GR_12_14__Geotech_Report_Segment_D.pdf

Geotechnical Report Historic Columbia River Highway

State Trail - Segment D Hood River, Oregon

Report - December 2014

Columbia River Gorge Commission 1988

FEDERAL HIGHWAY

ADMINISTRATION

Report to:

CH2M Hill

2020 SW 4th Avenue

Portland, OR 97201

GEOTECHNICAL REPORT

HISTORIC COLUMBIA RIVER

HIGHWAY STATE TRAIL

SEGMENT D (LINDSEY CREEK TO STARVATION CREEK)

December 2014

Submitted by:

Cornforth Consultants, Inc.

10250 SW Greenburg Road, Suite 111

Portland, OR 97223

December 2014 i Cornforth Consultants, Inc.

TABLE OF CONTENTS PAGE

1. INTRODUCTION

1.1 General

1.2 Scope of Work

1.2.1 Task 5.7 – Subsurface Investigation

1.2.2 Task 5.8 – Engineering Analysis and Geotechnical Report

2. SITE AND PROJECT DESCRIPTION

2.1 General

2.2 Trail Reaches

2.2.1 Lindsey Creek Drainage – Segment D (151+50 to 155+00)

2.2.2 Dog Hair Forest – Segment D (155+00 to 172+00)

2.2.3 Warren Creek Drainage – Segment D (172+00 to 189+00)

2.2.4 Moderate Vegetated Talus – Segment D (189+00 to 195+00)

2.2.5 Steep Vegetated Talus – Segment D (195+00 to 219+00)

3. GEOLOGY

3.1 General

3.2 Geologic Setting

3.3 Regional Geology

3.4 Site Geology

3.4.1 Unconsolidated Geologic Units

3.4.2 Bedrock Geologic Units

3.4.3 Structural Geology

3.5 Geologic Hazards

3.5.1 Seismic Hazards

3.5.2 Volcanism

3.5.3 Landslide / Mass Movement

3.5.4 Debris Flows / Flooding

3.5.5 Talus

3.5.6 Rockfall

4. FIELD EXPLORATIONS

4.1 General

4.2 Test Pits

December 2014 ii Cornforth Consultants, Inc.

4.3 Borings

5. LABORATORY TESTING

5.1 General

6. SUBSURFACE CONDITIONS

6.1 Trail Reaches

6.1.1 Lindsey Creek Drainage – Segment D (151+50 to 155+00)

6.1.2 Dog Hair Forest – Segment D (155+00 to 172+00)

6.1.3 Warren Creek Drainage – Segment D (172+00 to 189+00)

6.1.4 Moderate Vegetated Talus – Segment D (189+00 to 195+00)

6.1.5 Steep Vegetated Talus – Segment D (195+00 to 219+00)

7. GEOTECHNICAL DESIGN RECOMMENDATIONS

7.1 General

7.2 Seismic Design Parameters

7.3 Cut and Fill Slopes

7.4 Fill Materials

7.5 Rockeries

7.6 Geotextiles

7.7 Wet Weather Construction

7.8 Lindsey Creek Drainage – Segment D (151+50 to 155+00)

7.8.1 Anticipated Conditions

7.8.2 Retaining Walls

7.8.3 Settlement

7.8.4 Dewatering & Excavation Shoring

7.9 Dog Hair Forest – Segment D (155+00 to 172+00)

7.9.1 Anticipated Conditions

7.9.2 Settlement

7.10 Warren Creek Drainage – Segment D (172+00 to 189+00)

7.10.1 Anticipated Conditions

7.10.2 Settlement

7.10.3 Warren Creek Bridge

7.10.4 Retaining Walls

7.10.5 Lateral Resistance

7.10.6 Subdrainage

7.10.7 Dewatering & Excavation Shoring

December 2014 iii Cornforth Consultants, Inc.

