 Though not conducted to any standard procedure, it is of interest to note that a weak solution of

10% HCI applied to a block sample of clayey silt from the base of the preglacial succession reacted by fizzing and bubbling. This rough experiment suggests the presence of carbonates, which may form a cementing agent.

In summary, the preglacial lake sediments involved in the landslide can be characterized as lightly overconsolidated, stiff, clayey silt of low plasticity. The granular (sand and gravel) interbeds are very dense. Cementation may exist.

Figure 5.2.11. Consolidated, undrained triaxial compression tests of preglacial lacustrine clay.

2000

Effective Normal Stress (kPa)

100 Figure 5.2.12. Consolidated, undrained triaxial compression tests of preglacial lacustrine clay

2000

-

1500

-

0

>/ & & o 500 o

I i i 1 I I I 500 1000 1500 2000 2500 3000 p= (kPa)

5.3 Failure Behaviour

5.3.1 1973 Flowslide

Major deformations of the slope occurred prior to the earliest photos from 1945 (A8294-63). A

fully developed main scarp was present in 1952, showing a maximum normal displacement of

approximately 25m. The intermediate scarp was also observable, although largely covered by

vegetated debris. Another fresh internal scarp, about 10m high could be observed circumscribing

a slump feature 480m wide and 120m long, near the toe of the landslide. None of these features

changed appreciably between 1952 and 1973, although their appearance remained fresh,

suggesting continuing slow movements. Some re-activation was reported by local residents two

weeks before the 1973 failure (Evans et al., 1996). On May 26, 1973, between 23:45h and

101 23:55h PDT, farmers residing on the north side of the Peace River directly across from the

Attachie Slide reported hearing "a series of loud reports and thunder-like noises accompanied by the sound of rushing water" (Thurber Consultants, 1973).

Piecing together all observations, the following sequence of events is proposed. The 1973 flow slide initiated when the soil within the lower stage of the unstable area mobilized forward by an amount sufficient to launch 6Mm3 down the toe scarp to the floodplain. The distal part of the mass traveled first. A large treed flake of debris, which originally rested on the proximal part of the lower stage slope moved about 200m and came to rest just short of the toe scarp, underlain by remoulded debris. Similar phenomena have been observed in flake slides in Norway (Karlsrud et al., 1984) and quick clay flow slides of Eastern Canada (Evans and Brooks, 1994).

Blocks of till were arranged in the head region into sharp ridges parallel with the main scarp

(Figure 5.2.9). Similar ridges can be observed in the scars of some quick clay flows in Eastern

Canada (Carson, 1977) and north western B.C. (Geertsema and Schwabb, 1997), where they

result from the kinematics of retrogressive failure. Thus, what probably followed were one or

several retrogressive failures of the upper stage, due to undercutting of the toe of this mass by the

primary movements. The material in the upper stage retrogressions was less mobile, perhaps due

to the predominance of till. Therefore, much of the upper stage remained in its source area, while

the remainder filled the void behind the intact flake of the primary failure with debris, rich in till

blocks. The top of the slope retrogressed 61-76m.

102 Disintegration of the lower stage soil masses appears to have been largely completed before their

descent over the toe scarp into the flood plain. Only a limited portion of the ground surface

remained intact and appears to have been rafted over fluid, disintegrated debris.

5.3.2 Trigger

An obvious trigger of the 1973 event has not been identified. However, there may have been a

number of contributing factors.

As the base of the rupture surface is well above river level, it may be concluded that river erosion

did not play a significant role. Thurber Consultants Ltd. (1973) made the observation that there

were few human-made changes in the area. No significant earth tremors were recorded in this

area on the day of the slide.

Temperature and precipitation data since September 1971 from both Ft. St. John Airport and

Hudson Hope are attached in Appendix II. From 1971 to 1973 the total annual precipitation was

greater than the long term average (Figure 5.3.1), 1972 being the 2nd wettest year on record, and

the wettest year since 1957. Furthermore, for the last two winters (Sept. - May) the snowfall was

greater than the long term average (Figure 5.3.3), whereas the temperatures were below or close

to normal. With a cool spring preserving the snow pack, and subsequent rain on snow, this

condition could well have resulted in higher than normal groundwater levels. April precipitation

exceeded the mean by 168% and although May was a very dry month, on May 26, the day of the

slide, it rained steadily from 10 a.m. to 2 p.m., and drizzled for the remainder of the day (16.3

mm recorded at Fort St. John Airport and 2.8 mm at Hudson Hope).

103 Evans et al. (1996) reports that many landslides occurred in the Fort St. John area in the spring and early summer of both 1973 and 1974 following a very wet summer in 1972 (Figure 5.3.2).

The slide area had been slowly failing for many years, resulting in a disturbed ground surface

characterized by scarps, open surficial cracks, and minor slumping. A distressed state allows

rainfall to infiltrate and saturate the mass quickly, creating excessive pore pressures along critical

failure surfaces. Furthermore, the shear strength of the failing materials is lowered by softening

and mechanical degradation. Thus the slope may have had a very low factor of safety prior to

failure and may have been triggered by a relatively modest, short duration rainfall or rain on snow

event.

Figure 5.3.1. Monthly precipitation, Fort St. John Airport, May 1972 to May 1973.

104 Figure 5.3.2. Monthly rainfall, Fort St. John Airport, May 1972 to May 1973.

160 T ——

Month Figure 5.3.3. Monthly snowfall, Fort St. John Airport, May 1971 - May 1973.

100

MJ JASONDJ FMAM

Month 5.3.3 Landslide dam

The flow of the Peace River was blocked for about lOh by the debris that descended the shale scarp and crossed the floodplain. Debris in the valley covered an area of 83.5 ha and had an average depth of 7.6m (Evans, 1996). The debris dam was breached gradually, without catastrophic release and consequent flooding. Post-slide aerial photographs show that a well defined overflow channel, about 150 m in width, had been established within 40 h of the slide occurring (Figure 5.2.2).

"The elevation of the silt line formed by the pondage upstream of the debris was 7 m above

normal river level (Thurber Consultants, 1973). The average discharge of the Peace River at

Hudson Hope (34 km upstream of Attachie) for the duration of the blockage was in the order of

900 m3/s (Environment Canada data); a minimum estimate of the volume of water impounded by

the Attachie debris dam is therefore 32.4 Mm3" (Evans et al., 1996).

5.4 Failure Analysis

The Attachie Landslide developed in two phases of movement. The first was an initial

widespread ductile sliding of slopes under drained conditions. This was followed by a sudden

flow slide, considered a brittle, undrained phase. Consequently, this chapter consists of two parts.

The first studies the mechanics of the ductile phase using limit equilibrium analyses of the pre-

slide slope. The second part studies the mechanism of the flow slide using a dynamic analysis of

the displaced mass. The analyses considered the primary failure, involving the lower stage of the

instability only, below the internal scarp.

106 Assumptions regarding reconstruction of the rupture surface, the prefailure surface, phreatic surface and material strengths are inherent in the use of the models. Therefore, it is not possible, and is not the intent, to accurately model the conditions of the slope and obtain specific values.

However, the analyses are used to test assumptions regarding the parameters and the hypotheses concerning the mechanism of failure.

5.4.1 Limit equilibrium analysis of initial widespread sliding

Conditions of the first-time failure are not known. Progressive failure may have played a role, or

possibly pre-shearing by glacial thrusting. In any case, the main rupture surface was pre-sheared

prior to 1973. Presumably the slope movement was slow and intermittent, similar to that

monitored in the downstream slide area (Section 5.2.3) and thus drained conditions applied.

The soil properties used in the CLARA analysis are summarized in Table 5.4.1. Due to pre-

shearing, it was assumed that the frictional strength along the rupture surface had been reduced to

residual with no cohesion. For simplicity, the strength is assumed to be uniform for each of the

main segments of the rupture surface. A residual friction angle of 14.5° was assigned to the sub-

horizontal bedding plane that the failure surface exploits. This is a lower value than the 17°- 25°

obtained from laboratory test results as discussed in section 5.2.4. If a (j)'r of 17° is selected to

conform to laboratory results, an equivalent piezometric surface, ru, of 0.35, or depth to

piezometric surface of 15m is required to initiate movement. This value is considerably higher

than levels recorded in any of the piezometers. Morgenstern (1992) noted that laboratory

simulations may not reproduce in-situ conditions and tend to indicate residual strengths higher

than those existing in-situ. Also, it is possible that the weakest horizon in the stratigraphic sequence was not tested. The selection of a lower value is justified by these arguments and by published correlations between residual friction angles and liquid limit (Stark and Eid, 1994;

Mesri and Cepeda, 1986).

While the weak, subhorizontal surfaces are believed to exploit bedding of plastic clay, the steeply dipping segments of the failure surface cut across bedding and therefore have been assigned a

higher

The configuration of the pre-failure ground surface (a) and sliding surface (b) is shown in Figure

5.4.1.

Pore pressure distribution

Using the above selected material strengths in CLARA, an equivalent piezometric surface

corresponding to ru = 0.26 was required to maintain stability of the slope, corresponding well with

the measured pore pressures. This corresponds approximately to a constant depth of 26.5 m

below ground surface.

Table 5.4.1. Summary of soil properties used in CLARA

Material Soil Unit weight, y Friction angle, (j)'r r„ (kN/m3) (degrees) Plastic clay sliding surface - 14.5 0.26 1 Cross-bedded clay and silt sliding 19 17.5 0.26 surface 2 Clay Till 21 23 0.26

108 Figure 5.4.1 Configuration of the (a) pre-slide ground surface and (b) rupture surface of the Attachie landslide. The 3D version of CLARA uses Bishop's Simplified Method. Given the strongly angular shape of the sliding surface, it is considered that Bishop's Method may be excessively conservative, as it neglects internal strength (Hungr et al., 1989). Further analysis was therefore carried out with a two-dimensional model of a typical cross-section of the lower stage mass, shown in Figure 5.4.2.

The Bishop Factor of Safety for this cross-section and the same parameters is 0.92, while the

Spencer's Method, which includes internal strength, yielded 1.05. Similar results were found with the Morgenstern-Price Method by Thurber Engineering (1973) and Evans et al. (1996).

Thus, the slope instability can be explained using conventional Limit Equilibrium analysis, with plausible input parameters. What remains to be explained, however, is the sudden acceleration and flow slide of May, 1973. The motion of the flow slide was simulated using the dynamic analysis model DAN (Hungr, 1995).

Figure 5.4.2. Two dimensional model of Attachie landslide lower stage.

0 50 100 150 200 250 300 350 400 (m)

110 5.4.2 Dynamic analysis of flow slide

DAN models the displacements of the mass resulting from a drop in the shear resistance at the base of the mass to a residual and possibly time dependent value, and a change of the sliding mass into a frictional (Rankine) fluid. The internal stresses vary around the geostatic value, depending on strain. The frictional (rate independent) rheology was used.

The objective of the dynamic analysis was to determine the reduction in shear resistance required

to yield the mobility and the post-failure geometry exhibited by the flow slide. This was

accomplished by increasing ru along the rupture surface. An accurate representation of the post-

failure profile was obtained with a residual soil strength between 14.5° and 17.5° along the main

sliding surface, and ru = 0.45.

Figure 5.4.3. Output from DAN analysis of Attachie landslide.

700 Pre-failure surface

Post-failure surface

600

s ei I 3 500

400 Figure 5.4.4. A simplified, approximate analysis of the balance of forces operating at the initiation of the flow slide.

