PSU Little Wind River Landslide Geology Study

Document Sample
PSU Little Wind River Landslide Geology Study Powered By Docstoc
					  BIG SLIDES            ON THE           LITTLE WIND:
ASSESSMENT OF THE LOWER LITTLE WIND
   RIVER NORTH BANK LANDSLIDES,
   SKAMANIA COUNTY, WASHINGTON


                           July 5, 2011




                             Prepared by:

             Portland State University
                        Department of Geology
                             P.O. Box 751
                        Portland, Oregon 97207




                             Adam Reese
                              Geologist




                            James Randall
                              Geologist




                        Under the Supervision of:
     Dr. Scott Burns, Portland State University Geology Department
Lower Little Wind River North Bank Landslides                                                                                    Page 2 of 33
Skamania County, Washington                                                                                                    March 22, 2007


                                                   TABLE OF CONTENTS

                                                                                                                                         Page

1.0      EXECUTIVE SUMMARY ........................................................................................................... 1

2.0      INTRODUCTION ......................................................................................................................... 2

3.0      SITE BACKGROUND ......................................................................................................5
         3.1   Previous Work .........................................................................................................5
         3.2   Geology ....................................................................................................................6
         3.3   Soil ...........................................................................................................................7
         3.4   Vegetation ................................................................................................................8
         3.5   Climate .....................................................................................................................9

4.0      METHODS ......................................................................................................................10

5.0      RESULTS .........................................................................................................................13
         5.1 Upper Slide A ........................................................................................................13
         5.2 Upper Slide B .........................................................................................................14
         5.3 Lower Slide #1 .......................................................................................................14
         5.4 Lower Slide #2 .......................................................................................................15
         5.5 Lower Slide #3 .......................................................................................................16
         5.6 Lower Slide #4 .......................................................................................................18

6.0      DISCUSSION ...................................................................................................................19
         6.1  Process ...................................................................................................................19
         6.2  Concerns ................................................................................................................19
         6.3  Causes ....................................................................................................................21

7.0      RECOMMENDATIONS.................................................................................................23
         7.1 Option 1: No Remedial Action .............................................................................23
         7.2 Option 2: Slope Stabilization by Revegetation .....................................................23
         7.3 Option 3: Slope Stabilization by Dewatering .......................................................24
         7.4 Option 4: Slope Regrading, Buttressing, and/or Rockfill .....................................26

8.0      CONCLUSION ................................................................................................................26

9.0      FUTURE WORK .............................................................................................................26

10.0     LIMITATIONS ................................................................................................................27

11.0     REFERENCES .................................................................................................................27




                                                                                                       Portland State University
Lower Little Wind River North Bank Landslides                                                                                   Page 3 of 33
Skamania County, Washington                                                                                                   March 22, 2007


                                                              FIGURES

1        Site Vicinity Map ................................................................................................................1
2        Aerial Photo .........................................................................................................................2
3        Photo: Overall Slide Area ....................................................................................................4
4        Large Landslides of the Columbia River Gorge ..................................................................5
5        Geologic Map.......................................................................................................................6
6        Photo: Soil Anomaly ............................................................................................................8
7        Precipitation Comparison ....................................................................................................9
8        Photo: Field Methods .........................................................................................................10
9        Site Plan: Area Layout ......................................................................................................11
10       Area Layout: Volume Calculation .....................................................................................12
11       Layout Feature Map ...........................................................................................................12
12       Slope Model Output: Slope/W ...........................................................................................13
13       Cross-section A-A’ ............................................................................................................14
14       Lower Slide #1 – Plan View ..............................................................................................15
15       Lower Slide #2 – Plan View ..............................................................................................16
16       Lower Slide #3 – Plan View ..............................................................................................17
17       Photo: Slump – Lower Slide #3 .........................................................................................17
18       Lower Slide #4 – Plan View ..............................................................................................18
19       Photo: Scarp – Lower Slide #4 ..........................................................................................19
20       Photo: Scarp – Upper Slide A ............................................................................................20
21       Photo: Ponding – Lower Road ...........................................................................................22
22       Photo: Cutbank Scarp ........................................................................................................22
23       Area Layout: Remediation .................................................................................................25
24       Trench Drain Details .........................................................................................................25


                                                               TABLES

1        Field Measurements – Upper Slides ..................................................................................30
2        Field Measurements – Lower Slide #1 ..............................................................................31
3        Field Measurements – Lower Slide #2 ..............................................................................32
4        Field Measurements – Lower Slide #3 .............................................................................33
5        Field Measurements – Lower Slide #4 ..............................................................................34


                                                          APPENDICES

A        Soil Lab Data – Sample #1 ................................................................................................35
B        Soil Lab Data – Sample #2 ................................................................................................36
C        Soil Lab Data – Sample #3 ................................................................................................37
D        Area Layout – Upper Slide Survey Stations ......................................................................38
E        Slide #1 Cross-Sectional Profile ........................................................................................39
F        Slide #1 Volume Calculation .............................................................................................40
G        Slide #2 Cross-Sectional Profile ........................................................................................41



                                                                                                       Portland State University
Lower Little Wind River North Bank Landslides                                                                        Page 4 of 33
Skamania County, Washington                                                                                        March 22, 2007


H        Slide #2 Volume Calculation .............................................................................................42
I        Slide #3 Cross-Sectional Profile ........................................................................................43
J        Slide #3 Volume Calculation .............................................................................................44
K        Slide #4 Cross-Sectional Profile ........................................................................................45
L        Slide #4 Volume Calculation .............................................................................................46




                                                                                              Portland State University
Lower Little Wind River North Bank Landslides                                                   Page 5 of 33
Skamania County, Washington                                                                   March 22, 2007


1.0      EXECUTIVE SUMMARY

In the winter of 2006-2007, a series of large landslides reactivated on the south facing slopes on the north
side of the lower Little Wind River in Skamania County near Carson, Washington. The large upper
landslides that make up the landslide complex boundaries have scarp heights of up to 4 meters (13 ft.),
scarp widths of 38 meters (125 ft.), and total length of up to 90 meters (295 ft.). The maximum flank
width is 76 meters (249 ft.) and the maximum length from upper scarp to toe is 90 meters (295 ft.). The
approximate volume of material displaced by these landslides is 59,459 cubic meters (77,770 cubic
yards). The overall landslide complex area is approximately 3,788 square meters (13,377 sq. ft.) and is
therefore classified as a medium landslide (Cornforth, 2005). The majority of the landslides investigated
in this assessment are classified as earthflows. Burns (1998b) describes an earthflow as a slow-moving
downslope flow of unconsolidated, water-saturated material that produces a well-defined scarp and a
tongue-shaped debris deposit. However, one of the slides has characteristics of a slump, with downward
slipping of material on a concave-upward failure plane with backward rotation (Burns, 1998b).

Five major elements combined to cause these slope failures to occur including steep slopes associated
with the site location within the Columbia River Gorge and the Cascade Range physiographic province,
geologic factors associated with the weak Ohanapecosh Formation and clay-rich Steever and St. Martin
Series soils, climatic factors associated with rainfall volume and intensity of November and December
2006, human factors associated with forest roads constructed in the vicinity of the site, and undercutting
of the slope (reduction of resisting forces) due to riverbank erosion of the slope buttress. The primary
contributing factor is assumed to be the increased water in the soil system, although the effects of the
lower road cut relative to the location of seeps also may have had a significant contribution.




                                                                             Portland State University
Lower Little Wind River North Bank Landslides                                                    Page 6 of 33
Skamania County, Washington                                                                    March 22, 2007


Landslides along the lower Little Wind River have the potential to introduce large amounts of sediment to
the lower Little Wind and lower Wind Rivers. Preventing sedimentation is important to water quality
protection in order to limit sediment influx, allowing aquatic species to see and breathe, as well as
maintaining sediment-free spawning beds. Stabilization of the slopes also protects mature vegetation
along the banks. While alive, the trees provide shade to the stream and keep water temperatures cool.
When fallen, the trees provide woody debris, which allows refuge and habitat for fish.

Mitigation options presented in this report include: placing tarps over the upper slopes, revegetation of the
slope, dewatering of the slopes via collector drains, slope buttressing, and slope regrading.

2.0      INTRODUCTION

Reactivation of landslides within the Lower Little Wind River North Bank Landslide Complex occurred
in November 2006, during a month of consistently heavy rainfall (Gundersen, 2007). Members of the
Wind River Watershed Council (WRWC) notified Dr. Scott Burns of the Portland State University (PSU)
Geology Department of the occurrence, and the authors of this study mobilized for fieldwork and site
reconnaissance on Saturday, February 10, 2007.