7.11 Moderate to Steep Talus Slopes – Segment D (189+00 to 219+00)

7.11.1 Anticipated Conditions

7.11.2 Settlement

7.11.3 Retaining Walls

8. CONSTRUCTION CONSIDERATIONS

8.1 Construction Surcharges

8.2 Groundwater Management

8.3 Obstructions

8.4 Rockeries

9. REFERENCES

LIST OF FIGURES

1. Segment D – Vicinity and Location Map

2. Geology Map

3. Geologic Hazards –Summit Creek to Starvation Creek

4. Design Details – Rock Inlay

5. Cut-Side Rockery Typical Section

6. Fill-Side Rockery Typical Section

7. Design Details – Lindsey Creek Wall

8. Design Details – Warren Creek Walls

APPENDICES

Appendix A Segment D DAP Plans prepared by CH2M Hill Appendix B Geologic Narrative prepared by CH2M Hill Appendix C Summary Exploration Logs and Foundation Data Sheets Appendix D Geotechnical Report by Allegiance Group Appendix E Laboratory Testing Data Sheets Appendix F Wonder Creek Flow Through Embankment Design

December 2014 iv Cornforth Consultants, Inc.

EXECUTIVE SUMMARY

Western Federal Lands Highway Division (WFLHD) is leading the design of a portion of the Historic Columbia River Highway (HCRH) Trail between Wyeth and Hood River. This report presents geotechnical design recommendations for Segment D of the trail between Lindsey Creek and Starvation Creek Trailhead.

A geologic review and subsurface explorations indicate the alignment is underlain by primarily granular soils. It is anticipated that surface soils, excluding topsoil, would be suitable for reuse as embankment fill. Surficial soils generally contain sufficient fines to be considered moisture sensitive, which will limit construction operations and the time of year that construction could be performed. It is anticipated that groundwater will impact temporary excavations as well as grading operations.

New trail segments are located along Interstate 84 and on the slopes south of the highway.

Embankments and retaining walls are proposed to maintain design grades and enhance user experience along the proposed alignment. Stability and construction impacts of retaining walls were evaluated using information available from a literature review and subsurface explorations. In general, gravity retaining structures (rockeries and MSE walls) proposed for the project have static and seismic factors-of-safety that meet AASHTO standards. One rockery near 185+00 is located on a steep slope, and has static and seismic factors of safety below AASHTO standards. An MSE wall could be constructed to meet AASHTO standards for global stability, but would require a deep subexcavation. The retaining wall at Lindsey Creek is located in the shoulder of Interstate 84, and may require excavation shoring to avoid encroaching on the travel lane.

Segment D contains one bridge at Warren Creek. Shallow foundations appear feasible at this location. Perched groundwater at Warren Creek will likely require dewatering. Due to the proximity of abutments to the existing stream channel, excavation shoring will likely be required to construct the bridge foundations.

December 2014 1 Cornforth Consultants, Inc.

1. INTRODUCTION

1.1 General

The Federal Highway Administration Western Federal Lands Highway Division (WFLHD) has retained CH2M Hill to provide biological baseline reconnaissance and preliminary trail design for the Historic Columbia River Highway State Trail Project. CH2M Hill retained Cornforth Consultants, Inc. to assist with geotechnical aspects of the preliminary trail design for Trail Segments A through E (Wyeth to Mitchell Creek). This report presents the results of field explorations along the proposed alignment for Segment D (Lindsey Creek to Starvation Creek Trailhead) and provides geotechnical design recommendations for retaining structures and stream crossing structures.

1.2 Scope of Work

Detailed descriptions of work tasks are outlined in the Statement of Work for Task Order T-13-001 of IDIQ Contract DTFH70-10-D-00020 dated February 26, 2013. The following paragraphs summarize geotechnical elements included in the Statement of Work that are addressed in this report.

1.2.1 Task 5.7 – Subsurface Investigation

Test pits, exploratory borings, and geologic mapping were completed to investigate soil and bedrock conditions at selected locations along the trail. In general, exploration locations were completed as outlined in a geotechnical exploration plan prepared under a previous phase of the work. Samples collected during explorations were delivered to the WFLHD testing laboratory in Vancouver, Washington for index and strength testing.

1.2.2 Task 5.8 – Engineering Analysis and Geotechnical Report Based on the proposed trail alignment and improvements, this geotechnical report was prepared to summarize geotechnical design recommendations. Task 5.8 also included preparation of foundation data sheets for explorations in Segment D.

December 2014 2 Cornforth Consultants, Inc.