150

— 100

— a 4 —^\ 50 o w "3 I s P > — H ! y \ 0 t R -50 1 1 II 1 1 1 1 1 1 1 1 1 I I I 1 1 1 1 MM 1. 1 1 1 MM

0 50 100 150 200 250 300 350 400 (m)

F = factor of safety = R/P = 0.50 P = yh2/2 R = WtancK

where y = bulk unit weight of active wedge = 19kN/m3 h = 84 m W = weight of passive block = 233.6kN/m

A simple approximate analysis was used to assign an index representing the balance of forces to

the post-failure conditions, in the context of the above-described model. At the onset of

movement, the flowing mass still has its original geometry, as shown in Figure 5.4.4. At first, the

resisting and driving forces are equal. However, movements are soon developed, because the

basal boundary strength has fallen to the value of b and the slide material has changed into a

heavy fluid, with a hydrostatic lateral stress. At this point, the driving force acting on the slide

112 2 mass is the heavy fluid pressure, yH /2, while the resisting force is W tan

obtained from the DAN analysis, that is 9.8° along the backscarp and 8.1° along the basal rupture

surface, the ratio between the resisting and driving forces is calculated as 0.50. Thus, the ratio of

resisting to driving forces (i.e. the Safety Factor) must have fallen from 1.0 at the onset of failure

to approximately 0.50 in course of failure, in order to produce the observed overall

displacements.

5.4.3 Flow slide mechanism

The renewal of movement on pre-existing shears is usually of modest speed and travel due to the

generally non-brittle nature of the stress-strain curve for such surfaces. Furthermore, rapid, brittle

flow is not the usual mode of movement in insensitive plastic soils (Skempton and Hutchinson,

1969). The following three hypotheses are proposed to explain the transition exhibited by the

Attachie slide from a slow, ductile failure to a rapid flow slide.

Hypothesis A: Undrained brittleness of the rupture surface

This is the usual cause of flow slides in brittle or collapsive soils, whether granular or plastic (e.g.

Hutchinson, 1988). The soil must possess a loose structure, which collapses in response to some

disturbance. The undrained tendency for volume reduction causes a sudden pore-pressure

increase and a loss of strength. At Attachie, however, no collapsive or sensitive soils existed near

the rupture surface, which had been compacted by Pleistocene ice and had already been pre-

sheared by ductile displacements of many metres. This hypothesis is consequently considered

improbable.

113 Hypothesis B: 'Macroscopic' brittleness

A process that leads to what can be described as "macroscopic" brittleness can be described as follows: The clay, silt and till soil develops a system of jointing and multiple open tension cracks

as a result of shear failure and slow displacements. The open cracks then fill with water and

softened, remoulded, or loosely deposited material released by loosening of the crack walls.

Furthermore, softening and mechanical degradation results in a lowered shear resistance adjacent

to the crack walls. A substantial portion of the previously intact soil volume is thus transformed

into a mass of blocks of intact clay and till separated by joints filled with highly disturbed, loosely

deposited matrix. Prior to catastrophic flow slide failure, the crack contents are saturated by

inflow of surface water. Liquefaction of the loose material then occurs, exerting high fluid

pressure inside the cracks. This suddenly increases the driving forces acting on the passive distal

segment of the slide mass.

A somewhat similar process leading to a flow slide from a dense glacial till stratum overlying

plastic clay has been described by Terzaghi (1950):

"The slide at Swir is an example of a process leading to a mechanical mixture of slide

material with water. Before the slide occurred, the glacial till overlying the Devonian clay

was firm and stable, and its porosity hardly exceeded 25%. The expansion of the underlying

clay broke the glacial till into large fragments. Rain water accumulated in the crevices

between fragments. It caused disintegration and collapse of the fragments; and the mixture

of water and till fragments flowed into the cut."

114 The Attachie Slide has some different attributes but the principle of gradual mechanical disturbance leading to spontaneous disintegration is the same as proposed by Terzaghi. The

mechanism requires a high degree of jointing and cracking of the slope mass due to preceding

sliding movements, which was observed at Attachie. A similar mechanism was proposed by

Savigny et al. (1992) to explain liquefaction of thick, glacial-drift derived colluvium. In this case,

the open, metastable structure was attributed to cryogenic and mass wasting processes in a

permafrost region.

The change of Factor of Safety corresponding to this process can again be represented with the

help of the model illustrated in Figure 5.4.4: The resisting force, using the friction angle of 14.5°

and an ru of 0.32 which produced a static F of S of approximately 1.0 using the Spencer's Method

is W (1-0.32) tan 14.5°, where W is the weight of the passive block (2.34xl05 kN/m). The

driving force again equals the hydrostatic thrust of a heavy fluid: yh2/2 = 6.68xl04 kN/m. The

ratio of the two is 0.62, showing that the factor of safety may have dropped by as much as 38%

as a result of the macroscopic liquefaction process. It must be noted, however, that such a

process is unlikely to have taken place simultaneously across the width and depth of the slide

mass. Thus, the given estimate of the magnitude of Factor of Safety reduction is likely an upper

bound.

Displacements from the dynamic analysis using an ru of 0.32 are not as large as exhibited in

actuality. Therefore, even with the assumption of fluid-like disintegration of the slide mass

subsequent to initial acceleration by the mechanism described by Hypothesis B, some boundary

115 surface strength reduction is also needed, Thus, Hypothesis B alone is insufficient to explain the mobility of the slide debris.

Hypothesis C - Internal strength of the slide mass

Hutchinson (1987) reviewed various mechanisms by which unexpected, rapid movement occurs.

One mechanism was illustrated by the 1963 Vaiont slide in Italy. According to Hutchinson

(1988) "it formed the basis for recognizing the group of compound slides released by internal

shearing occurring towards the rear of a highly brittle slide mass".

The essential features of the Vaiont slide were that it was a renewal of movement on a pre•

existing, deep-seated non-circular (listric) slip. Important to the stability is the great difference

between internal strength and the strength of the sliding surface (Hungr et al., 1989). While the

major part of the sliding surface followed clay seams at residual strength and was consequently

ductile, nearly 50% of the resistance arose from the more brittle strength of the limestone rock

mass. This helps explain the rapid initial acceleration of the slide, once the peak strength of the

limestone was exceeded. The factors which lead to such behaviour appear to be:

1. High normal effective stresses on the slip surface bounding the slide mass.

2. A compound (markedly non-circular) shape of the bounding slip surface, so that the

slide forms initially a kinematically inadmissible mechanism and must deform internally in

order to be able to move forward along the bounding surface.

3. A strong contrast between a low shear strength bounding slip surface and a high

strength slide mass.

\ 116 4. High values of brittleness on internal shear surfaces within the slide mass, particularly

towards its rear above the change of slip surface angle from the rear scarp to the sole of

the slide.

5. The build-up of a significant, stored deficit in the overall factor of safety on the

boundary shear surfaces as a result of the partial submergence of the slide toe and high,

rainfall-induced groundwater levels. Accumulation of this deficit was made possible by

items 2 and 4 above.

6. A sudden, brittle failure within the slide mass, triggering a sudden drop in overall factor

of safety to its deficit value and releasing the slide mass by transforming it into a

kinematically admissible mechanism.

All of the above conditions exist in the lower stage of the Attachie Slide. The strong, brittle element is thought to be the very dense, cemented silt facies observed at mid-slope. Although sheared along the rupture surface and displaced by a number of meters, this package of brittle soil may have retained sufficient rigidity to present resistance to motion at the sharp angular change of the rupture surface inclination in the rear of the slide. Partial saturation of the soil may also contribute to the brittleness of the silt.

An attempt to quantify this phenomenon was carried out using the Sarma limit equilibrium analysis (Hoek, 1987), which is able to fully account for frictional as well as cohesive internal strength of the sliding body. The 2D limit equilibrium analysis of the surface shown in Figure

5.4.2 was repeated with the Sarma Method. This geometry includes one internal slip surface which accounts for downward movement of an active wedge and outward movement of a passive

117 wedge along the weak base. For each analysis performed the orientation of the internal slip surface, a, was varied between 0 and 20° to the vertical to obtain the minimum factor of safety and thus the most likely orientation exploited. It was found that in each case a=10°-15° gave the lowest factor of safety. The selection of shear strength parameters along the internal slip surface were chosen to correspond closely to those obtained from the undrained compression tests of

Section 5.2.4 with c' = 176 kPa and 'P = 30°.

Table 5.4.2. Summary of soil properties used in Sarma limit equilibrium analysis. FOS Bounding surface Internal Shear ru Cross-bedded silt and clay Sub-horizontal clay surface c' c" c' ' 0 17.5° 0 14.5° 176 25 0.45 1.00 0 17.5° 0 14.5° 0 25 0.45 0.83

Using the same residual friction angle on the rupture surface, a pore-pressure corresponding to an

r„ = 0.45 was required to obtain a Factor of Safety of 1.00. This pore pressure ratio is relatively

high though it is possible that such high pressures developed along the basal rupture surface prior

to the 1973 failure. Alternatively, it should be considered that the shear strength mobilized along

the surface may have been lower than assumed. Brittle failure of the cemented silty layers can be

simulated by removing the internal cohesion, as if a continuous near vertical shear surface

developed through the silt strata. The Sarma Factor of Safety then falls to 0.83. This 17% loss in

overall force balance is thought to contribute to the acceleration of the slide mass leading to the

flow slide.

It is important to note that these same parameters, that is the residual friction angles and pore

pressure ratio summarized in Table 5.4.2, are the same required to model the post-failure

118 geometry exhibited by the flow slide in the DAN analysis. Thus, this mechanism provides the required reduction in strength to not only initiate acceleration of the slide mass, but to accurately model the post-failure geometry as well.

In summary, the occurrence of a large rapid earth flow is surprising, given the insensitive and relatively non-brittle nature of the stiff, clayey silt forming the bulk of the overburden. It is further unusual that the Attachie Slide rapid earth flow occurred at a location which had previously been subject to a large amount of disturbance by slow sliding movements. However, both Mechanisms B and C above can lead to substantial sudden changes in the balance of forces within the slide mass. The quantitative estimates of these changes are estimated above as up to

38% for Mechanism B and 17% for Mechanism C. It is not possible to state definitively that the

failure occurred by one of the suggested mechanisms, or that others do not exist. Furthermore, as

is often the case for most landslides, more than one mechanism may have occurred

simultaneously, combining to produce the spectacular flow slide event.

6.0 Comparison of two land slides

The two landslides had similarities in geological setting, geometry, material properties,

morphology and behaviour prior to the 1973 flow slide at Attachie. The key parameters are

summarized in Table 6.1.

119 Table 6.1. Comparison of key parameters, Attachie and Slesse Park landslide. Attachie Slesse Classification of lacustrine sediments non-plastic to moderate moderate to high plasticity plasticity Stratigraphy of source volume % clayey silt 48 61 % non-plastic silt 31 % sand 21 16 % till Geometry Height 162 53 Width 760 580 Length of rupture surface 570 160 Total length 1500 250 Depth to rupture surface 83 25 Volume 12.6 1.84 Pre-failure slope gradient 17° 16° Fahrboschung 7.7° 12° Failure behaviour Ductile Rapid Ductile Rapid phase phase phase phase Maximum rate of movement slow extremely slow rapid rapid Maximum displacement 25 900 21 30

The most significant difference between the two cases, apart from scale (volume) is the presence

at Attachie of large quantities of low plasticity to non-plastic, cemented silt. Though the

lacustrine sediments at Slesse are generally classified as medium plasticity silty clay, thick

sequences of non-plastic silt have not been found. It is the higher average plasticity of the soils at

Slesse which explains the more ductile, plastic movement.

The second difference between the two slides is the climate in which they exist, and the associated

ground water conditions. Given the wet, coastal climate, and the low permeability of most of the

soils the groundwater table at Slesse is high thus the slide mass remains mostly saturated. At

Attachie the climate is sub-humid. The groundwater table is low, and the coarser soils are

120 believed to be partially saturated. The rapid decrease in shear strength exhibited by silt upon wetting has been documented by Lum (1979).