The lower Little Wind River is located in the Columbia River Gorge, approximately 1.6 kilometers
(1 mile) east of the town of Carson, in Skamania County, Washington (Figure 1). The portion of the site
studied is located in Township 3N, Range 8E, Section 31 – Southwest Quarter, and can be found on the
Carson WA-OR USGS 7.5 minute Topographic Quadrangle. The landslide complex addressed in this
study is located approximately 0.40-kilometer (0.25-mile) upriver from the mouth of the Little Wind
River as shown in the attached aerial photograph (Figure 2) after Haagen (1990). Daniel Gunderson is the
primary property owner in the lower Little Wind River vicinity, while the USDA Forest Service manages
most of the Little Wind River watershed.




                                                                              Portland State University
Lower Little Wind River North Bank Landslides                                                     Page 7 of 33
Skamania County, Washington                                                                     March 22, 2007


For this study, the PSU Geology Department was asked by the WRWC to provide a complete report on
the landslide. The purpose of this study is to provide spatial and cross-sectional mapping of the landslide,
analysis of the processes and causes associated with the landslide event, and recommendations for
remedial options to mitigate the present and future hazards associated. This report represents the
fulfillment of the verbal agreement by PSU to provide these services.

Due to its geographic proximity to the Columbia River, the Little Wind River is a rare and important
tributary within the Wind River ecosystem. In the Columbia River Gorge, ecosystems such as this are
unique above Bonneville Dam, a result of the natural geomorphology and backwater flooding effects.
Streams at low elevation and within 1.6 kilometer (1 mile) of the Columbia River have a great potential
for biological diversity and provide habitat for threatened and endangered species such as Coho Salmon,
Winter Steelhead, and Cutthroat Trout. Because of the relatively low gradient of the Lower Wind River,
it has the potential to be a rearing ground for the aforementioned species, as well as having the potential
to support primary populations of Chinook and Chum salmon. The absence of urban development along
the Little Wind is another important and unique attribute of this area (WDFW, 2006).

Landslides along the lower Little Wind River have the potential to introduce large amounts of sediment to
the Little Wind and lower Wind Rivers. According to Tova Cochrane of the Underwood Conservation
District, prevention of mass wasting is important to water quality protection in order to have less
sedimentation in the water, allowing fish species to see and breathe (2007). Stabilization of the slopes
also protects mature vegetation along the banks. While alive, the trees provide shade to the stream,
keeping water temperatures cool. When fallen, the trees provide woody debris and refuge and habitat for
fish (Cochrane, 2007). The purpose of this study is to assess the current and previous slope failures on
the north side of the lower Little Wind River and provide recommendations for remedial options in order
to mitigate the present slope failures and reduce the likelihood of additional landslides within the study
area.

The scope of work for this project consisted of several phases of investigation. Prior to visiting the site,
investigation team members reviewed available topographic maps, soil survey information, and aerial
photographs of the site. Background information related to other landslides that occurred in the vicinity
of the Lower Little Wind River North Bank Landslide Complex was discussed. A site visit was then
conducted for the purpose of data collection and site reconnaissance. The initial field visit was conducted
on February 10, 2007, with field team members and Dr. Scott Burns of the PSU Geology Department.
During this initial field visit, it was determined that the landslide area encompassed slides both above and
below the Lower Road. The study area was limited to the area covered in this report, although additional
sliding may have historically occurred immediately east of the study area. A photo of the overall
landslide area is presented as Figure 3.

On February 15, 2007, a second site visit was conducted to delineate the slide area upslope of the Lower
Road. The study authors were assisted during this field event by Rex Whistler, a PSU Geology student.
During this visit, the main scarp, minor scarps, and flanks of the landslide were field delineated, as well as
determining proximity of the upper slide to relevant features such as the upper road and the Williams
Northwest Natural Gas Pipeline located north of the landslide area. Global Positioning System (GPS)
waypoints were collected at representative locations in the vicinity of the landslide; however, these data
were not used in delineation due to the limited accuracy of the GPS unit. It was determined during this
visit that the upper landslide area consisted of two large landslides separated by a central ridge.
Preliminary sketches were prepared for each of the landslides within the study area.

Following the initial site visit, members of the WRWC were contacted, including: Daniel Gundersen,
property owner; Tova Cochrane, facilitator with Underwood Conservation District; and Margaret Neuman



                                                                               Portland State University
Lower Little Wind River North Bank Landslides                                                   Page 8 of 33
Skamania County, Washington                                                                   March 22, 2007


of the Mid-Columbia Fisheries Enhancement Group. The WRWC is described as a “…partnership, which
encourages the use of land management practices which sustain and improve water quality, fish habitat,
and other natural resources while contributing to long-term economic and community sustainability
within the Wind River watershed” (WRWC, 2007). Each of these contacts provided additional
background information about the site, including land use history, prior slope failure, and information
regarding the ecological importance in preventing mass wasting (Cochrane, 2007; Gundersen, 2007;
Neuman, 2007).




Two subsequent field evaluations were conducted on February 22 and March 3, 2007. During these
visits, the lower landslides were field mapped. Field measurements included scarp and toe dimensions,
length of center line segments, width of transects, slope angle, overall landslide length, and average depth
of the material displaced by the landslide. Physical features such as the Little Wind River, cut bank
dimensions, and seeps were mapped. Soil samples were collected to determine Atterberg Limits and
moisture content in the landslide soils. Photographs were taken extensively during all phases of the site
investigation.

Landslide activity and associated hazards have been present and abundant in the recorded history of the
Columbia River Gorge (Palmer, 1977). Figure 4 shows the Large Landslides of the Columbia River
Gorge (Palmer, 1977). The slopes in several locations on the north side of the Columbia River Gorge,
including the area shown within Figure 4 known as the Wind Mountain (Collins Point) Landslide
Complex, have been particularly susceptible to slope failure. The basic reasons for this are the presence
of the three major elements typically associated with landslide occurrence in the Pacific Northwest.
These include: (1) steep slopes, (2) geologic conditions susceptible to landslides, and (3) climate



                                                                             Portland State University
Lower Little Wind River North Bank Landslides                                                 Page 9 of 33
Skamania County, Washington                                                                 March 22, 2007


conditions that bring excess soil moisture to the susceptible areas (Burns, 1998b). In addition, factors
such as human impacts and riverbank erosion also played significant roles in the landslides assessed in
this study.




Following compilation of these data, remediation and mitigation options were considered for the
landslides. These recommendations are presented based upon interpretation of the landslide processes in
effect, causes of the landslide event, and potential concerns regarding the impacts of future slope
movement.

3.0      SITE BACKGROUND

3.1      Previous Work

Any research-based engineering geology analysis inevitably draws on the work of others, and this is
particularly the case for this analysis of the Lower Little Wind River Landslide Complex. For conceptual
landslide background material, as well as characterization background and remedial design information,
the works of Burns (1998a and 1998b) and Cornforth (2005) were utilized. Geological information,
including historical geology and landslide background for the Columbia River Gorge, was gathered from
Allen (1984), Berri and Korosec (1983), Korosec (1987), and Palmer (1977). Background information
regarding the Little Wind River ecosystem and the potential for landslide impacts to habitat was found in
the reports of Cochrane and White (2005), the Washington Department of Fish and Wildlife (2006), and
the Washington Department of Wildlife et al. (1990). Additional site specific information was collected
through conversations with the Little Wind River stakeholders including: Daniel Gundersen, property
owner; Tova Cochrane, facilitator with Underwood Conservation District; and Margaret Neuman of the
Mid-Columbia Fisheries Enhancement Group. Soil background information was found in the Soil Survey
of the Skamania County Area (Haggen, 1990). Climate data were found on the websites of the National
Oceanic and Atmospheric Administration (2007) and the Western Region Climate Center (2007). For
field mapping methodology, the field-developed cross-sectional process of Williamson et al. (1991) was
followed. Volume calculation methodology was borrowed from the works of Cruden and Varnes (1996)



                                                                           Portland State University
Lower Little Wind River North Bank Landslides                                                  Page 10 of 33
Skamania County, Washington                                                                   March 22, 2007


and UNESCO (1990). Finally, the lectures and guidance of Dr. Scott Burns (2007) were followed for the
overall analysis and structure of this report.