2. SITE AND PROJECT DESCRIPTION

2.1 General

The Historic Columbia River Highway (HCRH) was constructed between 1913 and 1922 and was America’s first scenic highway. By the late 1940s and early 1950s, the initial construction of a water-level route that would become I-84 obliterated many sections of the HCRH, and left upland sections disconnected. The Columbia River Gorge National Scenic Area Act of 1986 directed the State of Oregon to connect the abandoned sections of the HCRH as a pedestrian and cyclist trail.

Since 1987, the Oregon Department of Transportation has been working with stakeholders including the HCRH Advisory Committee, Oregon Parks and Recreation Department, the State Historic Preservation Office, and Travel Oregon to restore the highway.

The HCRH State Trail Plan focuses on 11 miles of the trail between Exit 51 on I-84 (Wyeth) and Ruthton Park. The proposed alignment is subdivided into the following segments:

Segment A – Wyeth to Shellrock Mountain Segment B – Shellrock Mountain to Summit Creek Segment C – Summit Creek to Lindsey Creek Segment D – Lindsey Creek to Starvation Creek Segment E – Viento to Mitchell Creek Segment F – Mitchell Creek to Mitchell Point Tunnel Segment G – Mitchell Point to Ruthton Point Segment H – Ruthton Point to Ruthton Park

An overview of the proposed trail alignment is shown on the Vicinity Map in Figure 1. Figure 2 shows the proposed trail alignment overlaid on a geologic map of the area. A more detailed view of the Segment D alignment and the anticipated geologic hazards is presented in Figure 3. Plans for the Design Acceptance Package (DAP) were prepared by CH2M Hill for Segment D, and are included in Appendix A. The DAP plans show the proposed trail alignment and the location of proposed retaining and stream crossing structures.

2.2 Trail Reaches

Segment D was subdivided into geotechnical reaches with similar geologic conditions, topographic features, and/or proposed improvements. The following sections describe the existing conditions and the improvements proposed for the individual trail reaches.

2.2.1 Lindsey Creek Drainage – Segment D (151+50 to 155+00)

Alignment Overview. This reach of the trail roughly follows the shoulder of I-84 across the relatively flat base of the Lindsey Creek drainage. There is an existing concrete box culvert crossing the trail alignment between 152+25 and 152+40. The Lindsey Creek drainage is moderately vegetated with medium to large deciduous trees and low-lying brush.

Proposed Improvements. Fills up to 3 feet above existing grade with side slopes up to 2H:1V are proposed in this reach. Retaining Wall “W1” spans the existing concrete box culvert that carries Lindsey Creek under I-84. The wall is a concrete cantilever wall with a footing/moment slab

December 2014 3 Cornforth Consultants, Inc.

designed to avoid placing new load on the existing box culvert structure. Wall W1 is located on the right side of the trail approximately 8 feet from centerline, and measures approximately 92 feet long.

It begins near 151+88 at approximately 4 feet high, including 3 feet of embedment. Wall W1 increases to a maximum height of 10 feet, including 3 feet of embedment, on the west side of the box culvert. Between 152+20 and 152+45, Wall W1 measures approximately 7 feet high, and is supported on a footing/moment slab that spans the existing box culvert. On the east side of the box culvert, Wall W1 tapers from approximately 10 feet high to approximately 4 feet high at 152+80 where it terminates.

2.2.2 Dog Hair Forest – Segment D (155+00 to 172+00)

Alignment Overview. At the eastern edge of the Lindsey Creek drainage, the trail veers toward I-84 and runs along the base of a 1H:1V cut slope for approximately 450 feet. The cut slope is densely vegetated with small to medium evergreen trees. Between 159+50 and 161+00 the trail veers into a small drainage before returning to the shoulder of I-84. Between 161+00 and 170+00 the trail runs along the base of a densely wooded 1.5H:1V cut slope. At 170+00 the trail veers south and joins the alignment of an existing BPA service road and runs along the service road to the end of the reach near 172+00.

Proposed Improvements. Fills up to 6 feet above existing grade with 3H:1V side slopes are proposed in this reach. Most fills are 2 to 3 feet above existing grade.

2.2.3 Warren Creek Drainage – Segment D (172+00 to 189+00)

Alignment Overview. Between 172+00 and 175+50 the trail follows what appears to be an old alluvial terrace deposit. The top of the terrace is approximately 15 feet above the bottom of the drainage. The slope between the top of the terrace and the bottom of the drainage is approximately 1.5H:1V. At 175+50 the trail descends off of the terrace deposit into the bottom of the drainage.