Very rapid, undrained failure at Slesse is limited to the highly disturbed, weathered clay and silt at the toe of the main unstable mass and the mechanism of undrained failure is very different. A high degree of mechanical degradation combined with softening due to infiltration along extensive joints and cracks eventually saturates the material. The softened material, with a water content at or near the liquid limit, fails as an earth flow. The wet climate of the Chilliwack Valley contributes to this process.

Differences in the geometry of the slide surface may also> contribute to differences in the failure behaviour. The depth to the bounding slip surface is much greater in the case of the Attachie, a maximum of 75m compared to 25m at Slesse. Furthermore, at Slesse, the rupture surface was

interpreted as having more than two intersecting planar segments. Therefore, the Slesse slide is

not as likely to develop a kinematically inadmissible mechanism. The development of a

kinematically inadmissible mechanism is key to one of the three mechanisms proposed to explain

the development of a flow slide at Attachie.

7.0 Conclusion

A comparison was made of two large landslides in over-consolidated glacio-lacustrine deposits.

In each case, the flat-lying main part of the rupture surface was seated in an overconsolidated

stratum of glacio-lacustrine clay and silt, tens of meters above the present valley floor. The

primary instability was/is a ductile compound slide, which moves episodically at slow speeds, in

121 response to pore pressure changes triggered by extreme rainfall and especially snowmelt. The primary instability has been analyzed using conventional limit equilibrium analyses.

Both slopes developed sliding displacements of tens of meters, manifested by prominent scarps.

After this initial movement, the Attachie slide changed its character suddenly and produced a rapid flow slide, spreading out across the river in a matter of seconds and causing a displacement wave on the opposite bank. The Slesse Park slide has continued intermittent displacements of limited extent and velocity for over 20 years and up to the present time.

Two mechanisms have been proposed to explain the rapid transition from a slow, ductile failure to an extremely rapid flow slide.

The first is a process that leads to what can be described as "macroscopic" brittleness. In short, a

substantial portion of the previously intact soil volume is transformed into a mass of blocks of

intact clay and till separated by joints filled with a disturbed, loosely deposited matrix.

Liquefaction of the loose material then occurs, exerting heavy fluid pressure inside the cracks.

This suddenly increases the driving forces acting on the passive distal segment of the slide mass.

A simple, approximate analysis estimated a 38% reduction in the factor of safety as a result of the

macroscopic liquefaction process.

The second was proposed by Hutchinson (1987) to explain the 1963 Vaiont Slide in Italy. The

factors which lead to this behaviour include: 1) a compound (markedly non-circular) shape of the

bounding slip surface, so that the slide forms initially a kinematically inadmissible mechanism; 2)

122 a strong contrast between a low shear strength bounding slip surface and a high strength, brittle slide mass. These conditions may produce a stored deficit in the overall factor of safety on the boundary shear surfaces, which can be suddenly released by a brittle failure along internal shears in the slide mass. Quantification of the influence of internal strength to the overall factor of safety was carried out using a variation of the Sarma Method of analysis. Brittle failure of the cemented silty layers was simulated by decreasing the cohesion along an internal slip surface from 176kPa to 0. The response is a decrease in Factor of Safety from 1 to 0.83.

A realistic simulation of the post-failure geometry and mobility exhibited by the Attachie flow

slide was achieved using the dynamic analysis DAN, with similar input parameters as those obtained from the Sarma limit equilibrium analysis. The results suggest that the latter of the two

mechanisms can explain the rapid acceleration and subsequent flow slide. However, neither are

sufficient to produce the 50% reduction in Factor of Safety in order to produce the observed

post-failure geometry. Therefore, it is proposed that both of the mechanisms may have occurred

simultaneously in certain degree, combining to produce the spectacular flow slide event.

The main instability at Slesse is also the direct cause of rapid, undrained failure. However, they

are relatively small (<15 000 m3), limited to the highly disturbed clay and silt at the toe of the

main unstable mass, and occur by a very different failure mechanism. A high degree of mechanical

degradation combined with softening due to infiltration along extensive joints and cracks

eventually saturates the material which subsequently fails as a mud flow.

123 The main apparent difference between the two cases involves the presence of large quantities of low plasticity silt in the rapid Attachie event. Based on stratigraphic mapping of surface exposures, there does not appear to be the presence of low plasticity silt, or more importantly, cementation of this unit. It is the higher average plasticity of the soils which explains the more ductile, plastic movement. Other factors that may contribute to the difference in failure behaviour are: likely cementation of the silt and clay sediments in case of Attachie, contributing to brittle failure; and that the Slesse landslide is a shallower failure of which a significant portion of the slide mass has already been depleted by secondary movements. Though it is unlikely that the

Slesse Park landslide will exhibit failure similar in magnitude or mobility as the Attachie landslide, mudslides capable of partial damming of Chilliwack River should be considered likely.

The results of this research suggest that the presence of non-plastic units in a moving landslide involving overconsolidated lacustrine soils should be considered as a potential cause of

catastrophic flow sliding.

124 References

Beland, J. 1956. Nicolet Landslide, November 1955. Proc. Geol. Assoc. Can., 8 (1), pp. 143-156

BCHydro, 1981. Peace River development Site C project, Report on reservoir slopes. Report No. H1361.

Bishop, A.W. 1955. The use of the slip circle in the stability analysis of slopes. Geotechnique, 5(1), 7-17.

Bishop, AW. 1973. The Stability of Tips and Spoil Heaps. Quarterly Journal of Engineering Geology. Vol.6, pp.335-376

Bishop, A.W., and Bjerrum. 1960. The relevance of the triaxial test to the solution of stability problems. Proc. Research Conf. Shear Strength of Cohesive Soils, pp.437-501.

Bostock, H.S. 1948. Physiography of the Canadian Cordillera with special reference to the area north of the 55th parallel. Geological Survey of Canada. Mem. 247.

British Columbia Ministry of Forests. Map: Biogeoclimatic Zones of British Columbia. 1988

Buma, J. and Van Asch, T. In Landslide Recognition: Identification, Movement and Courses. Edited by R.Dikau, D. Brunsden, L. Schrott and M.L. Ibsen. John Wiley and Sons Ltd. pp. 43- 56

Carson M.A. 1977. On the retrogression of landslides in sensitive muddy sediments. Canadian Geotechnical Journal., 14, pp.582-602

Church,M. and Miles, M.J. 1987. Meteorological antecedents to debris flow in southwestern British Columbia; Some case studies. Geological Society of America. Reviews in Engineering Geology, Vol. VII, pp.63-79.

Clague, J.J. and Luternauer, J. 1982. Late Quaternary sedimentary environments, southwestern British Columbia, International Association of Sedimentologists, 11th International Congress, Field Excursion Guidebook 3 OA, 167pp.

Clague, J.J., Saunders I.R., and Roberts, M.C. (1988). Ice-free conditions in southwestern British Columbia at 16 000 years BP. Canadian Journal of Earth Sciences, Vol. 25, No. 6, pp. 938-941

Conlon, R. J. Landslide on the Toulnustouc River, Quebec. Canadian Geotechnical Journal, Vol. 3(3), pp. 113-144.

Craig, R.F., 1997. Soil Mechanics, 6th Edition. Chapman and Hall, London.

125 Cruden, D.M., Keegan, T.R., and Thomson, S. 1993. The landslide dam on the Saddle River nearRycroft, Alberta. Can. Geotech. J. 30: 1003-1015.

Cruden, D.M., Lu, Z-Y., Thomson, S. 1997. The 1939 Montagneuse River landslide, Alberta. Canadian Geotechnical Journal, 34: 799-810.

Cruden, D.M., Lu, Z-Y., Thomson, S. and Weimer, N. 1995. Alberta's largest historic landslide. 48th Can. Geotech. Conf, 2: 893-900.

Cruden, D.M. and Varnes, D.J. 1996. Landslides: Investigation and Mitigation, Special Report 247 Transportation and Research Board. Edited by AK. Turner, R.L. Schuster, pp 36-75

Drinkwater, T.A., Huestis, E.S., Istvanffy, D.E., Klawe, J.J., Laycock, AH. and Wonders, W.C. 1969. Atlas of Alberta, University of Alberta Press, Edmonton, 158p.

Duncan, J.M., Fleming, R.W., and Patton, F.D. 1986. Report of the Thistle Slide Committee to State of Utah Department of Natural Resources. U.S. Geological Survey Open-File Report 86- 505

Evans, S.G. 1982. The development of Big Slide, near Quesnel, British Columbia, between 1953 and 1982. Geoscience Canada. Vol. 9 (4), pp.22-223.

Evans, S.G. 1982. Landslides and surficial deposits in urban areas of British Columbia: A review. Canadian Geotechnical Journal. Vol. 19, pp. 269-288.

Evans, S., Hu, X.Q., Enegren, E.G. 1996. The 1973 Attachie Slide, Peace River Valley, near Fort St. John, B.C., Canada: A landslide with a high-velocity flowslide component in Pleistocene sediments. In Proceedings, 7th International Symposium on Landslides, Trondheim, Norway, pp. 715-720

Evans, S.G. and Brooks, G.R., 1994. An earthflow in sensitive Champlain Sea sediments at Lemieux, Ontario, June 20, 1993 and its impact on the South Nation River. Canadian Geotechnical Journal, Vol. 31, No. 3, p. 384-394.

Fulton, R.J. 1965. Silt deposition in late-glacial lakes of Southern British Columbia. American Journal of Science, 263, pp. 553-570.

Geertsema, M. and J.W. Schwab. 1997. Retrogressive Flowslides in the Terrace-Kitimat Area, British Columbia: From Early Post-deglaciation to Present and Implications for Future Slides. Proc. 11th Vancouver Geotechnical Society Symposium, ppl 15-133

Geertsema, M. 1998. Flowslides in waterlain muds of northwestern British Columbia, Canada. In Proceedings, 8th International I AEG Congress, Vancouver, Canada, pp. 1913-1921

Gongxian, W. and X. Bangdong. 1984. Brief Introduction of Landslides in Loess in China. 4th International Symposium on Landslides, Lausanne. 1, pp.3-35. 126 Hardy, R.M. 1963. The Peace River Highway Bridge - a failure in soft shales. High. Res. Rec. 17:29-39

Hendron, AJ. Jr., and Patton, F.D. 1985. The Vaiont Slide, a geotechnical analysis based in new geological observations of the failure surface (2 vols). Washington, D.C: Department of the Army, U.S. Army Corps of Engineers.

Hicock, S.R. and Armstrong, J.E. 1981. Coquitlam Drift: a pre-Vashon glacial formation in the Fraser Lowland, British Columbia. Canadian Journal of Earth Sciences, 18: 1443-1451.

Hicock, S.R., and Armstrong, J.E. 1985. Vashon Drift: definition of the formation in the Georgia Depression, southwest British Columbia. Canadian Journal of Earth Sciences, 22:748- 757. .

Hicock, S.R., Hobson, K., and Armstrong, J.E. 1982a. Late Pleistocene proboscideans and early Fraser glacial sedimentation in eastern Fraser Lowland, British Columbia. Canadian Journal of Earth Sciences, 19: 899-906.

Hicock, S.R., Hebda, R.J., and Armstrong, J.E. 19826. Lag of the Fraser glacial maximum in the Pacific Northwest: pollen and macrofossil evidence from western Fraser Lowland, British Columbia. Canadian Journal of Earth Sciences, 19:2288-2296.

Hoek, E. 1987. General two-dimensional slope stability analysis. In Analytical and Computational Methods in Engineering Rock Mechanics. Ed. E.T. Brown, pp. 95-114

Holland, S.S. 1964. Landforms of British Columbia, a physiographic outline; B.C. Dept. Mines Petr. Res. Bull. 48, 138p.

Hungr, O., 1997. Slope stability analysis. Keynote paper, Procs., 2nd. Panamerican Symposium on Landslides, Rio de Janeiro, Int. Society for Soil Mechanics and Geotechnical Engineering, 3: 123-136.Hungr, O. In review. Classification of landslides of the flow type.