3.2      Geology

The subject area for this study is located in the Columbia River Gorge, within the Cascade Range
Physiographic Province. The geologic foundation of the subject area is simply unconsolidated soil
deposits of Steever and St. Martin series, overlying Tertiary Volcaniclastic (Tvc2) bedrock made up
primarily of tuff and tuff breccia, as well as volcaniclastic breccia, volcanic sandstone, lapillistone, and
conglomerate (Korosec, 1987). The portion of the Hood River Geologic Quadrangle (1:100,000 scale)
for the study area vicinity is presented in Figure 5 (Korosec, 1987). The Tvc2 formation is at least 300
meters (985 ft.) thick in the vicinity of the study area (Berri and Korosec, 1983). Korosec (1987) refers
to the lithology of the site study area only as Tertiary Volcaniclastics of the of Upper Ogliocene age (~25-
32 Ma); however, previous studies have equated these volcaniclastics to those of the Ohanapecosh
Formation (Berri and Korosec, 1983). The formation underlying the study area is similar to the notorious
Ohanapecosh. However, the Ohanapecosh is typically older, more silicic, and less tuffaceous than the
Tvc2 formation in the lower Little Wind River area (Korosec, 1987).




Within the Tvc2 formation are breccia beds up to 15 meters (49 ft.) thick. These beds are made up of
materials representing pyroclastic flows, deposits from debris flows, surge deposits from explosive
eruptions, and mudflows of low energy (Korosec, 1987). Representing deposits of fluvial reworking, the
thinly bedded sandstones of this formation are poorly sorted and consist of volcanic debris (andesite clasts
and altered pumice particles) (Korosec, 1987). The 12 centimeter (4.7 inch) thick beds of lapilli stone
within this formation are composed of well sorted pumice lapilli and rounded andesite (Korosec, 1987).
Conglomerates of this formation contain porphyritic and aphanitic andesite clasts, as well as basaltic
andesite (Korosec, 1987). Alteration of most of the volcaniclastic rocks of this unit has taken place, with
extensive replacement by zeolites and clay minerals (Korosec, 1987).




                                                                             Portland State University
Lower Little Wind River North Bank Landslides                                                  Page 11 of 33
Skamania County, Washington                                                                   March 22, 2007


The beds of this formation dip south toward the Columbia River at 5 to 30 degrees and containing
interbedded clay and clayey paleosols. For these reasons, the rock units of this formation are considered
to be weak (Korosec, 1987; Palmer, 1977). Palmer (1977) describes the effects of excessive moisture on
these Tertiary volcaniclastics within the adjacent Wind River (Collins Point) Landslide Complex. In
cases where these clay layers become saturated, the plastic flow causes extension, subsidence, and block
faulting on the upper portions of the hillslopes that the Tvc2 formation lies beneath (Palmer, 1977).

Separating the lower Little Wind River north bank Tertiary volcaniclastics from the Quaternary Landslide
Deposits (Qls) found to the south of the Little Wind in the site vicinity, is the northeast-trending, south-
dipping Little Wind River Fault (Berri and Korosec, 1983; Korosec, 1987). The Little Wind River Fault
intersects the northwest-trending Wind River zone (a linear topographic feature with no evidence of
faulting, that includes aligned intrusions and hot springs) and cuts the augitic Buck Mountain intrusion
(Tidi) to the west of the study area (Berri and Korosec, 1983; Korosec, 1987).

3.3      Soil

The specific soil types for the subject area as found in the Soil Survey of the Skamania County Area
compiled by the USDA Natural Resource Conservation Service (Haagen, 1990), are Steever Stony Clay
Loam (30-65% slopes) and St. Martin Gravelly Silty Clay Loam (30-65% slopes).

The Steever Stony Clay Loam soils make up approximately 40% of the landslide area. These very deep
soils, found on back slopes, toe slopes, and foot slopes, are well drained and form in colluvial landslide
material derived from conglomerate, andesitic, and basaltic parent material (Haagen, 1990). They
typically consist of a layer of decomposing biomass (~ 5 centimeters thick (~2 inches)), above an upper
surface layer (~12.7 centimeters thick (~5 inches)) of very dark brown stony clay loam, above a lower
surface layer (~17.8 centimeters thick (~7 inches)) of dark brown gravelly clay loam. The upper subsoil
layer (~ 20 centimeters thick (~8 inches)) is a dark brown very gravelly clay loam, above a lower subsoil
layer (~258 centimeters thick (~10 inches)) of dark brown very gravelly loam (Haagen, 1990). The
substratum ( 152 or more centimeters thick (~60 or more inches)) is additional dark brown very gravelly
loam (Haagen, 1990). Permeability of Steever Series soils is moderate within the soil layers (Haagen,
1990).

The Steever Stony Clay Loam soils within the slope range of 30%-65% have several factors that make
them prone to landslide hazards. Steep slopes, severe runoff and erosion potential, and moderately high
soil moisture capacity combine to allow potential for slope failure (Haagen, 1990). Steever Stony Clay
Loam soils have a moderate to high liquid limit (25%-40%) and a low plasticity index (5 – 15%)
(Haagen, 1990). A soil sample collected from the headscarp of the upper landslide (presumed to be
located in the Steever soil zone) on February 24, 2007, had a moisture content of 33%. This soil sample
had a liquid limit of 24.3%, a plasticity limit of 23.9%, and a plasticity index of 0.4%. Soil sample
laboratory data sheets are presented in Appendix A through C.

The St. Martin Gravelly Silty Clay Loam soils make up approximately 60% of the landslide area. These
very deep soils, found at the base of slopes that have been disturbed by landslides, are moderately well
drained and form in colluvium derived from andesitic parent material (Haagen, 1990). They typically
consist of a layer of decomposing biomass (~ 5 centimeters thick (~2 inches)), above a surface layer (~10
centimeters thick (~4 inches)) of black gravelly silty clay loam, above an upper subsoil layer (~17.8
centimeters thick (~7 inches)) of dark grayish brown silty clay loam (Haagen, 1990). The lower subsoil
(~152 or more centimeters thick (~60 or more inches)) is light olive brown and light yellowish brown clay
(Haagen, 1990). Permeability of St. Martin Series soils is slow within the subsoil layers (Haagen, 1990).
An anomalous soil layer found within the St. Martin series soils, suspected to be an ash layer, is shown in
Figure 6.


                                                                             Portland State University
Lower Little Wind River North Bank Landslides                                                  Page 12 of 33
Skamania County, Washington                                                                   March 22, 2007



The St. Martin Gravelly Silty Clay Loam soils within the slope range of 30%-65% have a combination of
factors that make them prone to landslide hazards. A seasonally high water table (within a depth of 0.3 to
0.6 meters (1 to 2 ft.)), steep slopes, high runoff and erosion potential, high soil moisture capacity,
potential for shrinking and swelling, and low soil strength and load-bearing capacity combine to allow
high potential for slope failure (Haagen, 1990). In addition, the clay-rich subsoil requires drainage when
locating structures on it due to low permeability (Haagen, 1990). St. Martin Gravelly Silty Clay Loam
soils have a moderate to high liquid limit (35%-55%) and a moderate plasticity index (15 – 35%)
(Haagen, 1990). A soil sample collected from the toe of the landslide (presumed to be located in the St.
Martin soil zone) on February 24, 2007, had a moisture content of 40%. This soil sample had a liquid
limit of 33.6%, a plasticity limit of 28.0%, and a plasticity index of 5.6%. Soil sample laboratory data
sheets are presented in Appendix A through C.




3.4      Vegetation

Vegetation information for the soil types formed in the subject area is also found in the Soil Survey of the
Skamania County Area (Haagen, 1990). Both Steever Stony Clay Loam and St. Martin Gravelly Silty
Clay Loam soils found on the 30 to 65% slopes of this area are best suited Douglas fir (Haagen, 1990).
However, natural regeneration of Douglas fir trees can be inhibited by brush invasion allowed by an
opened canopy in both soil types, as well as by the high seasonal groundwater table found in St. Martin
series soils (Haagen, 1990). Other species capable of growing in Steever Stony Clay Loam soils are
western hemlock, red alder, grand fir, big leaf maple, vine maple, Oregon grape, red huckleberry, trailing
blackberry, creambush oceanspray, western hazel, Pacific dogwood, common snowberry, thimbleberry,
and dwarf rose (Haagen, 1990). Other species capable of growing in St. Martin Gravelly Silty Clay Loam
soils are red alder, grand fir, western redcedar, big leaf maple, vine maple, salal, common snowberry,
dwarf rose, western hazel, honeysuckle, creambush oceanspray, trailing blackberry, and western
swordfern (Haagen, 1990).