The bottom of the drainage slopes gently down toward Warren Creek between 2 and 5 degrees from horizontal. The basin is densely wooded with medium to large deciduous and evergreen trees. The proposed alignment crosses Warren Creek near 177+90. Warren Creek flows in a shallow, narrow channel 10 to 20 feet wide and 2 to 3 feet below the surrounding ground. Near 184+00 the trail ascends out of the bottom of the drainage to follow the base of a vegetated talus slope. The proposed alignment roughly follows an existing footpath to the end of the reach at 189+00. The areas on either side of the footpath are densely wooded with large evergreen trees.

Proposed Improvements. Between 172+00 and 175+50, the trail follows the existing terrace and utilizes 2H:1V cuts and sliver fills less than 3 feet from existing grade on the right side to achieve the required trail footprint. Once the trail descends off the terrace bench, fills up to 8 feet above existing grade with side slopes varying between 2H:1V and 3H:1V are proposed. Near 184+00, the trail transitions to a cut section with cuts between 1 and 7 feet below existing grade. Cut slopes between 1.5H:1V and 2H:1V are proposed on the high side of the trail. A low-side fill rockery is proposed between 184+75 and 185+10 where the trail crosses a narrow drainage. The rockery has a maximum exposed height of 4 feet and is founded on an existing 1.5H:1V slope.

The proposed bridge at Warren Creek is a single-span structure supported on shallow foundations.

Abutment fills are retained by four cast-in-place concrete wing walls that run parallel to the bridge

December 2014 4 Cornforth Consultants, Inc.

approximately 9 feet from centerline. The wing walls are 14 to 15 feet tall, including minimum embedment of 9 to 10 feet at the ends of the bridge. Retaining walls and bridge foundations bear at elevation 87 feet for hydraulic and scour considerations.

2.2.4 Moderate Vegetated Talus – Segment D (189+00 to 195+00) Alignment Overview. The proposed trail roughly follows an existing footpath between 189+00 and 192+00 as it approaches I-84. The footpath follows the base of a densely wooded slope that drops down to the northwest between 2H:1V and 4H:1V. Near 192+50 the trail begins to run along the shoulder of I-84 and the off ramp to the Starvation Creek Trailhead. There is a densely vegetated talus slope south of the trail that drops to the north at approximately 1.5H:1V.

Proposed Improvements. Fills up to 12 feet above existing grade with 2H:1V slopes are proposed in this reach of the trail to provide trail users a view point for the historic ovens on the left side of the alignment. Between 189+00 and 191+40, the embankment fill increases from zero to approximately 12 feet on the left side of centerline before transitioning to Retaining Wall W4. Wall W4 runs from 191+40 to 193+25 and has a maximum height of 14 feet, including 2 feet of embedment, near 191+50. At 193+25, Wall W4 transitions back to an embankment fill that tapers to existing grade near 195+00. Wall W5, a 1H:1V vegetated slope, is proposed to retain the embankment on the right side of the trail between 190+79 and 193+25. Wall W5 extends up to 12 feet above existing grade.

2.2.5 Steep Vegetated Talus – Segment D (195+00 to 219+00)

Alignment Overview. Between 195+00 and 219+00 the trail follows the base of a steep, vegetated talus slope that varies between 2H:1V and 1H:1V. The existing trail is supported on a rubble rockery with an exposed height of approximately 4 feet between 198+00 and 200+50. Near 210+50 there is a 10-foot-tall berm between the trail and the toe of the talus slope. The berm has side slopes slightly flatter than 1H:1V and appears to have been constructed to contain rockfall originating from the cliffs at the top of the talus slope. The rockfall berm extends to approximately 216+00. From 216+00 to 219+00 the proposed trail traverses the base of a near-vertical, 80-foot-tall rock cliff.

Proposed Improvements. This reach of the trail would roughly follow the existing line and grade of the Starvation Ridge Trail #414. In general, proposed cuts and fills are less than 2 feet from existing grade and utilize slopes up to 2H:1V. A 1H:1V vegetated slope up to 5 feet tall would retain the left side of the trail between 197+50 and 198+14. Between 212+60 and 215+30, a high-side rockery with an exposed height of approximately 2 feet would be used on the right side of the trail. The rockery would remove the toe of the existing rockfall catchment berm. Rockeries would also be used to retain fills on the right side of the trail from 217+05 to 217+35 and 218+05 to 218+30. The rockeries are less than 5 feet high.