Hungr, O. 1995. A model for the runout analysis of rapid flow slide, debris flows, and avalanches. Canadian Geotechnical Journal, Vol. 32: 610-623

Hungr, O., Evans, S.G., Bovis, M., and Hutchinson, J.N., Review of the classification of landslide of the flow type. Submitted to Environmental and Engineering Geoscience, June, 2000.

Hungr, O., Salgado, F.M., and Byrne, P.M. 1989. Evaluation of a three-dimensional method of slope stability analysis. Canadian Geotechnical Journal Vol. 26: 679-686

127 Hutchinson, J.N. 1961. A landslide on a thin layer of quick clay at Furre, Central Norway. Geotechnique, Vol. 11:69-93

Hutchinson, J.N. 1968. Field meeting on the coastal landslides of Kent, 1-3 July 1966. Proc. Geol. Assoc., 79:227-237.

Hutchinson, J.N. 1987. Mechanisms producing large displacements in landslides on pre-existing shears. 1st Sino-British Geological Conference, Taipei, Memoir of the Geological Survey of China, No. 9:175-200.

Hutchinson, J.N. and Bhandari, R.L., 1971. Undrained loading, a fundamental mechanism of mudflow and other slope movements. Geotechnique, Vol. 21, pp 353-358.

Hutchinson, J.N., Prior, D.B., and Stephens, N, 1974. Potentially dangerous surges in an Autrim mudslide. Quarterly Journal of Engineering Geology., Vol. 7, pp. 363-376.

Hutchinson, J.N. 1988. General report: Morphological and geotechnical parameters of landslides in relation to geology and hydrogeology. In Proceedings, 5th International Symposium on Landslides, Lausanne, Vol.1 pp. 3-35.

Hutchinson, J.N. 1992. Flow Slides from Natural Slopes and Waste Tips. Ill Simposio Nacional sobre Taludes y Laderas Inestables. Vol. 3:827-841

Hutchinson, J.N. 1995. Rapid landslides: Prediction of initiation and motion. The International Symposium on Prediction of Rapid Landslide Motion. Working Group for Prediction of Rapid Landslide Motion. Kyoto, Japan.

IAEG Commission on Landslides. 1990. Suggested nomenclature for landslides. Bulletin of the International Association of Engineering Geology, No.41,pp.l3-16.

Iverson, R.M. and Major, J.J. 1986. Groundwater seepage vectors and the potential for hillslope failure and debris-flow mobilization. Water Resour. Res. 22:1543-48.

Karlsrud, K., Aas, G, and Gregersen, O. 1984. Can we predict landslide hazards in soft sensitive clays? Summary of Norwegian practice and experiences. In IV International Symposium on Landslides, Lausanne. Canadian Geotechnical Society, Vol. 1 pp. 107-130.

Keefer, D.K. and Johnson, A.M. 1983. Earthflows: Morphology, mobilization and movement. Us.S. Geological Survey Professional Paper 1264, 56p.

Lambe, T.W. 1991. Soil Testing for Engineers. John Wiley & Sons, New York, 165pp.

Lambe, T.W. and Whitman, R.V. 1969. Soil Mechanics. John Wiley & Sons, New York, 553pp.

128 Lu, Z.Y., Cruden, D.M., and Thomson, S. 1998. Landslides and preglacial channels in the Western Peace River Lowland, Alberta. Proceedings 51st Canadian Geotechnical Conference, Canadian Geotechnical Society, pp. 267-274

Lum, K.K.Y. 1979. Stability of the Kamloops silt bluffs. Unpublished Masters of Applied Science Thesis, University of British Columbia, Department of Civil Engineering. 124pp.

Maber, CT. and Stewart, W.O. 1976. The Peace River Hill Landslide. Proceedings, 29th Canadian Geotechnical Conference, Canadian Geotechnical Society, Vol. 4 pp. 35-47

Mathews, W.H. 1978. Quaternary stratigraphy and geomorphology of Charlie Lake (94 A) map- area, British Columbia. Geological Survey of Canada Paper 76-20, 25 p.

Mesri, G. and Cepeda-Diaz, AF. 1986. Residual shear strength of clays and shales. Geotechnique, London, England. 36(2), 269-274.

Meyerhof, G.G. 1957. The mechanism of flowslides in cohesive soils. Geotechnique, 7, pp.41-49 Mitchell, R.J. 1978. Earthflow terrain evaluation in Ontario. Ontario Ministry of Transportation and Communications RR213.

Mitchell, R.J. and A.R. Markell. 1973. Flow sliding in Sensitive Soils. Can. Geotech. J., 11, pp.11-31

Mitchell, R.J. 1978. Earthflow Terrain Evaluation in Ontario. Ontario Ministry of Transportation and Communications RR 213

Monger, J.W.H., 1989. Geology of Hope Map Area. Geological Survey of Canada Map 41- 1989. 1:250 000 scale.

Morgenstern, N.R., 1992. The Evaluations of slope stability - a 25 year perspective. In Stability and Performance of Slopes and Embankments, G.S.P. 31, ASCE, New York, 1,1-26.

Morgenstern, N.R. and Price, V.E. 1965. The analysis of the stability of general slip surfaces. Geotechnique, 15, (1), 79-93.

Mullineaux, D.R., Waldron, H.H., and Rubin, M. 1965. Stratigraphy and chronology of late interglacial and early Vashon glacial time in the Seattle, Washington. United States Geological Survey, Bulletin 1194-0.

Nasmith, H. 1964. Landslides and Pleistocene deposits in the Meikle River valley of northern Alberta. Can. Geotech. J. 1:155-166.

Odenstad, S. 1951. The landslide at Skottorp on the Lidan River. Proceedings of the Royal Swedish Geotechnical Institute 4: 1-38.

129 Quigley, R. 1976. Mineralogy, chemistry and structure of the Penticton and South Thompson silt deposits. Research Report to British Columbia Department of Highways.

Richmond, G. 1989. Halfway River Slide, Internal Memo, BC Hydro. File 1016-1206.0

Rigg, G.B. and Gould, H.R. 1957. Age of Glacier Peak eruption and chronology of post-glacial peat deposits in Washington and surrounding areas. American Journal of Science, 255:341-363.

Sarma, S.K. 1973. Stability analysis of embankments and slopes. Geotechnique, 23, 423-433.

Saunders, I.R., 1985. Late Quaternary Geology and Geomorphology of the Chilliwack River Valley, B.C. Unpublished M.Sc. thesis, University of British Columbia.

Saunders, I.R., Clague, J.J., and Roberts, M.C. 1987. Deglaciation of Chilliwack River valley, British Columbia. Canadian Journal of Earth Sciences, Vol. 24, No. 5, pp. 915-923

Savigny, K.W., Sego, D.C, and Maclnnes, K.L. 1992. The Little Doctor Lake landslide, an example of coseismic reactivation of a landslide in permafrost terrain. In Geotechnique and Natural Hazards, Vancouver Geotechnical Society and Canadian Geotechnical Society, pp. 203- 209

Seed, H.B. and Wilson, S.D. 1967. The Turnagain Heights landslide, Alaska. Journal of the Soil Mmechanics and Foundations Division, Proceedings, American Society of Civil Engineers, 93, SM4:325-53.

Sitar, N., Clough, G.W., Bachus, R.C. 1980 Behaviour of weakly cemented soil slopes under static and seismic loading conditions. United States Geological Survey.

Skempton, A.W., 1952. The Colloidal "Activity of Clay". Proceedings 3rd International Conference on Soil Mechanics and Foundation Engineering. Switzerland. Vol. 1 pp. 57-61

Skempton, A.W., and Brown, J.D. 1961. A landslide in boulder clay at Selset, Yorkshire. Geotechnique 15, pp. 221-242

Skempton, A.W., and J.N. Hutchinson. 1969. Stability of Natural Slopes and Embankment Foundations.In Proc, Seventh International Conference on Soil Mechanics and Foundation Engineering, pp. 291-340

Skempton, A.W., 1985. Residual strength of clays in landslides, folded strata and the laboratory. Geotechnique 35, No. 1, 3-18.

Stanton, R.B. 1898. The great landslides on the Canadian Pacific Railway in British Columbia. Proc. Institution of Civil Engineers. Vol. 132, 1-20.

Stark, T.D. and Eid, H.T., 1994. Residual strength of cohesive soils. Journal of Geotechnical Engineering, ASCE. Vol.120, No.5 130 Stott, D.F., 1982. Lower Cretaceous Fort St. John Group and Upper Cretaceous Dunvegan Formation o fhte Foothills and Plains of Alberta, British Columbia, District of Mackenzie and Yukon Territory. Geological Survey of Canada, Bulletin 328, p. 13

Stumpf, A.J., Broster, B.E., Levson, V.M., Geertsema, M. and Schwab, J.W. 1998. Stability of glacial silt and clay deposits in central British Columbia. Proc. 8th Congress of the International Association for Engineering Geology and the Environment. Vol. 3:1897-1912.

Terzaghi, K., and Peck, R.B., 1948. Soil Mechanics in Engineering Practice, 1st ed. John Wiley and Sons, New York.

Terzaghi, K. 1950. Mechanisms of landslide. In Application of geology to engineering practice, Geological Society of America, pp.82-123.

Thomson, B., 1991. B.C. Ministry of Environment, Fisheries branch. File No. 0965.

Thomson, B., 1999. Slope failures - Chilliwack River drainage basin: A mass movement classroom. 8th Congress of the International Association for Engineering geology and the Environment. 2:1097-1101

Thomson, S. and Mekechuk, J. 1982. A landslide in glacial lake clays in central British Columbia. Canadian Geotechnical Journal. Vol. 19: 296-306.

Thomson, S. and Morgenstern, N.R., 1977. Factors affecting distribution of landslides along rivers in Southern Alberta. Canadian Geotechnical Journal, 14:508-523

Thurber Consultants Ltd., 1973. Inspection of the Attachie Landslide. Report to the Water Resources Services, Department of Lands, Forests and Water Resources, British Columbia.

Thurber Consultants Ltd., 1981. A report on field inspection of overburden landslides. Report to B.C. Hydro & Power Authority. File 15-2-169-A

Thurber Consultants Ltd., 1982. Feasibility of stabilizing the Attachie slide and adjacent areas, Site C Reservoir. Report to B.C. Hydro & Power Authority. File 15-2-169

Thurber Engineering Ltd. 1997. Geotechnical assessment of Slesse Park landslide. Report to Department of Fisheries and Oceans, Vancouver, Canada

Watson, A. 1999. Stratigraphy at Slesse Park Clayslide, and system for monitoring movement. Unpublished BASc thesis, University of British Columbia, Vancouver, Canada

Varnes, D.J. 1978. Slope movement types and processes. In Landslides - Analysis and Control. Edited by R.L. Schuster and R.J. Krizek. Special Report of the National Research Council and Transportation Research Board, 176:11-33.

131 APPENDIX I

Slesse Park landslide Table AI. 1 Slope displacements at Slesse Park landslide July 1997 to September 1999.