                                                                             Portland State University
Lower Little Wind River North Bank Landslides                                                  Page 13 of 33
Skamania County, Washington                                                                   March 22, 2007


3.5      Climate

Climate data comparison is of major importance to landslide investigations, because increased water in
the soil system is the primary contributing factor to most landslides (Burns, 1998b). When combined
with other major factors of steep slopes, landslide-prone geology, riverbank erosion, and human factors,
increased soil moisture is often the trigger that causes the onset of the landslide event. Precipitation for
the winter of 2007 (November 2006-February 2007) was well above average, with especially high
precipitation volume occurring in November and December. Average annual precipitation for the site
vicinity (observed at Bonneville Dam) is approximately 195.6 centimeters (77 inches) (WRCC, 2007).
For the purpose of this study, comparisons can be made between average rainfall for November through
February, the amount of rainfall for this period in 2007, and precipitation for the same period in 1996 -
considered to be the 100-year event for landslide occurrence in this region (Burns, 1998b). Figure 7
shows the precipitation data collected at NOAA’s Portland International Airport weather station (NOAA,
2007). These data include daily and cumulative precipitation for November through February 1996 and
2007, as well as the climatic average for this period. Precipitation data from this location were included
only because they provided a complete set of regional data for comparative analysis and should be viewed
only as examples of precipitation magnitude. Data from the Bonneville Dam or Stevenson (WA) were
not available for all of the data sets evaluated.




As shown in Figure 7, both the 1996 and 2007 winter seasons had above average precipitation. The high
incidence of landslide activity of 1996 can be largely attributed to the nearly 30.5 centimeters (12 inches)
of precipitation that fell between February 6th and 9th of that year (Burns, 1998b). 2007 had nearly 61
centimeters (24 inches) of cumulative precipitation, compared to the average of approximately 53.3
centimeters (21 inches) for the period of November through February. However, a more important factor
was the 22.8 (9 inches) of rain occurring in the first two weeks of November; nearly four times the
amount expected for this period. Slope failures are often the result of a single storm event occurring
when soil is already saturated (Burns, 1998b; Cornforth, 2005). The high volume of precipitation in


                                                                             Portland State University
Lower Little Wind River North Bank Landslides                                                  Page 14 of 33
Skamania County, Washington                                                                   March 22, 2007


November and December included both consistent daily rainfall and high intensity short-duration events,
resulting in high antecedent soil moisture followed by shallow slides triggered by the high intensity short-
duration events. Ponding and ephemeral streams were observed above and within the north bank
landslide area during all field events in January and February 2007.

4.0      METHODS

An initial visit to the landslide site was made on February 10, 2007, for the purpose of site
reconnaissance. Prior to the initial field reconnaissance, topographic maps and aerial photos were
studied. The Puget Sound LIDAR Consortium and the Oregon Department of Geologic and Mineral
Industries were also contacted prior to field activities; however, no LIDAR data is currently available for
the site vicinity. During the following visit on February 15, 2006, systematic measurements were taken
using a Brunton compass, laser rangefinder, and measuring tape. Global Positioning System (GPS)
waypoints were collected at representative locations in the vicinity of the landslide; however, these data
were not used in delineation due to the limited accuracy of the GPS unit. Field mapping methods were
used as described by Williamson et al. (1991).




A representative center line was staked down the center channel of the landslide slope, placing stakes at
each slope break to establish survey stations. Up to seven transects were measured perpendicular to the
center line on each of the four smaller landslides downslope (south) of the lower road. A photo showing
the field team using this methodology is presented as Figure 8. The slope angle, orientation, and length of
each center line and transect segment were recorded, as well as additional slide-specific measurements
such as scarp and toe dimensions. For the larger slides (those originating upslope from the lower road),
the slope angle, orientation, and length of line segments around the perimeter of each landslide area were
measured. In many cases, the flanks of the upper landslides were not obvious topographic features.
Therefore the primary method of determining the boundaries of these slides was observation of vegetation
change fr,om fir species (outside the landslide boundary) to alder and thick brush species (within the
landslide boundaries). This method was chosen due to the greater size of the two upper slides, as well as
the prohibitive thickness of vegetation on the landslide mass. A plan view map showing survey points for




                                                                             Portland State University
Lower Little Wind River North Bank Landslides                                                   Page 15 of 33
Skamania County, Washington                                                                    March 22, 2007


the upper landslides is found in Appendix D. Field measurements for each of the landslides are presented
in Tables 1-5.

Data and observations collected in the field were compiled, and plan view maps of each landslide and the
overall landslide area were constructed. The site plans consist of a scale drawing of each landslide
constructed using slope angle, center line (low point) segment lengths, transect widths, and orientation of
transects relative to one another. In addition to the site plans, slope profiles were drafted for each slide.
The plan view map showing the area layout is presented as Figure 9.




Volume calculations were made using average lengths, widths, and depths of ellipses representing the
displaced material of the landslide before and after failure. This method was first outlined by the
UNESCO Working Party on World Landslide Inventory (1990) and further detailed by Cruden and
Varnes (1996). Based on this methodology, the landslide volumes were calculated using the following
equation:

Landslide Volume = π Dr Wr Lr
                        6
Where:

         Lr = Length of the Surface of Rupture (Minimum distance from toe of surface of rupture to
                 landslide crown)
         Wr = Width of the Surface of Rupture (Maximum width of flanks of landslide perpendicular to
                 length Lr)
         Dr = Depth of the Surface of Rupture (Maximum depth of surface of rupture below original
                 ground surface measured perpendicular to a plane containing Lr and Wr)

Using a best fit ellipse, Lr and Wr were calculated directly using field-measured data. In order to estimate
Dr, (estimation required in the absence of subsurface investigation data) the average depth of the lowest
point in the “center channel” of the landslide at the base of the headscarp. Figure 10 shows the volume




                                                                              Portland State University
Lower Little Wind River North Bank Landslides                                                   Page 16 of 33
Skamania County, Washington                                                                    March 22, 2007


calculation method as performed on the upper landslides. In addition to performing volume calculations,
field observations were used to classify the landslide, as well as to determine the processes that formed it.




In addition to mapping the landslide dimensions, physical features indicative of landslide causal factors,
such as seeps and cutbank scarps were also mapped, as shown in Figure 11. Regional climate data were
obtained from NOAA (National Oceanic and Atmospheric Administration) databases for purpose of
comparison to average precipitation and with years of high landslide occurrence (1996). Laboratory
methods were used to determine the moisture content of the sample collected. Atterberg limits and other
soil parameters were available in the NRCS Multnomah County Soil Survey (Haagen, 1990). The soil
laboratory data sheets are presented in Appendix A through C.




                                                                              Portland State University
Lower Little Wind River North Bank Landslides                                                     Page 17 of 33
Skamania County, Washington                                                                      March 22, 2007


This investigation also employed a modeling tool (Slope/W) developed by Geo-Slope International. A
profile was constructed from a transect through Upper Slide A and Lower Slide #1 (A-A’) using
measurements collected in the field. Lab soil analysis data was also entered (soil unit weight, cohesion,
and internal angle of friction). The result was a factor of safety of 0.96, indicating the need for
remediation (see recommendations. The Slope/W output is presented in Figure 12.




5.0      RESULTS

The Little Wind River North Bank Landslide Complex trends in a general south-southeast direction
downslope, originating approximately 50 meters (164 ft.) south of the upper road, as shown previously in
Figure 9. Although the slide consists of one continuous downslope unit, it can be subdivided based on
descriptive properties into two distinct sections, hereafter described as “Upper Slide A” and “Upper Slide
B”. These two “upper” landslides are separated by a southeast trending ridge that forms a distinct
boundary between the slide masses. In addition to the topographic distinction of the ridge, there are
several fir trees along the ridge distinguishing it from the alder coverage of the landslide masses.

A secondary forest road trending northeast-southwest, described within this report as the “lower road”,
separates the less recently active upper portion of the landslide from the more recently active lower
portion. Within the lower portion of the slide are several distinct smaller landslides. For the purposes of
this study, detailed investigation was limited to the four largest of these smaller landslides slides, focusing
on two slides beneath each Upper Slide.

5.1 Upper Slide A

A field-developed cross-sectional slope profile of the lateral extent and change in character of Upper
Slide A (Cross-section A-A’) is attached as Figure 13. Upper Slide A extends south-southeast from its
head scarp, with an average scarp height of approximately 3.0 meters (9.84 ft.) and a scarp width of
approximately 14.9 meters (48.8 ft.). The body of the Upper Slide A is approximately 55.2 meters (181
ft.) in length and 31.2 meters (102.4 ft.) wide, at an average slope angle of 30 degrees. The area of Upper
Slide A is approximately 1,498 square meters (16,125 square ft.) and would therefore be classified as a



                                                                               Portland State University
Lower Little Wind River North Bank Landslides                                                 Page 18 of 33
Skamania County, Washington                                                                  March 22, 2007


small landslide (Cornforth, 2005). The calculated material volume of the pre-rupture mass of Upper
Landslide A is 23,512 cubic meters (30,753 cubic yards).