December 2014 5 Cornforth Consultants, Inc.

3. GEOLOGY

3.1 General

The regional and site-specific geology of the proposed trail alignment was studied by CH2M Hill and used by Cornforth Consultants, Inc. to prepare the geotechnical design recommendations included in this report. In addition to geology of the proposed alignment, CH2M Hill also evaluated potential geologic hazards along the trail. The following sub-sections present excerpts from a geologic narrative prepared by CH2M Hill. A complete copy of the geologic narrative is included as Appendix B.

3.2 Geologic Setting

The project is located in the Columbia Gorge National Scenic Area approximately 50 miles east of Portland, Oregon. The Columbia Gorge was formed when the Columbia River downcut through uplifted basalt flows of the Columbia River Basalt Group. The Columbia Gorge was further eroded and widened by a series of catastrophic floods during the late Pleistocene, which left behind steep basalt cliffs.

3.3 Regional Geology

The geology of the region is dominated by the Columbia River Basalt (CRB) Group. The CRB flows in the Columbia River Gorge are Miocene in age (~17 to 6 million years old) and cover an area of roughly 200,000 km2 in Oregon, Washington, and Idaho. These basalt flows erupted from north- to northwest-trending fissures and vents in northeastern Oregon, eastern Washington, and western Idaho. The combination of large volume and low viscosity of the erupting lavas enabled them to cover large areas and distances. The basalt flowed into lowland areas and followed existing stream or river channels. The CRB flows entered western Oregon and Washington through a broad low area which existed where the northern Oregon Cascades are found today. After cooling, the lava flows have been uplifted and faulted along the length of the Columbia River Gorge.

3.4 Site Geology

Figure 2 presents a geologic map of the site and surrounding vicinity, adapted from Beaulieu, 1977 and Bela, 1982), using GIS data from the Oregon Geologic Data Compilation provided by DOGAMI (Jenks et al, 2008). The geologic units that are shown on the map, plus other geologic units not shown on the map due to scale, are described below.

3.4.1 Unconsolidated Geologic Units

Qal. Quaternary Alluvium: Unconsolidated gravel, sand, and silt, located above floodplains of major streams and valley fill of smaller stream valleys, and includes several terrace levels of varying ages.

Along the proposed alignment this unit is primarily located at the west end between Wyeth and Shellrock Mountain. Alluvium is also deposited along tributary streams in the Gorge that cross the proposed alignment. Most of the younger Quaternary-age alluvium along the Columbia River is flooded beneath the reservoir formed by Bonneville Dam.

Qta. Talus: This unit consists of uniformly sloping, unconsolidated rock and soil debris accumulating at base of cliffs primarily by rockfalls and rock slides. This unit consists of coarse, angular rubble

December 2014 6 Cornforth Consultants, Inc.

near or at the surface with increasing proportions of silt at depth. The estimated thickness generally exceeds 50 feet in many areas. Thinner deposits are widespread but are not mapped on the geologic hazard maps due to scale issues. Other types of slope deposits include locally-derived and gravity-driven colluvium that consists of weakly stratified silt, sand and rock fragments. This unit accumulates on slopes that cover steep bedrock. Thick talus along the proposed alignment is mostly situated along the base of Shellrock Mountain, and on slopes formed from fractured basalt such as Mitchell Point.

Qaf. Alluvial Fans (not shown on the geologic map): Moderately sloping, subaerial deposits of unconsolidated, poorly-sorted, angular gravel and sand occurring as fan deposits at mouths of steep canyons cut into the basalt. These fans have a maximum thickness of up to 50 feet. Smaller fans in the Columbia River Gorge are present but are not indicated on geologic mapping. Their presence can be inferred, however, at the mouths of all torrential flood channels indicated on geologic hazards maps (Beaulieu, 1977).

Qls. Mass Movement/Landslides: Landslides in the area include both deep bedrock slides and slump/earthflow landslides. The deep bedrock slides are characterized by large, down-dropped blocks of bedrock that are both active and inactive. These are formed by faults, joints, and weak interbeds in the bedrock. Slump/earthflows consist of moderately sloping, hummocky terrain with irregular drainage and soil distribution.