Station Slope Displacement (m) Total displacement (m) July 97 April 98 June 98 Sept. 98 Dec. 98 Mar. 99 July 99 Sept. 99

A1-A2 12.8 _ 12.9 12.9 12.9 12.9 12.9 12.9 0.1 A2-A3 22.8 23.6 23.6 23.7 23.7 25.3 25.3 25.3 2.5

B1-B2 57.8 58.2 NR 57.9 58.0 58.0 NR 58.0 0.1 B2-B3 48.3 48.5 NR 48.2 48.2 48.2 48.2 48.2 0 B3-B4 26.4 26.4 NR 26.5 26.5 26.5 26.5 26.5 0.1 B4-B5 48.3 50.0 NR 48.3 50.1 48.3 48.3 48.3 0

W1-W2 20.9 20.5 20.5 20.5 20.5 21.9 21.9 21.9 0.9 W2-W3 31.4 31.1 31.3 31.3 31.3 31.3 31.0 31.0 -0.3 W3-W4 20.6 20.9 21.0 20.9 20.9 21.1 21.1 21.1 0.5 W4-W5 31.8 32.0 32.0 31.9 37.1 32.0 31.9 '• 31.9 0.1 W5-W6 35.7 35.4 35.5 35.3 29.8 34.9 34.9 34.9 -0.8 W6-W7 8.4 8.6 19.5 8.7 8.7 W7 destroyed 0.2 W7-W8 28.8 28.6 28.6 28.6 28.6 -0.1 W6-W8 37.2 37.3 37.3 37.3 37.5 37.5 37.5 0.3

Note: Numbers in italics are suspected to be errors in data recording NR measurements not recorded Air photo interpretation

1940 (BC209;59,60)

The oldest airphotos available, from 1940, show a well developed back and western lateral scarp in approximately the same location as they are today. It may be concluded that failure of the main body of the slide occurred prior to that time.

These photos also show that an unstable area extended further to the west, beyond the lateral margin of the currently active area. Evidence is provided by a lobe of colluvium that appears to have blocked a channel of the river, resulting in the river shifting to the south. The colluvium is revegetated suggesting that the failure occurred several years prior to 1940. This failure can be

evidenced in the field today by a hummocky colluvial fan protruding into the river. A steep scarp

(station 13) in which sheared, discontinuously laminated clay can be observed, is located at the

head of the colluvial fan and is believed to be the source of this failure.

1952 (BC1623;11,12)

What is most notable in airphotographs from 1952 is relatively recent activity on the western

portion of the landslide. An area of approximately 4 ha, one third of the total landslide area, is

almost entirely devoid of vegetation. It is not possible to determine the maximum displacements,

rate of movement, or wether slope failure was surficial, or exploited the main rupture surface.

However, it appears that some material was displaced over the bluffs and reached the river. An

island of trees remains at the base of the slide area, at the top of the bluffs at an elevation lower

than the ground surface on either side. Because the ground surface is freshly exposed above this

clump of trees and material appears to have flowed over the bluffs below, the vegetation appears

134 to cover a block of intact material which has been displaced from upslope. Although the 1940 air photographs show this area to be covered by vegetation, a lateral scarp is quite well defined and the area appears to be depressed suggesting that this area was subject to mass wasting prior to

1940.

Besides activity of the main instability, active erosion of the bluffs and development of a flow bowl where West and East Twin Creek merge is evidenced by a lack of vegetation and fresh soil exposures.

1979 (BC79005;170,171)

Most notable in these photographs is a linear ridge feature that parallels the back scarp 40m downslope in the vicinity of Central Creek. This ridge appears to be approximately 150m long.

There appears to be continuing disturbance within the 1952 landslide area as well as fresh

exposures along the eastern half of the bluffs, and erosion of a flow bowl at the base of East Twin

Creek.

What is also important to note in these larger scale photographs, taken during the winter when the

leaves are off the deciduous trees, is that it appears that the back scarp extends some 200m

beyond the western lateral margin to Nonie Creek. There is a horst and graben feature below the

back scarp just to the west of the western margin of the 1952 slide.

1997 (McElhanneyl6366-0, R332, 1-10, 1-11)

135 Air photographs were taken in March, 1997 to document the earthflows occurring in the colluvial apron at the base of the bluffs. An earthflow occurred January, 1997, partially blocking the river, in material that appeared to have been stable in all previous air photos. There are multiple fresh exposures along the bluffs.

By 1997, a large flow bowl had developed at Central Creek, extending as far back as to intersect the ridge feature described in photos from 1979.

The island of vegetation noted in the 1952 air photos appears in the 1997 airphotos as coniferous trees, more mature than adjacent vegetation. There is continued shallow sliding and slumping in the disturbed 1952 landslide area. Distortion of photos means that it is difficult to see in stereo.

136 Photogrammetry

Figures All through AI.4 are the result of photogrametric mapping of principal landslide features.

From the mapping of back scarp it would appear that as much as 22m of retrogression has

occurred since 1940. However, the accuracy of the mapping is only as good as the interpretation

by the user as to the location of the back scarp on the photo. In the older air photos, the back

scarp is not as clearly defined. It is possible that this is not only due to the quality and scale of the

photo, but that the scarp may not have been as well developed. The western lateral margin has

not shifted since 1940 whereas most recently there was an earthflow in material that had

remained vegetated and seemingly undisturbed since at least 1940. This earthflow extended what

previously had appeared to have been the eastern limit of disturbance by approximately 60m. It

does not appear that the bluffs have retrogressed significantly apart from the development of a

large flow bowl at Central Creek. This does not seem accurate, as colluvium eroded from the

crest of the bluffs has backfilled the berm and developed an apron along the base of the bluffs.

This erosion must have resulted in retrogression of the crest of the bluffs.

It was intended that from transects taken through the 1952 landslide area, a profile of the pre

(1940) and post (1952) failure ground surface could be interpreted. From the pre and post failure

surface profile, one may be able to determine whether the landslide was a deep seated failure

exploiting the main failure surface, or a series of more shallow slumps and flows. However, due

to the poor quality of the air photos from 1940 and 1952, possibly taken with a fish eye lens, there was distortion of the images that could not be eliminated by the stereoplotter. As a result, elevations vary as much as 110m at a location that, according to the air photos, does not appear to have been displaced. Therefore, the transects were considered unreliable.

Photogrammetry indicates that there has been a shift in the course of the river since 1940. A slump in the bluffs downstream of the current landslide activity blocked a channel of the river, forcing the river to shift course further south, resulting in a sharpening of the bend and an island left between the former channel and the current channel. By 1952 the ground surface of the former channel had been completely vegetated and by 1979 had been developed with houses.

Since then the river has shifted to the south and west, eroding part of the colluvium that had blocked the former channel. The channel is slowly migrating towards its former course.

As well as the river having obviously shifted over time, so have the creeks flowing over the landslide area. Central creek flowed slightly east of its current position in photos from 1940 and

1952. West Twin Creek does not appear to have incised a gully other than what has occupied in older air photos. There did not appear a stream flowing through the 1952 landslide area prior to

1997. Again, the location of these features is only as accurate as the person mapping interprets of the location on the air photos.

138

Figure AI.2 Pho-togramnetrlc napping of\Slesse landslide - 1952 Figure AI.3 Photogrannetrlc napping of Slesse landslide - 1979

Slide scar

(41 Figure AL4 Pho-togranne-trlc napping of Slesse Landsllde-1997 * tN * * 00 * IT) \ * * VO eg * rH + * rH * o rJ -* rH CJ * * r-t m x + * + CC * CO + CD Q X CO + * r> M + 2 Cu * rH CO +* 2 Q EH * O O + M rH H r- rn 0> H i/i * < •J + in oj in co oj in * E-t CO CJ CO + tr W (J> N T r- in oi vo o m E- CJ + rH rH rH H rH OJ OJ O 2 + I I I I 1 I I I I 1 I I * 2 CO + > \ \ \ \ >v. \ -v. ^v. \ \ \ \ * a + -+• + •+- + + + 4- -+- 4- + + CO M + a< ^ CD OJ iD M O rH r- o & r- rH rH CTl \£) CC Cu CM m o in r- co 00 fO 6- * 2 * in 2 * o + CM VD Cfi r~ CD H co cn CO U5 Lf) CO * CJ + N i/l h O W T 1— rH T OO •—i T CJ + H W H N CM C-J ^H oj oj OJ m m <#> + CO + o -r + m p~ *r T m r^ if) M r LO I—4 v w m o r- v X< + X in ro ^ o m r- EH CT\ in rH 0\ -rr n + + - •J I—I + t"0 in r~- c ' i—i n rH oj OJ rn CO CJ CJ -* * + < + + Q< T in M ifi c\ ro ^? -3- m o o a M< 00 o o o a vD T r- <» r~. •H O rH o >i) o~> < a > * i-i co r- o r- oj cr> o cr\ co r- ro ro vD ro oj rH u < + + r-t m un oo cr» VD o *3- r— o ro o co co o in o rH U + + rH oj c-j oi ro i"0 ro ro T in tn ^ u< + + > LJ CJ * + T m r- h in CO a + CO O * T •—I r0 KD CJ> C-! oi M m in r- o"i rH rn in CA oi VD CA ro Cu CO + rH rH rH OJ CJ M + * I I I I I I i i i i > * * >- \ \ \ \ ^s. \. 2 2 M + * + -r H- -r 4- 4- 4- 4- 4- 4- 4- o o a * a< oo in in OJ co ID \D r- ro in *r r- cc M cn m co OJ in io rH in o m M E- * + m u * .—. in rn CNJ O cr> r~ r- o CA rH vo OJ 2> CC< o m o io r- T> oi u? -a- oo «n CA T CJ rJ Si + rH OJ Cv] OJ OJ ro oi m m o c- CC >- CJ * CO -I co r- T -*r O in ifi in ^ CJ E-* * CJ + X w CJ * 3 -r r> io r- co ro Oi CO + eg -J + O! tr W CO O OJ rH rH Ol OJ CO V) o «-t a! + I I I I I 1 O CJ a: + >- fn Q X > < X M T H a + o< a H r- H co co in in 7 * H? M o O 05 ("O rH r- oi T i—I T OJ IO (D < +* O rH + t-J CJ + CC in r- o OJ "X> in CA \D o vo r- CO •J X + LO + T co ri in r- o n o in o T co EH Cu OJ CM m ro T < +* o + >- 2 •—I i—I OJ CN OJ rn M CQ + a CC + "7 CTl '.0 Ol CO CTl rH VD rH O OJ OJ OJ CC< Q id E- T in oi i— T C-J CJ + CJ + u) H m o T co CC U CC > o i~n *T MD r~- •H rH OJ Ol OJ + X + I I I I 1 I Cu< + tH + >• >- CJ < CC O + a< r-< o rH (7« r~- T OD in o -T vo a r- r- r— T OJ +* + in r— in O T in C-.J CM co co Cu E- O C-J vi? iO f,t

H + ci o in i— —• -^r oo T CA ro o •* io r- CA c-i in r- * + X -r * E- 4 Lj 2 OJ O .c 1 * + u + 4-J Cu + + OJ o + t-J X + < < * (•-* + Cu * • 4 2 rH + M * + < +* (J cc 03 oc * 3: + •rH Q 3 rH o s: rH rH Q a rH +* •rH CC + O O o xz CO + >-H < * u Cb + CC s: * + CJ * -* Cu CJ -X 4 -J * 2 1 cc * rH E- c-i in o in o -3 r> E— O CC Cu LJ CC CC * o Chilliwack River Hatchery Total Precipitation (mm)

Year Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

1984 Msng 161.2 132.9 93 173.5 79.4 13.7 46.8 103 159.7 244 176.2 1985 45.2 85.2 66.6 113.6 92.3 82 4:8 22.8 92.8 345.2 133.4 44

1986 232 216.1 137.2 101.7 114.6 72.6 88.1 3.1 75.7 145.9 343.2 123 1987 225.9 58.5 134.9 150.9 120.8 41.5 99.2 14.6 47 15.6 124.7 191.6 1988 168.9 108 180.1 179.7 119.9 55.3 51.8 33.2 118.6 152.1 263.1 144.6 1989 Msng 30.7 185.9 87.8 94.2 72.9 32.2 92.2 21.6 130.4 376.2 191.1 1990 258.2 237.4 79.8 63.9 91.3 125.9 46.7 108.2 28.2 271.3 496.6 228.5

1991 186.8 225.7 136.4 166.2 100.8 74.6 49.2 150.2 16.2 42.4 255.2 138.5 1992 316.8 102 26.7 161.3 44.8 86.4 96 76.4 115.6 107.3 207.9 117.4 1993 145.1 10.5 153.3 130.8 121 146.5 86.9 38.3 21.5 119.3 114 190.1 1994 119.8 204.3 135.5 73.2 56.6 109.8 39.4 39.3 74.4 139.2 166.3 261.2 1995 143.2 175.2 132.1 67.1 21 61.2 74.3 56.8 44.8 207.4 419.8 189.9