At the base of Upper Landslide A, significant zones of accumulation have formed due to due to the recent
activity of the smaller slides below the Lower Road. Upper Slide A has a well defined scarp, hummocky
topography, and vegetation on the primary mass of the upper portion suggesting that the major slope
failure occurred prior to the 100-year climatic event of 1996. However, the fresh character of the upper




scarps suggest that slope creep may still be occurring. Upper Slide A does not have a distinct snout at its
base, however it is probable that the Little Wind River would rapidly erode any significant material
accumulation in this region. Indicative of this is the well defined cut bank scarp, with a maximum
thickness of approximately 3 meters (9.84 ft.).

5.2 Upper Slide B

Upper Slide B extends southeast from its head scarp, with average scarp height of approximately 2.0
meters (6.56 ft.) and a scarp width of approximately 10.8 meters (35.43 ft.). The body of the Upper Slide
B is approximately 89.3 meters (293 ft.) in length and 36.0 meters (118 ft.) wide, at an average slope
angle of 24 degrees. The area of Upper Slide B is approximately 2,290 square meters (24,650 square ft.)
and would therefore be classified as a medium landslide (Cornforth, 2005). The calculated material
volume of the pre-rupture mass of Upper Landslide B is 35,947 cubic meters (47,017 cubic yards).

At the base of Upper Landslide B is a small creek that flows between the base of the slope and an alluvial
terrace extending approximately 30 meters (98.4 ft.) south to the Little Wind River. Upper Slide B has a
moderately well defined scarp, hummocky topography, and vegetation on the primary mass of the upper
portion, suggesting that the major slope failure occurred prior to the 100-year climatic event of 1996.
Upper Slide B does not have a distinct snout at its base, with the exception of the material slumping
within the mass of Lower Slide #3. As with Upper Slide A, it is probable that the small creek flowing
west toward the Little Wind River would erode any material accumulating at the base of the slope. An
ephemeral stream flowing through Lower Slide #4 has probably transported material from the former toe
(zone of accumulation) at the base of this slide.



                                                                             Portland State University
Lower Little Wind River North Bank Landslides                                                 Page 19 of 33
Skamania County, Washington                                                                  March 22, 2007


5.3 Lower Slide #1

Lower Slide #1 extends south-southeast from its head scarp, with average scarp height of approximately
3.5 meters (11.5 ft.) and a scarp width of approximately 12.5 meters (41 ft.). The body of the Lower Slide
#1 is approximately 14.8 meters (48.5 ft.) in length and 9.7 meters (31.8 ft.) wide, at an average slope
angle of 29 degrees. The area of Lower Slide #1 is approximately 144 square meters (1,550 square ft.)
and would therefore be classified as a very small landslide (Cornforth, 2005). The calculated material
volume of the pre-rupture mass of Lower Slide #1 is 261.7 cubic meters (342.3 cubic yards). A plan view
map of Lower Slide #1 is presented as Figure 14. A slope profile of Lower Slide #1, as well as graphical
representation of the volume calculation, is found in Appendix E and Appendix F, respectively.




This very small landslide has a toe at its base with a material volume of approximately 36.3 cubic meters
(47.5 cubic yards). This section (zone of accumulation) represents the accumulated material displaced
from the upper slope (zone of depletion). Approximately 6 meters (19.7 ft.) downslope from the toe of
Lower Slide #1 is the top of a cutbank scarp along the Little Wind River. This erosional feature has an
average height of 3 meters (9.8 ft.) and a maximum height of 4 meters (13 ft.). In addition to a well
defined toe, Lower Slide #1 has a well defined scarp, hummocky topography, and growth of vegetation
on the primary mass suggesting that the major slope failure occurred at the time of the 100-year climatic
event of 1996.

5.4 Lower Slide #2

Currently, the most active slide of the complex, Lower Slide #2 extends south from its head scarp, with
average scarp height of approximately 3.7 meters (12.1 ft.) and a scarp width of approximately 7.0 meters
(23 ft.). The body of the Lower Slide #2 is approximately 17.2 meters (56.4ft.) in length and 6.6 meters
(21.6 ft.) wide, at an average slope angle of 34 degrees. The area of Lower Slide #2 is approximately 114
square meters (1,227 square ft.) and would therefore be classified as a very small landslide (Cornforth,


                                                                            Portland State University
Lower Little Wind River North Bank Landslides                                                Page 20 of 33
Skamania County, Washington                                                                 March 22, 2007


2005). The calculated material volume of the pre-rupture mass of Lower Slide #2 is 216 cubic meters
(283.1 cubic yards). A plan view map of Lower Slide #2 is presented as Figure 15. A slope profile of
Lower Slide #2, as well as graphical representation of the volume calculation, is found in Appendix G and
Appendix H, respectively.




This very small landslide has a toe at its base with a material volume of approximately 38.2 cubic meters
(50.0 cubic yards) and an oblong secondary toe along its eastern lower flank with a material volume of
approximately 102.9 cubic meters (134.6 cubic yards). These sections (zone of accumulation) represent
the accumulated material displaced from the upper slope (zone of depletion). Approximately 6 meters
(19.7 ft.) downslope from the toe of Lower Slide #2 is the top of the cutbank scarp along the Little Wind
River, as described in the discussion of Lower Slide #1. In addition to well defined toes, Lower Slide #2
has a well defined scarp and hummocky topography. During field events, minor cracking was noted
above the headscarp, however crack monitoring did not indicate extensional movement during the one
month span of field visits. No vegetation is found on the slide mass due to the recent activity. However,
the eastern portion of the upper slopes of Lower Landslide #2 have minor growth of alder and grasses,
indicating that these slopes have had more significant failure in recent years.

5.5 Lower Slide #3

Lower Slide #3 extends east-southeast from its head scarp, with average scarp height of approximately
2.7 meters (8.8 ft.) and a scarp width of approximately 14.7 meters (48.2 ft.). The body of the Lower
Slide #3 is approximately 21.0 meters (68.9 ft.) in length and 15.4 meters (50.5 ft.) wide, at an average




                                                                           Portland State University
Lower Little Wind River North Bank Landslides                                                Page 21 of 33
Skamania County, Washington                                                                 March 22, 2007




slope angle of 41 degrees. The area of Lower Slide #3 is approximately 315 square meters (3,390 square
ft.) and would therefore be classified as a small landslide (Cornforth, 2005). The calculated material
volume of the pre-rupture mass of Lower Slide #3 is 798.7 cubic meters (1044.7 cubic yards). A plan
view map of Lower Slide #3 is presented as Figure 16. A slope profile of Lower Slide #3, as well as
graphical representation of the volume calculation, is found in Appendix I and Appendix J, respectively.




Below the head scarp, Lower Slide #3 has a series of two flat terraces of up to three meters ( 9.8 ft.),
grading to steep minor scarps beneath. The lower terrace appears to be buttressed by the deep root system
of a Douglas fir tree. This large tree is angled northwest toward the head scarp at an angle of
approximately 13 degrees to the vertical, indicating backward rotational movement of the slope material.
A photo of the upper portion of Lower Slide #3 showing the slumping is presented as Figure 17.



                                                                           Portland State University
Lower Little Wind River North Bank Landslides                                                 Page 22 of 33
Skamania County, Washington                                                                  March 22, 2007


Approximately 19 meters (62 ft.) downslope from this tree is a small creek at the base of the slope, as
mentioned in the description of Upper Slide B. It is likely that this small creek flowing into the Little
Wind River would slowly erode any material accumulating at the base of the slope if the tree buttressing
the slump were to fail. Of the lower slides, Lower Slide #3 has the greatest potential to deliver a large
amount of sediment to the river.

5.6 Lower Slide #4

The largest in area of the lower slides, Lower Slide #4 extends southeast from its head scarp, with average
scarp height of approximately 4.5 meters (14.7 ft. ) and a scarp width of approximately 15.7 meters (51.5
ft.). The body of the Lower Slide #4 is approximately 35.0 meters (115 ft.) in length and 14.2 meters
(46.5 ft.) wide, at an average slope angle of 32 degrees. The area of Lower Slide #4 is approximately 425
square meters (4,574 square ft.) and would therefore be classified as a small landslide (Cornforth, 2005).
The calculated material volume of the pre-rupture mass of Lower Slide #4 is 1,045.1 cubic meters
(1,366.9 cubic yards). A plan view map of Lower Slide #4 is presented as Figure 18. A slope profile of
Lower Slide #4, as well as graphical representation of the volume calculation, is found in Appendix K and
Appendix L, respectively.