3.4.2 Bedrock Geologic Units

Bedrock in the project vicinity consists of basalt flows and intrusive diorite plugs. These are as follows:

Tgn1/Tgr2/Tgn2. Grande Ronde basalt: Dense, hard, dark-gray, jointed, fine-grained basalt flows.

Flows range from 3 feet to more than 160 feet thick, but are generally between 50 and 80 feet thick in the study area. The total thickness is unknown but exceeds 2,000 feet in some areas. There are local thin interbeds between the flows; these weather more quickly. At Mitchell Point, near the east end of the proposed alignment, the basalt includes a cemented basaltic conglomerate with a quartz sandstone matrix. The flows have both normal and reversed magnetic polarity, and are thus subdivided into normal (n) and reverse (r) flow designation. The Grande Ronde basalt crops out along most of the proposed alignment and is exposed in many of the road cuts.

Twf. Frenchman Springs Member of Wanapum Basalt: Dark gray basalt this is typically blocky to columnar-jointed but exhibits entablature-colonnade-style jointing in the Columbia Gorge area. This unit consists of about a dozen flows with normal magnetic polarity. The Frenchman Springs member is exposed near the eastern end of the project, west of Mitchell Point.

Tsp. Saddle Mountains Basalt: Consists of 10 chemically diverse members and is the youngest formation in the Columbia River Basalt Group. Typically the flows are black, dense, and fine grained. These flows have localized thick sedimentary deposits between the flows. This formation overlies the Frenchman Springs member near the eastern end of the project.

December 2014 7 Cornforth Consultants, Inc.

Piqd. Intrusive igneous rocks: Pliocene-age, fine- to medium-grained micro-dioritic plugs, dikes, and irregular intrusive bodies that are dense and coarsely-jointed. Shellrock Mountain is the largest intrusion in the study area and is composed primarily of quartz diorite porphyry.

Other basalt flows and volcanic rocks form the higher mountains around the site and are shown on the geologic map. However, these are not in the project alignment and are thus not described.

3.4.3 Structural Geology

The term geologic structure refers to folds, faults, and joints that affect the disposition and arrangement of geologic units in the study area. The structure is significant to interpreting rock mass characteristics and mass movement potential.

Folds. Major folds in the vicinity include the Columbia Hills anticline which splays into several compressional structures east of the project area. Several smaller unnamed folds are present in the Gorge. The CRB flows dip generally to the east along the proposed project alignment.

Faults. Bela (1982) mapped several northwest-trending dip-slip faults near the eastern end of the proposed alignment. It is likely that, due to tectonic deformation and uplift of the Cascade Mountains, several small faults are present that are not mapped because they are either too small to show or are obscured by vegetative cover. The west side of Mitchell Point is bounded by a fault down-dropped to the west.

Joints. Jointing in bedrock strongly influences mass movement and water movement through bedrock. Joints in the CRB include 6-inch to 12-inch vertical cooling fractures, flat sheeting joints at the tops of individual flows, and large regional joints up to 10 feet wide and several miles long. In addition, field observations showed densely-jointed basalt that forms fractured “dice rock”. The dipping beds of CRB have once-vertical columns that are now dipping at angles perpendicular to the dip of the bedding. The dip angles of these joints will influence rock slope stability and the joint density will influence rockfall potential.

3.5 Geologic Hazards

Based on site visits, aerial photograph reconnaissance, regional earthquake data, soil data, and geologic hazard maps of the area (Beaulieu, 1977; DOGAMI), the geologic hazards that may affect the site include seismic hazards, volcanism, landslide susceptibility, debris flows and flooding, and rockfalls and rockslides. This section describes potential geologic hazards along the proposed alignment. Figure 3 shows a generalized geologic hazard map of the site and surrounding vicinity, adapted from Beaulieu (1977), GIS data provided by DOGAMI, and reconnaissance-level site observations.

3.5.1 Seismic Hazards

Seismic hazards include ground shaking and ground rupture caused by earthquakes, surface fault rupture, soil liquefaction and lateral spread resulting from ground shaking, and earthquake-induced landslides and rockfalls.