1996 244.4 155.6 71.2 172.5 127.3 36.2 43.4 43.4 105.5 232.4 241.9 293.2 1997 316.9 103.2 260.3 139.3 91.7 143.4 97.6 28 173.7 183.5 116.5 175.8 1998 210.6 104.5 102.8 57.1 Msng 97.6 47.1 Msng 31 Msng Msng Msng

Max 316.8 237.4 260.3 179.7 173.5 146.5 99.2 150.2 173.7 345.2 496.6 293.2 Year 1992 1990 1997 1988 1984 1993 1987 1991 1997 1985 1990 1996 Mean 191.4 131.9 129 117.2 97.8 85.7 58 53.8 71.3 160.8 250.2 176.1 Min 45.2 10.5 26.7 57.1 21 36.2 4.8 3.1 16.2 15.6 114 44 Year 1985 1993 1992 1998 1995 1996 1985 1986 1991 1987 1993 1985 StDev 72.4 71.3 56.5 42.9 38.2 33.6 30.6 40.2 46.4 85.7 120.6 62.4

Chilliwack River Hatchery Total Rainfall (mm)

Year Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

1984 Msng 160.2 132.9 93 173.5 79.4 13.7 46.8 103 151.8 224.9 112.9 1985 40.2 57.2 64.1 113.6 92.3 82 4.8 22.8 92.8 345.2 114.6 21.7

1986 230 133.1 137.2 97.7 114.6 72.6 88.1 3.1 75.7 145.9 338.2 123 1987 223.9 56.5 127.4 150.9 120.8 41.5 99.2 14.6 47 15.6 124.7 154.6 1988 147.8 94.5 180.1 176.2 119.9 55.3 51.8 33.2 118.6 152.1 259.7 129.8 1989 Msng 18.7 185.9 87.3 94.2 72.9 32.2 92.2 21.6 130.4 376.2 191.1 1990 226.7 137.8 71.8 63.9 91.3 125.9 46.7 108.2 28.2 271.3 488.1 166.5

1991 117.3 224.7 75.4 166.2 100.8 74.6 49.2 150.2 16.2 42.4 254.2 138.5 1992 315.8 96.5 26.7 160.8 44.8 86.4 96 76.4 115.6 107.3 207.9 57.9 1993 98.6 6 153.3 130.8 121 146.5 86.9 38.3 21.5 119.3 110 189.6 1994 119.8 167.8, 133.5 73.2 56.6 109.8 39.4 39.3 74.4 139.2 135.3 246 1995 143.2 159.2 116.1 67.1 21 61.2 74.3 56.8 44.8 207.4 417.8 183.4

1996 186.4 147.6 60.2 172.5 127.3 36.2 43.4 43.4 105.5 232.4 161.4 168.2 1997 Msng 103.2 240.3 139.3 91.7 143.4 97.6 28 173.7 183.5 116.5 171.8 1998 146.7 102.5 102.8 57.1 Msng 97.6 47.1 Msng 31 Msng Msng Msng

Max 315.8 224.7 240.3 176.2 173.5 146.5 99.2 150.2 173.7 345.2 488.1 246 Year 1992 1991 1997 1988 1984 1993 1987 1991 1997 1985 1990 1994 Mean 166.4 111 120.5 116.6 97.8 85.7 58 53.8 71.3 160.3 237.8 146.8 Min 40.2 6 26.7 57.1 21 36.2 4.8 3.1 16.2 15.6 110 21.7 Year 1985 1993 1992 1998 1995 1996 1985 1986 1991 1987 1993 1985 StDev 73.9 59.4 56.2 42.6 38.2 33.6 30.6 40.2 46.4 85.7 124.5 56.9 Chilliwack River Hatchery Total Snowfall (cm)

Year Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

1984 Msng 1 0 0 0 0 0 0 0 7.9 19.1 63.3 1985 5 28 2.5 Trace 0 0 0 0 0 0 18.8 22.3

1986 2 83 0 4 0 0 0 0 0 0 5 Trace 1987 2 2 7.5 0 0 0 0 0 0 0 0 37 1988 21.1 13.5 Trace 3.5 0 0 0 0 0 0 3.4 14.8 1989 Msng 12 Trace 0.5 0 0 0 0 0 0 0 Trace 1990 31.5 99.6 8 0 0 0 0 0 0 0 8.5 62

1991 69.5 1 61 0 0 0 0 0 0 Trace 1 0 1992 1 5.5 0 0.5 0 0 0 0 0 0 0 59.5 1993 46.5 4.5 Trace 0 0 0 0 0 0 0 4 0.5 1994 0 36.5 2 0 0 0' 0 0 0 Trace 31 15.2 1995 Trace 16 16 0 0 0 0 0 0 0 2 6.5

1996 58 8 11 0 0 0 0 0 0 0 80.5 135 1997 Msng 0 20 0 0 0 0 0 0 0 0 4 1998 63.9 2 Trace 0 Msng 0 0 Msng 0 Msng Msng Msng

Max 69.5 99.6 61 4 0 0 0 0 0 7.9 80.5 135 Year 1991 1990 1991 1986 1984* 1984* 1984* 1984* 1984 * 1984 1996 1996 Mean 25 20.8 8.5 0.6 0 0 0 0 0 0.6 12.4 30 Min 0 0 0 0 0 0 0 0 0 0 0 0 Year 1994 1997 1984 '1984 *1984* 1984* 1984* 1984* 1984 * 1985 * 1987 * 1991 StDev 27.6 30.6 15.9 1.3 0 0 0 0 0 2.1 21.7 38.6 a * to the right of a year indicates the first of several occurrences 146 Table AI.2 Classificatio n of soils sampled from Slesse Park landslide O >. ra Sample No. Location of Sample Description Unified Liquid Plastic Plasticity Liquidity Activity _i '-i '-i Soil Class. Content % E E Index Index

In situ laminateid sediments CO ci CO o 0 CM LO Ol CO si 8 CO to o I Thurber 1 directly above Grey, moist, high-plastic sheared surface clay with some silt 5M 8 ci CM o to 8 CO CM CO CO CO O _l Thurber 2 directly below Grey, dry clay, some silt sheared surface Ci LO r- CM CO CM CO o X o Thurber 5 f-c sand grades up into Light gry, moist to dry, silty clay, trace of fine sand

sand and silty clay I it I? i 0. CO Thurber 6 below 5 Grey, moist sand with trace of silt i-» CM CO ci 5 9 CM to o _l Thurber 7 below 6, v. stiff silty clay Grey, dry clay with some silty clay, trace of fine sand CM m CO o ci LO LO LO to o X Thurber'91 grab sample grey silty clay; hard, moist homogeneous iri Ol co i-» csi o oi Golder 2 backscarp layered silty sand and gravel CL-ML -0.58 o o r- CM CM "

Soil Class. Content % Limit Limit Index Index u u

Flowslide collu' tun o SL10-3 Flowslide colluvium, bottom of CC g sandy, gravelly mud CL (fines) 22.0 23.0 16.0 0.86

Slesse 6 Flowslide colluvium, Central C. silty clay with fine sand CL 13.2 27.0 17.0 10.0 -0.38

Colluvium in main scarp r- CM Thurber 8 colluvium filled gully Grey, moist silt, some ML 26.8 0.97 gravel, traces of sand and clay CO co 9-22-01 Till from Central Creek very stiff diamicton CL-ML (fines) 43.5 12.9 22.5 15.7 -0.41 0.16 Table A1.3 Specific gravity of soil particles. Micromeritics Multivolume Pycnometer 1305

Sample name Attachie P1 (psi) P2 (psi) Vs (cc) mass of container (g) 22.1221 19.584 10.348 7.955716327 mass of container + sample (g) 44.2095 19.502 10.33 7.832806149

Vcell (cc) 31.474 19.546 10.352 7.839147705

Vexp (cc) 20.991

mass of sample (g) 22.0874

Average Vs (cc) 7.876

Density (g/cc) 2.804

Sample name Slesse P1 (psi) P2 (psi) Vs (cc) mass of container (g) 22.1221 19.581 10.706 6.152349746 mass of container + sample (g) 39.5884 19.583 10.699 6.194541535

Vcell (cc) ' 31.474 19.574 10.693 6.20018365

Vexp (cc) 20.991

mass of sample (g) 17.4663

Average Vs (cc) 6.182

Density (g/cc) 2.825 (sjunoo) un CD 03

O O _>> D) i CM CD CO CO 0 •1 CO

— x

a 2 2 8 (siunoo) up 151 APPENDIX H

Attachie landslide Sunmaary of BC Hydro Borehole 63-2

Moisture content V. 20 30 40 Elev. 479m

30-

40-1

50-

60

70 K>WvH Shale bedrock t>n°n°n<1 Cobble/gravel I 1 Laminated clay and silt 80-

^ Silt

Efrfrl-IH Sand \\,;JV\ Till 11=11=11=1 Slide debris 1 Summary of BC Hydro Borehole 63-3

Moisture content 7. Elev. 654m •-==- 20 30 40

10-

20

30-;';.; ':•

40-•••

60 F-.T

KXX>Cx>1 Shale bedrock Cobble/gravel I 1 Laminated clay and silt fggggg sat E^**i>1 Sand ] Till Slide debris Siimimary of BC Hydro Borehole 63-4

Moisture con-tent Z Elev. 663m 80 30 40 50

* *

20H A. . *.• •

3(H

K»^»1 Shale bedrock 40H EggS?S4 Cobble/gravel I 1 Laminated clay and silt sat 3 Sand so- I.W>1 Till reoei suae debris so- Summary of BC Hydro Borehole 63-5

Moisture content 'JL Elev. 610.9m

KXXXrVI Sbalebedrock

ro°n°o^ Cobble/gravel I 1 T aminatfid clay and silt

Eggogg sat KftfrVfr, Sand |.W v-l Till fieOO Slide debris 60 Summary of BC Hydro Borehole 63-6

Elev. 538m

Shale bedrock

b^o^q Cobble/gravel

r - — — ] Laminated clay and silt sat K%frHK1 Sand

\-:,;S-Y\ Till

tOOO Slide debris

/57 Siammmry of BC Hydro Borehole 63-7

Moisture content '/. Elev.

rsrXXXrt Shale bedrock

Pn°o°nq Cobble/gravel l 1 Laminated clay and silt

RK^XHX^ Silt K&^HK1 Sand l.'-.-^-y • v H Till Bm Slide debris

15b Summary of BC Hydro Borehole 63-8

Elev. 6562m

10J

2o-i:

304:

4LH'

\^>y>M Shalebedrock Ktttt Cobble/gravel 1 Laminated clay and aih •I Silt ISff^rffl Sand 3 Till fOffemi Slide debris 80H 3£j

3 fc)

J c O is O IS

I 2 oa •f .3 -S c ••3 -° e E •a o o Z * O.

?! i •a "> N 5 = £

•8 8 8. •a 8. a 8.5 a 1/1 c E s e a cs ° O TJ w S _ o u ,E o .5 CS _ to <3 < o n . •a u 00 —J 'JZ t- •rf •rf G\ dd* •a-a •u i S « O M s fl 00 C\ CO o E <-« > en 5 2 >, § 7) >i >> < ITJ f > O) o >- Si V£> — P, ° > -5 > Z d 2 ^ j 1 *u Si §•8 S 2 2 E i & O Cr) o 2 £ E 2 i CQ < ll 5 s 6 fe U '5 2 (S £ 1C Q '-5 00 -C o « oa E

© E «3 CS

•o5.