This slide is located immediately adjacent to Lower Slide #3 and the two slides are likely to be related.
Lower Slide #4 has a well defined scarp, hummocky topography, and growth of vegetation on the primary
mass suggesting that the major slope failure occurred at the time of the 100-year climatic event of 1996.
It is probable that an ephemeral stream flowing down the center channel of the Lower Slide #4 has
transported material from the former toe (zone of accumulation) at the base of this slide into the small
creek on the alluvial terrace.




                                                                             Portland State University
Lower Little Wind River North Bank Landslides                                                 Page 23 of 33
Skamania County, Washington                                                                  March 22, 2007


6.0      DISCUSSION

6.1      Process

Using a classification system described by Burns (1998b), each of the landslides within the landslide
complex was classified. All of the slides, with the exception of Slide #3, were classified as earthflows.
An earthflow is described as a generally slow-moving downslope flow of unconsolidated, water-saturated
material (Burns, 1998b). Cornforth (2005) similarly categorizes this type as an “earth” subdivision of the
“flow” landslide classification, because the area of movement consists of predominately fine material.
Other features of these landslides characteristic of earthflow are well-defined scarps and tongue-shaped
debris deposits (Burns, 1998b). The well defined headscarp of Slide #4 is shown in the Figure 19
photograph.

Upper Slides A and B, as well as Lower Slides #1, #2, and #4, each had features indicative of an
earthflow, including fine-grained material, well-defined scarps, well-defined toes, and hummocky
topography. However, based on the field observations of apparent rotational movement of the soil, Slide
#3 was classified as a slump. A slump is described by Burns (1998b) as the downward slipping of
unconsolidated material along a concave-upward plane of failure, usually with a backward rotation.




6.2      Concerns

Of primary concern related to the hazards associated with the Little Wind River North Bank Landslide
Complex is the potential for additional recession of the main and minor scarps, along with the potential
for remobilization of the displaced material within the slide mass and zones of accumulation. Landslides


                                                                            Portland State University
Lower Little Wind River North Bank Landslides                                                  Page 24 of 33
Skamania County, Washington                                                                   March 22, 2007


along the lower Little Wind River have the potential to introduce large amounts of sediment to the lower
Little Wind and lower Wind Rivers. Preventing sedimentation is important to water quality protection in
order to limit sediment influx, allowing aquatic species to see and breathe, as well as to maintain
sediment-free spawning beds. Stabilization of the slopes also protects mature vegetation along the
banks. While alive, the trees provide shade to the stream, keeping water temperatures cool. When fallen,
the trees provide woody debris, allowing refuge and habitat for fish (Cochrane, 2007).

The potential loss of functional property in the vicinity of the slides, such as the lower road, is an
additional concern. In order to prevent additional recession of the head scarps of the lower landslides, the
water infiltration above the head scarp on the lower road must be prevented to avoid further recession of
the scarps into the road. This can be accomplished temporarily by placing impermeable plastic sheeting
over the upper portions of the scarps. In addition, mitigating the groundwater seeps and ephemeral
streams on the lower road above Lower Slide #2 and Lower Slide #4 is important to prevent future slope
failures within the landslide complex. The further recession of the cutbank scarp at the base of Upper
Slide A (below Lower Slides #1 and #2) has the potential to reduce resisting forces and significantly
decrease the Factor of Safety for the western portion of the landslide complex.

In addition to the concern of slope failure on the slides below the lower road, of greater concern is the
potential for reactivation of the upper slides. Field observations of freshly exposed material on the head
scarp of Upper Slide A, as seen in Figure 20, indicate that creep or recent extensional movement may be
occurring. When an ancient landslide reactivates, movement will first be observed at the top and bottom
of the slope (Burns, 2007). Ancient landslides are significant concerns, because if a slope has moved in
the past, it has a high potential for additional future movement (Burns, 2007). For this reason, the upper
and lower slopes of this landslide complex should always be monitored for movement during winter
months.




                                                                             Portland State University
Lower Little Wind River North Bank Landslides                                                  Page 25 of 33
Skamania County, Washington                                                                   March 22, 2007


6.3      Causes

The major causes of landslides in the Pacific Northwest are: (1) steep slopes, (2) geologic conditions
susceptible to landslides, and (3) climate conditions that bring excess soil moisture to the susceptible
areas (Burns, 1998b). Each of these elements were present and played a significant role in the landsides
of the lower Little Wind River north bank, combined with additional factors such as human impacts and
riverbank erosion.

The first two elements described above, slope steepness and geologic conditions, are static for a location.
The slopes are naturally steep in the vicinity of the site due the geomorphology of the Columbia River
Gorge at the mouth of the Little Wind River. Slopes within the Columbia River Gorge are typically in
excess of 30 degrees. Secondly, the geologic setting of the lower Little Wind River vicinity includes clay-
rich soils of the Steever and St. Martin Series above weak volcaniclastic bedrock often associated with the
Ohanapecosh Formation (Berri and Korosec, 1983; Haagen, 1990; Korosec, 1987; Palmer, 1977). The
Steever and St. Martin Series soils have low soil and load bearing strength, shrink-swell potential, and a
high seasonal groundwater table, within a depth of 0.3 to 0.6 meters (1 to 2 ft.) of the surface (Haagen,
1990). With beds dipping south toward the Columbia River at 5 to 30 degrees, as well as the presence of
interbedded clay and clayey paleosols, the rock units of the volcaniclastic geologic formation are
considered to be weak (Korosec, 1987; Palmer, 1977). In the cases where these clay layers become
saturated, the plastic flow causes extension, subsidence, and block faulting on the upper portions of the
hillslopes that they underlie (Palmer, 1977). Additional future landslides could result from the effects of
seismic activity on the Little Wind River Fault, described in Berri and Korosec (1983) and Korosec
(1987).

The third element of landslide occurrence, climatic conditions causing excess moisture, was the ultimate
trigger of the 2007 landslide activity on this site. The winter of 2007 (November 2006 - February 2007)
had above average precipitation, not only above-average total rainfall, but also with significant short-
duration high intensity events. The first two weeks of November 2006 had nearly four times the amount
of rainfall expected for this period. According to Burns (1998b) and Cornforth (2005), landslides are
often the result of a single storm event occurring when soil is already saturated (Burns 1998b; Cornforth,
2005). The high-intensity short-duration associated with the storms in early November 2006, were
certainly of consequence. In addition, the high groundwater table and antecedent moisture resulting from
the consistent precipitation of November and December also played a key role in the ultimate failure of
the slope.

Beyond these geographic, geologic, and climatic elements, human influences were additional contributing
factors in the Lower Little Wind River North Bank Landslide Complex. The practice of placing material
from the roadcut on the upper portion of the lower slopes had significant slope failure implications.
Surcharging slopes with fill materials can cause failure as a result of increased load on the slope, as well
as from the negative effects of artificially steepening the existing slope (Cornforth, 2005). The direct
trigger of the earthflow is likely to have been the loss of soil strength caused by elevated soil moisture
conditions; however, the root causes are likely to have been the locally oversteepened slope and the
weight of the introduced fill (Cornforth, 2005). In addition, the lower road prevents water from the seeps
and ephemeral streams from moving from the upslope to the downslope side of the road (WDFW, 2006).
This results in ponding as shown in Figure 21, followed by infiltration into the soils behind the head
scarps of the landslides. The Wind River Watershed Council has previously identified the need to
remediate these conditions caused by the lower road (WDFW, 2006).




                                                                             Portland State University
Lower Little Wind River North Bank Landslides                                                   Page 26 of 33
Skamania County, Washington                                                                    March 22, 2007




As mentioned previously, an additional factor causing the historic and future landslides, as well as
creating the potential for major slope failures in the future, is the river bank erosion at the base of Upper
Landslide A. The erosion of the cutbank, as shown in Figure 22, has the potential to reduce the
buttressing effect of the material at the base of the slope. This would decrease the resisting forces, reduce
the Factor of Safety, and cause the failure of both the upper and lower slopes.




                                                                              Portland State University
Lower Little Wind River North Bank Landslides                                                    Page 27 of 33
Skamania County, Washington                                                                     March 22, 2007


7.0      RECOMMENDATIONS

In order to prevent additional upslope and downslope impacts from the earthflows and slump of this
landslide complex, mitigation and remedial measures are recommended. Remedial measures should be
completed in phases, beginning with immediate mitigation to avoid further slope failure, followed by
several phases of slope stabilization measures to prevent future slope movement. When considering
remediation of landslides, Cornforth (2005) suggests that two basic factors should be considered. First,
the causes of the landslide need to be identified. The second factor that needs to be identified is the
amount of remediation required to maintain stability for anticipated future conditions (Cornforth, 2005).
In the case of the lower Little Wind River vicinity, these future conditions include the possibility of
impacts to slopes triggered by earthquake events. Four basic remedial options are presented in this report
and provide answers to these questions with varying degrees of confidence. Three of these are
recommended, and a fourth alternative is presented for comparison. These are presented in order of
lowest to highest cost and amount of slope stability achieved. These options include (1) no remedial
action, (2) slope stabilization by revegetation of the slope surface, (3) slope stabilization by dewatering of
the slope, and (4) slope regrading, buttressing, and backfilling the slope with coarse gravel.

7.1      Option 1: No Remedial Action

Although not recommended, one possible remedial option involves taking no immediate action toward
mitigation of the landslides potential future hazards. Although the landslides have the potential to impact
ecological systems and habitat within the lower Little Wind River and lower Wind River drainage, the
landslides assessed in this report do not present threat to human life or downslope structures. Therefore,
taking no action is the least expensive plausible means of addressing these landslides. In this scenario,
the property owner, in conjunction with qualified consultants, should monitor the slope, accumulated
material, and associated hazards to determine if future remedial action should be taken (Cornforth, 2005).
However, the consequences of continued slope failure and potential for remobilization of accumulated
downslope material are significant due to the potential impacts to the threatened and endangered species
discussed previously. Although inexpensive, this scenario will have associated costs that could include
installation of monitoring instruments such as piezometers and inclinometers, as well as consulting fees.
It is important to reiterate that the option of taking no remedial action is not a recommendation of the
authors of this study.

7.2      Option 2: Slope Stabilization by Revegetation

A second remedial option involves revegetation of the slopes of the landslide complex to allow increased
cohesion for the surface soils, moisture control, and addition of overall stability to prevent future slope
movement. This option involves planting mature (as plausible) Douglas fir trees on the upper hillslopes
and mature (as plausible) western red cedar trees on the lower hillslopes, where there is a high seasonal
water table in the St. Martin Series soils. Correlating the Douglas fir planting to the Steever Series soils
and western red cedar to the St. Martin series soils may be a good approach, but will require a soil
scientist for soil interpretation. This remedial action should be conducted in the summer season when the
slopes are dry, and the water table has receded. In addition to the fir and cedar planting, smaller ground
cover, such as willows or grasses, should be planted to reduce erosion. This will serve to stabilize the
hillslope through cohesion created by the root systems and to decrease the amount of soil moisture present
in the months of soil saturation.

The slope revegetation remedial option provides a low cost option, with a significant factor of safety
improvement over the option of no remedial action. To improve the effectiveness of this option, some
level of drainage improvement or slope dewatering should be performed to reduce the chance of
additional slope failure on the slides below the lower road. Preferential drainage channels should be


                                                                               Portland State University
Lower Little Wind River North Bank Landslides                                                     Page 28 of 33
Skamania County, Washington                                                                      March 22, 2007


developed below the ephemeral stream, working with the natural pathways and trends of surface drainage
to minimize the potential for future soil saturation. Due to the impacts of the local elk population on prior
plantings as described by Tova Cochrane (2007), additional protection of the species planted will be
required, particularly for the cedar species. The remedial measure of slope revegetation is recommended
as a minimum action for mitigation of the lower Little Wind River north bank landslides. A hazard not
addressed comprehensively by this method is the potential for reactivation of the upper landslides due the
removal of resisting forces resulting from the riverbank erosion at the base of the slope.

7.3      Option 3: Slope Stabilization by Dewatering

A third remedial option that would provide a more effective means of stabilizing the slopes, especially
those slides downslope of the lower road, would be dewatering of the upper portion of the slopes through
drainage mitigation. This level of remediation is necessary to reduce the amount of soil moisture on the
upper slopes, reducing the weight of the slopes, and therefore increasing the Factor of Safety by reducing
the driving forces (Burns, 2007). Drainage of the upper slopes would be accomplished through the
installation of trench drains in locations above the slopes where seeps, ephemeral streams, and ponding
have been observed. Installation of trench drains is a relatively low cost method of remediating landslides
that have high groundwater tables, as is the case with the slides within the Lower Little Wind River North
Bank Landslide Complex (Cornforth, 2005). The suggested locations for trench drain installation are
presented in Figure 23.

Within this landslide complex, high groundwater surface expression has been observed along the lower
road in the form of seeps, ephemeral streams, and ponding. Trench drains are highly effective in
remediating shallow landslides (depths of 7 meters (23 ft.) or less) where drains fully penetrate the
landslide debris into the stable ground beneath (Cornforth, 2005). The preferred method of trench drain
installation for this slope will be as interceptor drains oriented across the slope, installed along the lower
road in the zones of groundwater expression at the head of the landslides as shown in Figure 23 (Burns,
2007; Cornforth, 2005). Trench Drain #1 will be approximately 30 meters (98 ft.) in length along the
lower road above the head scarp of Lower Slide #4. Trench Drain #2 will be approximately 40 meters
(131 ft.) in length along the at the lower road above the head scarp of Lower Slide #2 and Lower Slide #1.
Each trench drain will be approximately 1 meter (3.28 ft.) wide and 1.5 to 2 meters (4.9 to 6.5 ft.) in
depth. A greater stability margin may be provided with increased depth of up to approximately 4 meters
(13 ft.) in this case, if the stability of the upper slopes is also taken into consideration. Additional trench
drains could also be installed above the upper slide head scarps in order to provide greater protection
against future mass movement of Upper Slides A and B.

Each trench will be lined with geotextile fabric prevent infiltration of fines and contain a 15.24 to 30.5
centimeter (6- to 12-inch) diameter flexible, perforated pipe at the bottom of the trench (Burns, 2007;
Cornforth, 2005). The bottom of the trench drain should be graded so that no ponding will occur within
the drain itself (Cornforth, 2005). Following installation of the perforated drainpipe, trenches should be
backfilled with free draining rock fill, such as 0.63 to 1.27 centimeter (0.25- to .50-inch) pea gravel
(Burns, 2007; Cornforth, 2005). It is important that the rock fill used contain a minimum of fines (less
than 3%) (Burns, 2007; Cornforth, 2005). Finally, from the downgradient end of each trench drain, flex-
tubing (10.16 centimeter (4”) diameter) should be extended to the river, in order to expel the water
collected by the drain system (Burns, 2007). The pipe to the river is an important element of the system
and can be run above or below the surface (Burns, 2007). An example of a trench drain is provided in
Figure 24 (Cornforth, 2005).




                                                                               Portland State University
Lower Little Wind River North Bank Landslides                                                    Page 29 of 33
Skamania County, Washington                                                                     March 22, 2007




The trench drain remedial option provides an intermediate cost option, with a significant factor of safety
improvement over the first two options described. In addition to the trench drains, it is recommended that
these installations be followed by slope revegetation, as described previously (Burns, 2007). To improve
the effectiveness of this option, placement of a boulder buttress at the base would provide an even greater
factor of safety and stability margin (Burns, 2007). This would also address the hazards not addressed
associated with the potential reactivation of the upper landslides due the erosion the riverbank buttress at
the base of the slope. However, the necessity of road excavation to the lower slopes in order to place the
boulders would likely be a cost-prohibitive option. In addition to the suggested remediation, because the
large Douglas fir tree on Slide #3 appears to be adding stability to the slope, it is important that this tree
remain in place (Burns, 2007).




                                                                               Portland State University
Lower Little Wind River North Bank Landslides                                                   Page 30 of 33
Skamania County, Washington                                                                    March 22, 2007


7.4      Option 4: Slope Regrading, Buttressing, and Rockfill

While the previous two remedial options (excluding the “No Action” option) may be effective in
stabilization of these slopes if the future land use for the vicinity is to remain the same (undeveloped), a
third remedial option is presented as a more comprehensive solution to the mitigation of this hazard.
Although at a higher cost than the other options presented, the option of buttressing the slope base,
regrading the slope, and backfilling the slope with filter materials would allow the greatest possible factor
of safety of the options presented in this report. This scenario involves regrading slope to an angle of 30º
or less. This will reduce the driving forces of the landslide by removing weight from the upper portion of
the slope and increase the resisting forces for the lower portion of the slope (Cornforth, 2005). A buttress
made of large boulders would then be constructed at the base of the slope to support the base and further
increase the resistance to sliding (Cornforth, 2005). The native soil of the slope would then be covered
with a geotextile filter fabric. This will serve as a barrier between the native soil and the rockfill to be
placed above, preventing migration of fines from one layer to the other (Cornforth, 2005). Finally, a free-
draining rockfill, such as a well-graded gravel ( 6.35 centimeter minus (2.5 inch minus)), should be
installed on the slope and compacted in 30.48 to 60.96 centimeter lifts (12 to 24 inch lifts) (Burns, 2007;
Cornforth, 2005). If desired, topsoil can be placed over the rockfill and grasses or other vegetation can be
planted on the slopes.