3 CS 31. 6 32. 0 31. 1 |

« c 1430. 7 827. 4 620. 6

?« 1703. 8 1956. 8 871. 5 8s 16. 0 18. 3 12. 8 A t max . deviato r stres s 'S1 0.29 6

T T 7 O O O 2.74 x 4.57 x r-i 1954. 7 1572. 1 991. 5 |

o o 125 0 | 200 0 !

tio n OV C~ cs rt 1017. 0 | 1585. 9 | 2344. 3 i Consolid a b ^

o VO to 00 tri VO

_ vo 128. 7 113. 8 j CO 99. 7 o o 25. 5 24. 5 | 27. 2 75 4> •9 e ss 1.80 2 1.65 8 1.67 7 1 En d o f Tes t

vo VO 00 •s. Cv o\ ss 00 00 o\ 95. 8 95. 8 96. 8 I o 00 30. 7 31. 8 •o 30.7 M

>0 .B2 46.3-46. 9 46.3-46.9 46.3-46. 9 Dept h (m ) cs .S 2 vo 63- 5 63- 5 Hol e Dril l "63- 5 73 '§•'3 R 3 Tes t No . oi 2 O H twn a e

is

a bi

IS ^

•ss

CS

a. c\

I o 00 ca en e 3 H CS ca -«o* CS u u s Z 5 U E CS u c cs a 2 u is g O ii u 1_ OS eu a. <2< 2 eu Ml s u < H oo "ci 8 CO CO CO CO CO r- CO CN CN in co •- in O) CO CO in h- CO cn • CO CN IO CN d CM csi N* oi ci d co' co* in oi d I-* ih oi CO CN CO *-* CN co in CO co O) CM o CN CN T o N- O) oi Q co CM 8 CO CM CO 9 CO CO •

•vT co CN in CO r- T CO l- •vT •vT co «- CO CM co CO co co CM CN co CN CO CO ,8 co' CO d cd CO •vT I-* CO oi •vT co' oi •vT —i iri iri T' CN iri r-' d oi CO co' iri iri Q —i T— CN CM CSI in CM CN •vT CM CM CN CO CO in •vT in CO •vT co CO CM •vT CN CM •* g ri - ^

•if vr CM i- in co in cp -vT in CO CO •vT CO co CO h- CO CO »— CO CN •vT CO in T r- co ,- o> CN d oi •vT CO •vT CO •vT oi iri co iri •vT oi iri d CO r~' CO* CN co CM CN in CO CN CM " 5 CN CO CN ? in CO CN •vT CN in 9 CM

•vT m CN m CN CN •vT CM •vT CN CN CN CO o> CO •vT CO CO CO cn in CN co cn co in co iri co* CO CO iri CM CO oi d CM •vT* oi m* CO CN* •vT csi I--* r-* ic-o co iri co oi iri iri CM CN CM T— •vT CN •vT co in CM •vT CN CO CN CN CM CO f- CO T- CN co >-

Q. 01 •vT •vT •vT CO co CO Ol CO in CO CO CN cn CO CO T co CO in cn N* CO co CO CO* ci co' oi iri d ^_ •vT* CM* CO i»* csi CO* d CM* co* iri W CO co "—C O •vT CN CN CO CM •vT CO •vT CN co co ? m co i- in

cn CO co cn CM •vT CO •T CO f- •vT co CO cn co in CO r» •vT v- t>- CSI CN m in CN CO O) •vT CO CN CM cb CO •T CO csi oi cb co CO* CO oi oo' to cb -vr csi co co* f— CO d r-* CO < IO co CM r» CO CM in CO CN CO 9 in CO co co CO CN o CN CN CO in

CO CO ,- CO CN in CO r~ •vT cnc o ,- co in •vT CO ^- co •vT co CO •vT Ol co to CO CN CN cb •vT* cri CO co csi •vr* CO* r-* csi •vT d d iri iri •vT r»* oi csi co* oi •vT* csi N* csi cb iri oi CO CO in in o CN CN 58 o co *" co CO CO co CM •vT CO t- co t~- r» CN CO 00 i CO CM •vT CM CO

CO •vT N. Ol r-- r~- m Ol co CSI in in r- CO co CO CO co cnr - co oo in CO CN co co in Ol in CO CM CO oi r»* CO csi CO* oi CO* CO csi CN* CO CO m •vT 1-^ csi in co i»* iri cb CN CO CO CO d CO iri Ol m 9 •vT CN •vT •vT CO CM CN in in CO co l>- CO •VT O CM *~ "

CO cnC M CO Ol co CO •vT CN t- Ol Ol •vT co co Ol CO CN CN CN CN Ol m CO •vT CO O) CO oi CO r»* CM oi co* csi iri CO r-' CO CO* O) in CO CO CO* 1^ CO CN cb d csi co in in -VT CO CN CO CN CO CN CN •vT CN CN CN CO CM CM CN ?

co cnc o CM CM CO CM in Ol in co i^. CO in co •vT r- in CN cp CN co CN in to CO co •vT iri •vT* CO* ico- 00* CO •vT* iri d CO csi csi csi d CN CO* CO CO d CN* iri cb d iri cb •— CO CN CN CO to 3 CN Ol CN CO CM co in CN

^ oo in in in •VT IN co CO in CN in CO CO o CO CO 00 •«T in 01 •vT co CN r- •vT l^ CO to oo csi CM r>* d csi d co CN oi d co oi -VT iri r-»* iri CO CO CO oi d cb CO* cb to cb csi m in •vT CO to •vT CN CN m CM 4 CN CN CO T

1*^ •vT r» CO CO in co Ol m CN h- in tn co o> co in m m •VT in CO •vT in oo CM CO co* CM tri •vT* oi CO* oi oi i--* i--* oi co •vT i>-' oi CO CO d r» d CO iri co co cb CO CO oi in iri CM •VT co CO co co CO CO in CN •vT CM CN CN

2 cn CO m CO in •vT CN CO CO CO oo CO CO CO co l-» h- CO CO Ol CN in CO CN co CO i-~ CO* i^ iri CM* CO Cs1i- CO* co csi co co CO |vi •vT* in CO oi Csi CO cb iri d d co •vT CO CN r- in in CO 4 CN CM 5? CO CO CO 9 9 CN CM ?

(MnvjmiONtooiOv- cNco-vTintor-cooi (NnfiniONcooiOv- CN co T T-vT'vTT'vT'vT^'vTmminininminininm m to lOIOIOlOlOlOCOtDNNSNS C^010)0101O101OlOlC^cnO)OlcncncnOlOlO101O^0)01OS )01C^Cn01C7)0101OlOl

/63 TT CN CO CO CO r- CO CO oo CO r~ o TT CO in Tf co CO uo Tf cb uo" oi _ uo "> ib CM cb cb to CO CO uo cb d CM T- Ol CO TT uo o CO Tf CO CO co io Tf CN CO Ol Tf co in TT CO Tf uo CO CO uo 5! uo CO 3 Tf r-

r- r- uo Tf r- O) CO uo CO cn CM cn o r- CO CN uo Tf CD UO uo' CN cb Tf uo cb TJ- ib d H in 00 CM CN CN T? m CO CM Ol

CN TJ- CO O) CO Tf r- CM cn co CO oo r- 01 O CO CO co uo cd id oi CM CM cb d cb uo CO CM CN ob Tf cb CO »- CM CM CM CM CO CN CO UO UO CM co CM r~

CM uo r- co cn TJ" r- co O) r- Ol O CO CM CM uo CN TJ- uo CO cb cb co oi CO cri CO cb d ob oi UO TT CO CM CN CM uo UO Tf Tf CM CO CO T" T- CN *f CN

N r- [O N IO a> CM 00 00 00 TJ oo CM CO oo o co co Tf CO CO uo CM cb vi cb cb 00 CO UO (o d CN UO CO CO CO r~-

CO

CO T- 00 CN r- uo 00 CN CO cn cn Tf Ol CN O CO cn r- S CN ^ co• uo CM CN d d cb Tf Tf TT cb cb UO r~' cb d uo CO UO co uo r- o CM CO Tf T— io CO CO CN Tf CO co CM

CO UO CN UO r- Tf Ol 00 in co co co co uo T— O CO Ol 00 cb CM o ai co o CM uo uo CN cb uo cb Tf cb cb 00 "1 in cb oi uo N f s CM in cn r- TT CN r- uo r- CO CN CO Tf CN CD E *—

r~ uo T TJ- OO CN CO Tf CN CO CO co cn CO co o co co Tf CO CO r r «i d CO UO oo UO cb d d CO Tf 10 r- oi cb o TT o Ol 5? uo o CD CM CM CM CN ro CO T- UO CO T— c I O c uo CO CO CM uo o 01 oo co _ CO Ol co CO

CO CN f- Tf CO cn CN CO CN CN O UO eo CM CO TT O UO cb cb TT" CO CN cb d CO ~ w d d CO T— UO T~ CO co CN Ol uo UO

U UO

CN CO CO Tf I— O) CN CN CO oo CO CO CO r- CO uo CO O oo CO UO d Tf CD CM CO CM ai CO r~ cb CM oi cb cb cb TT UO uo ai uo CO CN T— CN CM CO CO CM CN co uo CM " i ro CO CO CO TJ- TJ-CO UO co. CN co co CN a> CM uo co CM Oi UO OO co co 13 o d TT cb cb uo TT cb CM cb cb uo CO TT' UO ^ N* Tf Q T- r- LL »- CN CO CN CO CM CO CO CM a; X 13 J

O CN CN CO r~- CO uo co CN Tf uo CO CO CM Tt CO TT Tf O CO Tf r- UO oi uo N uo TT ai cb Tf d d d TJ" CM cb O ^ CM d cb cb ™ y o to CN CM TT CO co r- CO 01 ro CN —T T- c < I ro ra c O X > I U0CDr~-CO0iOT-CMCOTfu0C0r-CO01OT- o SNNNNCOCOtOCOeOCOeOCOCOCOOlO) oioioioioioioioioioioioioioioioiai —>O 35 CO h-co ro 01 W -5. r- 2 - "8 s > is i s x 29 CD o 1(0*1 ro Z CT)CO CN CO •vr CN cn9 CO in CM •vr i^ O) CO •vT •vr cn CO 00 CO CD CO CO CO cn cn •vT •vT cb oi cb csi cb in csi oi •vf csi CO* d in oi csi cb oi cb csi 1^ cb in •vf CN •vf cb cb oi C co cn o CO CO i^ g o o to O) CN i^ o •vT in co CO g in CM CO CN CM CN CN CN CN CO CM CO CN CM CO CN CO CO CM CN CO CN CO CO CN A n

o O CO o o o o o o in o o r— o CO o l- r- CO r- •vT r- o r- O I- o CO r- o o o 1—

> m CO CO o 00 m CO CO in o CO r- CO CO CM r— t-CM CO CO t-r- cn t- h- h- O •vf CN cb d CN d in in d CN •vf i^ cb d O d cb d O o d o O oi o o o wz. d d d d d d d d

CO CO i^ CO co in co cn •vT T CO CO in •vT r- in in in CO CO CO cn cn CO •vT CO CD CO CM d oi cb cb d •vf in in d V* oi CO CM O in cb •vf in •vf cb •vf csi r--' oi csi d o CN 1—' CM d CN CN

CL cn "vr LO m CO CO CO •vr CO •vT f- t>- tn CN m T O) O) CO LO •vT CD l~- O) CM r~' co cb oi csi d cb oi in •vf Csi cb cb oi Csi CO d d CN oi 00 CO (/) CO LO 1- CO •VT CN CN CO •vT CN CN CN CO CO •>f in 1- CN m •vf m

D) co 00 O) CM co •vT CN r- •vT CO CO O) to in CO i^ •vT CN CN LO m CO CN CO •vT ZI cb CN CM cb cb co •vf CN cb cb cb CO 2 cb oi cb cb cb •vf 00 csi cb •vT cb tb d i-^ cb •<<£ in CO CM CN T— in co CN co ** in 00 CO 00 CO CN o CN CN •vT m