This final option offered represents a rough order of magnitude cost much greater than that of the
previous options. In addition to the costs of the excavation and materials associated with engineering this
slope alteration, costs of developing a temporary road to allow for placement of the rockfill and buttress
will be significant. However, this option is the only option presented that would permit future
development in the vicinity of these slopes. This could be a major consideration with regard to the
increasing land values and recreational development occurring with the Columbia River Gorge.

8.0      CONCLUSION

In conclusion, in November 2006, a small landslide on the south facing slope of the Lower Little Wind
River North Bank Landslide Complex reactivated. This landslide complex has scarp heights of up to 4
meters (13.12 ft.), scarp widths of 38 meters (124.7 ft.), and total length of up to 90 meters (295.3 ft.).
The maximum flank width is 76 meters (249.3 ft.) and the maximum length from upper scarp to toe is 90
meters (295.3 ft.). The approximate volume of material displaced by these landslides was calculated to be
approximately 59,459 cubic meters (77,770 cubic yards). Within this complex, the landslides are
subdivided into two larger landslides, both classified as earthflows. Contained within the two larger
landslides, four smaller landslides are found on the lower slopes along the Little Wind River. Five major
elements combined to cause these landslides to occur including (1) high rainfall, (2) human impact, (3)
riverbank erosion, (4) steep slopes, and (5) geologic environment. Due to the ecological implications of
additional mass wasting, remedial options must be considered to prevent future reactivation of these
landslides. These include: placing temporary impermeable barriers over the upper slopes, revegetation
with trees, installation of trench drains to prevent accumulation of water above the scarps, slope
buttressing, slope regrading, and free-draining rockfill placement.

9.0      FUTURE WORK

Due to time constraints, the scope of the Lower Little Wind River North Bank Landslide Complex
assessment was limited to the aspects presented in this study. Several additional study aspects could be
explored for this site if additional work were to be performed. The scope of this project was limited to
this section of the Little Wind River north bank; however, additional landslide activity has occurred
further east along the lower road. These landslides could be mapped and assessed as part of a separate
study or an addendum to this report.


                                                                              Portland State University
Lower Little Wind River North Bank Landslides                                                     Page 31 of 33
Skamania County, Washington                                                                      March 22, 2007



If a more exhaustive study of this landslide were to be performed, soil borings to determine the failure
plane, piezometers to determine exact piezometric surface and groundwater gradient, as well as
inclinometers to quantify and monitor slope movement, would be installed in linear array along the slope.
Within the landslide complex, additional slope modeling could be performed to determine the individual
Factor of Safety for Upper Landslide B and each of the lower slides. Furthermore, while the major
landslides of this complex were analyzed in this study, at least one smaller slide is present between Lower
Landslide #2 and Lower Landslide #3. This slide could be examined in a report addendum for more
comprehensive coverage of the study area. Finally, the results and recommendations of this assessment
could be combined with those of the concurrent south bank assessment for a comprehensive assessment
of the lower Little Wind River valley.

10.0     LIMITATIONS

Our services have been performed, our findings obtained, and our recommendations prepared in accordance
with customary principles and practices in the field of engineering geology. Portland State University is not
responsible for the independent conclusions, opinions or recommendations based on the information
presented in this report.

In preparing this report, the authors have relied upon and presumed certain information (or absence thereof)
about the site and adjacent properties provided by the Wind River Watershed Council, and others identified
herein. Except as otherwise stated in the report, Portland State University and the authors of this report have
not attempted to verify the accuracy or completeness of any such information.

The data reported and the findings, observations, and conclusions expressed in the report are limited by the
scope of work, including the lack of subsurface exploration and other tests. The scope of work was defined
through the requests of the Wind River Watershed Council, the time and budgetary constraints present, and
the availability of access to the site.

11.0     REFERENCES

Allen, J.E., 1984, The Magnificent Gateway: A Layman's Guide to the Geology of the Columbia River
Gorge, Timber Press, Forest Grove, Oregon.

Berri, D.A. and Korosec, M.A, 1983, Geological and Geothermal Investigation of the Lower Wind River
Valley, Southwestern Washington Cascade Range, Washington Department of Natural Resources:
Division of Geology and Earth Resources, Olympia, Washington, 48 pp.

Burns, S.F., 1998(a), “Landslide Hazards in Oregon,” in S.F. Burns, editor, Environmental, Groundwater
and Engineering Geology: Applications from Oregon, Star Publishing, Belmont, California, pp 303-315.

Burns, S.F., 1998(b), “Landslides in the Portland Area Resulting from the Storm of February, 1996,” in
S.F. Burns, editor, Environmental, Groundwater and Engineering Geology: Applications from Oregon,
Star Publishing, Belmont, California, pp 303-315.

Burns, S.F., 2007, Personal Communication: Conversation regarding the Little Wind River north bank
landslides.

Cochrane, Tova, 2007, Written Communication: Email correspondences regarding the Little Wind River
north bank landslides.



                                                                               Portland State University
Lower Little Wind River North Bank Landslides                                                 Page 32 of 33
Skamania County, Washington                                                                  March 22, 2007


Cochrane, Tova, and White, Jim, 2005, "Wind River Watershed Restoration Project; Underwood
Conservation District", 2004-2005 Annual Report, Project No. 199801900, 17 electronic pages,
(BPA Report DOE/BP-00005480-3)

Cornforth, D.H., 2005, Landslides in Practice: Investigation, Analysis, and Remedial/Preventative
Options in Soils, Wiley, Hoboken, New Jersey.

Cruden, D.M., and Varnes, D. J., 1996, - “Landslide Types and Processes,” in Turner A.K. and Shuster
R.L., editors, Landslides: Investigation and Mitigation, Transportation Research Board, Special Report
247, pp. 36-75.

Gundersen, Daniel, 2007, Written Communication: Email correspondences regarding the Little Wind
River north bank landslides.

Haagen, Edward, 1990, Soil Survey of Skamania County Area, Washington. Soil Conservation Service,
National Printing Office, Washington, D.C.

Korosec, M. A., compiler, 1987, Geologic map of the Hood River Quadrangle, Washington and Oregon:
Washington Division of Geology and Earth Resources Open File Report 87-6, 40 p., 1 plate, scale
1:100,000.

National Oceanic and Atmospheric Administration (NOAA), 2007, Observed Weather Forecast Database,
www.wrh.noaa.gov/pqr, Portland, Oregon.

Neuman, Margaret, 2007, Written Communication: Email correspondences regarding the Little Wind
River north bank landslides.

Palmer, L.A., 1977, “Large Landslides of the Columbia River Gorge, Oregon and Washington,” in D.N.
Coates, ed., Landslides, Geological Society of America, Reviews in Engineering Geology, vol. III, p. 69-
83, 1977.

UNESCO Working Party on World Landslide Inventory (WP / WLI), 1990, “A suggested method for
reporting a landslide.” Bulletin of the International Association of Engineering Geology, No. 41, 5-12.


United States Department of the Interior – Geologic Survey, 1994, Carson Quadrangle, Washington-
Oregon, 7.5-Minute Series (Topographic), U.S. Geological Survey, Denver, Colorado.

Washington Department of Fish and Wildlife (WDFW), 2006, “Little Wind River Protection Project”,
Landowner Incentive Program - Application for Project Funding 2005/2006 for Daniel Gunderson,
WDFW, Olympia, Washington

Washington Department of Wildlife (WDW), Confederated Tribes and Bands (CTB), and Washington
Department of Fisheries (WDF), 1990, Wind River Subbasin Salmon and Steelhead Production Plan,
Columbia Basin System Planning, NPPC, Portland, OR.

Western Region Climate Center (WRCC), 2007, Monthly Precipitation Database – Bonneville Dam,
www.wrcc.dri.edu, WRCC, Reno, NV.

Wind River Watershed Council, 2007, “Monthly Meeting Minutes For January 17, 2007”



                                                                             Portland State University
Lower Little Wind River North Bank Landslides                                            Page 33 of 33
Skamania County, Washington                                                             March 22, 2007


Williamson, D.A., Neal, K.G., and Larson, D.A., 1991, The Filed-Developed Cross-Section: A systematic
Method of Portraying Dimensional Subsurface Information and Modeling for Geotechnical Interpretation
and Analysis. Association of Engineering Geologists, Proceedings, 34th Annual Meeting, pp. 719-738.




                                                                        Portland State University

				
DOCUMENT INFO