3 CO CO CO CN m CO i*- •vT O) co i^ CO in •vT CO CO •vT 1^ 1^ CO CO •vT O) CO to CO CM CN cb •vf oi CO cb CN cb i^ CN •vf d d in in 55 •vf l»* vO oi csi CO oi •vf csi i-: csi cb in oi cb ) i— CO •vr in CO i-- CO in I-- CN CO CO o CN CN •vf o CO • CO CD CO CN CO •vT CO CM CO CM

c CO Tf 1^ CD r-- m CO CO CN m m r- CO CO CO cb CO r~ CO CO in CO CM CO CD in cn in CO CN CN cb oi in cb cb r»* *J CN csi cb oi cb cb CN csi d ob cb CO m "2 •vf r— O) m 92 •vr CM CO in CO •vT co i— CN CN LO m to CO to CO l^ co 00 i^- • •vf o CN

CO co cnm co •vr in to co CN in r— cn cn CO CN cn to CO CO CO to O) tN CD •vf CO CO CO ui cb cb CM oi in Q 1^ d o cb oi tb in CO d in d in oi •vf CM •vf tb d csi co CN CM <2 i^n •vf CM d co CN CM in CN CM •vT

o 1^ CO in CO CO co m r— O) to CO r- oo CO CO r— H •vT CO to co 1- to in •vT to oi d CO cb •vT O in •vf O CN o oi O csi o oi Csi csi d o cb d CN < CN d d d d d d CM

(0 o o in o H o in r- in o CO o CO o H CO CO co CO o r- CO o r- •vT o oo r- CN o r- o d O d d CN o d CN d O d O o in ,_• O CO O d d d d d d d d

z>c o o o o CO o o o o in o l- o o o r- CO r- r- r- CO r— o r- m r- d r- m o d h- CL) Li. d d o O Csi O O O ,J O O d o O o d d d d d d d d d d

C o h- in o in r- o o o co o o o o o CO o o I- CO r— o cn l- r- in o o CO o o CO CO —J o d d O CN O d O oi o O d in d d d d d d d

D! 2 !2 22 2 o CM CO •vf in to i^ co cno CM CO •vT m CO i^ oo cno CM co •vT if •vf •vf •vr •vf T in in in m m m in in in m to CO CO CO CD co to CD to 1^ i^ i^- >- • CD cn O) O) O) O) CD cn cn cn o> O) CD O) O) 5cn cn tn CD O) CD O) cn o> cn cn cn cn o>

/65" m in T- to N OltlMnNOlttNWN^ co "rr uo C.O. ^ooinoioNoioiovori csi d cb fS CO CN»-coini-iocoo>cocM3r-CMm csi co 01 CM COCMCOCOCNCNCOCOCNCNTJ-CNCOCN

uo o o o co H r- h- O O CM I-O CM CO r- O Q uo TJ" uo < a> r~ co d (si o o o O o o 0 d uo' d d odd d «- d CM co

C01/)(NOIt>-tNt(NOCOeOtSOM Z

inncoiniNcouicoiO'-Nionco'criN^ O CN CD J : T cb^^cbcNcb'TrT} cp csi in cb cd cb d csi T- T-

r-T-COr-COCOOlCMCMCMlOCOTj-CMCO'r-CO cn a> uo o m oi N^BlNU)dN^6lboitNlflS0«)co' cb ai d oi co o r n in MO r- t ii) o to co

oil NsmstNtOT-oiaiToi < at uo Q O UO TJ" m CM CN o cb < m r- o CM CO TT TJ" TJ" co 3 uo d ui TJ- 1- in co co CN

cqincNi/)r-Tto>cor-.T-inT-cocococoin T- CO t- CO T- cb CN d d cb d CN uo ih CM ^ d in m TT cb cb oi uo P uo N^NNWOli-Nr-VCNSinNNnM co uo o T-

NS^t>-Ol(N[O'»Nn>-COtOO)Nfl0 O CD CN T- ^ °> cb o r^coinoiincocOT-cgoo co in co a> csi uji

(OcooNtuifiNnmocnco^nrjS ^-TT^CON^CCi^COi O CO 2 TJ" CO CO UO CO CO T- I £ Tf co a>

T- r~ T- CO h- T-COCDCOUOTrcOTJ-CQCOTj" < co oi uo cb oi P d TT cb CN O pT-^aicNddcMcbcbT-: 8 T- uo uo CN CO CN T-

r— r— r— r— ^ O) TJ- O O T- t- O TJ- TJ- o o oo 5 to N O O O O o -1 CO T- CD cb T-; o d d d r-^ uo oi in dddd co in d CM

l-l-COr-Or-l-Or-r-OI-COOOOCD UL CM CO 01 Ol CN UO OO O OO OO O o O uo TT "2 dd dad odd Tf CO

UOCOT-I— OOr-OO-TrcOI—OOh-OCO O IO Ol Ol 5 Tf Tf uo o'dP ° dd° °d odd o o o d T- co cn

c UOCOr~COO>OT-CNCOT}"uOCOr~coa>OT- cs 5 SSNSSCOCOtOCOCOCOCOIOtOCOOlO) E 2 to nj oioioicnoioioioioioioioioioioooioi 15 i • 5 CO uo r»- •>T ir O) T CO CO co ir m r- in CO CM Ol CM r~- CO m CM CM ~~ ir co CM 3 CO CN oi to' rf Ol O) iri if if d CO oi cb CM d IT iri CM co to CM co if to oi d C CO CO CO CO IT O) CO Ol CM CM co CO H Ol O CM oo CO co C CinM CoM CO CM CM mCM CoM CirM CoM CM CoO CM CoN CoO CM CN CM CN CM CM <

o ir to Ol in to r- CO r- Ol if CO ~- CM CO co in r- CO CO co ir CM 01 *- 00 in CO CM iri co oi to CO tb oi CO oi if iri m £ if CM iri i-' if oi CO if if iri •>f O) ai to ^—* Q CM m CM CM ir CM CM CM co CO in CO ir to 5' CM m CO co CM If ? If CN 1-

CO CM CO CO CO in to .— ,- CO co co i— CO Ol CM CO ir in CM to to if O) if oi m ^ i^ eb oi cb cb CM cb CO iri cb cb CM tb d iri CM CM i— co CO CM 3 CM CM CM *—' CM CO s? co CO CM CO CM in 3 ir in

o co ir ir— I CO CO Ol i— Ol O) Is- to cn 00 CO ir i-~ ir oi ir in ^ pi b 6 cd csi d ^ ^ m s O. Cb S Oi b S IN CO O ob cb to 5 Ol 9 CM CM d CM CM i~ CO to CO CO CM 1- 1-

r-00101r-COr-COincO|—r>-c001iTOr-r— OtOp— i-r-COOOOCOiTCOi-OCO CO o cbcb0d°CNd'-0ibif'<-J'-: °. °. o P if °. *- d cb ~ w o d d oi- CMOO oo i- 5

Ol OOOOh- Tr-Oi-CMOOOOOOOOOOOOOOOOOOOOOCOO o. ^ cb d |3 O i- CM CM

OOOOOOOOOOOOOOOOOOOOOOOO.OOOOOOOOO

T- OOOOOOOr— r— OOOOOOOOOOOOOO O O O o o o O O oi to O O d d

O r- O O "T O r— r- to co ir co J (0 o •>- co CO in CO CO CO o CM i- co h- CO CM o *- cb cb CM d d o d d CM cb o ^ if 1" N CN ri CO IO csi 12 d d i- CO

i- to 00 Ol r- co if m in in ir CD to CM in oo ir ir CN T- Ol i- m CO *t *- t ^ ffl O) N co rg *- cb csi if T r-^ oi csi iri •f iri csi d d iri tb 1^ CM d cb •>f cb in oi oi ^ d CM i- CM CM CM CM —* i- i— IT CO —* •>f i— CM

oo i- CO to m ir CN CO CO O CN CN CO CO O IT in CM ir ir in co IT io in ir in oi T- CMo i co m csi CN d i" CN "> if Oi I- N oi if cb in iri i- to d cb iri cb CM »- CO JQ cb cb i— T CN CN in m ir CM t- co CM CO IT 5

ir CO co in co Ol O i- CM h- in in h- Ol 00 in Ol in in CM in i- co CO oi Ol CM d cb ob if oi h-' d cb in i- cb oi csi oi CN cb if in 00 cb 5 5 CO co co co co CO CO in CM ir r»- i~ CO co CM i- 1- CM m to

Ol co m co in ir CM CN CO co i^. co CO CO ir CO CO f oo to co in in CM Ol in co cb i^ if csi in csi cb CO CM cb cb in cb if tb tb in cb csi CO d ob i^ CO iT co v— CN CO CO tr CM co in ir in to co in CM co CN ir

o CM CO if in to co Ol o CN CO T CO CO to to to CO CO to co to to r- 1^ 12 Ol o Ol Ol Ol o> Ol Ol Ol Ol Ol Ol Ol Ol oi

It, loioionooioNoioiuiui TJ" CM CO T— CO O O CN CO ibiririvcdnoidaicd cb -rr co

lOr-UOUOCOCOCOT-Oli-CNr-COCOlOCMO CM T- O < T- uo co r- UO CO' ""T d CO CO TJ-' t> *~ CN CO CO S uo "> d "TT K CO CO CN CO CN CO CM T- co co T- T- r-

CNT-r-TrcoT-o)r-couoTro)coTrocoo) CN o> r- in ; n TJ" O CO T- cd uo cd cd cd r-^ cb 9 I CN CN CM CN CM UO UO 3

CNCNr-coooTfCMcococncocDTrcNTr'TrcN LL Ol CSI T- CD cb cb T— TJ* cb d o co cb CM uo cb cb cb cb TJ* CO CO CO CJ) T- T- CM CM CM T- T-UOT- UO CD CM CO

OOOOOTj-l—OUOCNCOTj-Or-CM o oi io in N cb P cb cb d eg cb TT CO o 3 CN •* CO CO

ooooooooo o o o o o o o TJ" T- O d UO CO CM 3 T|

ooooooooooo o o o o o CN CO T- UO 10 oi oi id Tf CM CO

oooooooot-oooo r- O o o O N t CO 10 d d cd ib N CM T- 00

r- O - r- W CO T- CO I ro T CO CO CN CN 1-OHCOTj-Or-CO o d csi P oi cb cb d 5 Tj" P T^ cb P d T- O T- T" O T- O

COUOT-COCMUOTj-Tj-t^cjOCMr-COUOTj-cOt- co r- TJ" T- csidcbr^csiuocsiTr'^d^oT-T-^gjp Tj" l< 8 < o o T- cb T- CM CO T- T- T- O T- r-. CM

COO)CMCOCOCncOr-COCDO)COU0COT-Tj"Tt" r~ in o —> o o o o NdcoNN'tr-bscdtbsTfiooaiDTr CN UO TJ" CN T- CO CM CM T- CM TJ- ,- CO CM T- r-

COCNCOCOCOOO)r~r-r~COUOTro>TrO)CD T- CD O) O O UO CN d *~ ds^NooiribTfriob^iri^riiri cb oi ^ T-COT- COT-CMCOCOT-T-T- T-T-Tf CM CO

N r- in r- Ol N CM r-r-COCNCNCMCOOOOCO m N oi 2 o O CM CD NHJCONNOib'T CM CO -o -," " m CN co s ' cb Tj" CO CO CO CM T— CN T— T- T- CO

uocDr-coo)OT-cMcOTruocor-cocnoT- SSNSSlOCOCOCOIOCOtOCOtOtOC'IO] O>O)O0O0O)O)O)O>O)O)O>O>O)O>O)O)O) E g CO rax': CD CD O Ol E E Z

Figue Al. Slesse Bark Landslide Geomorphologieal Map. See oversize.