December 2014 8 Cornforth Consultants, Inc.

3.5.2 Volcanism

The Cascade Mountains are home to a large number of active volcanoes. The closest ones to the Project area are listed below, with approximate distances to the project site:

Mount St. Helens – 40 miles northwest Mount Hood – 22 miles south Mount Adams – 35 miles north

Mount Hood is the closest volcano to the site, and has erupted as recently as the 1800s. Mount Hood has had two significant eruptive periods recently, one about 1,500 years ago and the other about 200 years ago (Gardner, et al., 2000). These eruptions consisted of steam and ash explosions, pyroclastic flows, lahars in valleys, and minor tephra falls. Lahars swept down the Sandy River drainage approximately 1,500 years ago.

Hazards from Mount Hood include proximal hazards and distal hazards. Proximal hazards include lava flows, pyroclastic flows, debris avalanches, and lahars. Distal hazards include lahars that travel a great distance down river valleys and ashfalls. The proposed alignment is not within the Mount Hood Hazard Zonation Map for proximal or distal hazards. However, depending on the prevailing wind direction at the time of the eruption and the source of the eruption, ash fallout in the region surrounding the project may occur.

The Mount Hood Coordination Plan has been developed to coordinate efforts among governmental agencies in the event of volcanic unrest at Mount Hood (Mount Hood Facilitating Committee, 2005).

This plan provides vital Mount Hood volcanic event response information for the areas that will be most affected by a volcanic event.

Mount St. Helens has erupted more than once in the last 200 years (USGS, 2011). During the past 4,000 years, Mount St. Helens has erupted more frequently than any other volcano in the Cascade Range. The May 18, 1980, eruption was the most economically destructive volcanic event in U.S.

history. The volcanic ash cloud from this eruption drifted east across the United States in 3 days and encircled the globe in 15 days. The ash fall caused problems for transportation operations, decreased visibility that closed highways and roads, and disrupted air transportation. The fine-grained, gritty ash also caused problems for internal-combustion engines and other mechanical and electrical equipment and caused short circuits in electrical transformers, which in turn caused power blackouts.

The removal and disposal of ash from highways, roads, buildings, and airport runways were monumental tasks for local communities.

Mount Adams has been less active during the past few thousand years than neighboring Mount St.

Helens, Mount Rainier, and Mount Hood. Future eruptions will probably occur more frequently from vents on the summit and upper flanks of Mount Adams (Scott, et al., 1995).

The USGS has established a National Alert Notification System for Volcanic Activity, which consists of advisories, watches, and warnings (USGS, 2011). The alert-notification system has been standardized and the goals are to (1) communicate a volcano’s status clearly to lay persons, (2) help emergency response organization determine proper mitigation measures, and (3) prompt people and businesses at risk to seek additional information and take appropriate actions. In the event of a

December 2014 9 Cornforth Consultants, Inc.

volcanic eruption that could damage or impact the project, the trail would be likely be shut down until safe operating conditions return.

3.5.3 Landslide / Mass Movement

Documentation of potential landslide hazards in the vicinity of the project has been established by reviewing published geologic and landslide hazard reports. Some of these reports focused on identifying documented historical landslides, others focused on identifying landslide topography, and a third group focused on identifying potential landslide areas. The following four primary sources of landslide hazard information, published by DOGAMI, have been used to identify the potential hazards for slope instability along the proposed HCRH Trail:

Statewide Landslide Information Database for Oregon (SLIDO-2) Slope Failures in Oregon: GIS Inventory for Three 1996/97 Storm Events. Oregon

Department of Geology and Mineral Industries, Special Paper 34 Potential Rapidly Moving Landslide Hazards in Western Oregon. (IMS-22) (Hofmeister, 2002) Geologic hazards of parts of northern Hood River, Wasco, and Sherman Counties. Oregon

Department of Geology and Mineral Industries, Bulletin 91 (including Hazard Maps).

A description of each primary publication used to assess landslide hazards is provided below.

Locations of these observed features are summarized below and shown on the geologic hazard maps in Figure 3. Field observations were also conducted to observe slopes and areas of present or future slope failure and are discussed in this section.

SLIDO2. In 2011, DOGAMI released a new publication series called Statewide Landslide Information Database for Oregon (SLIDO-2) (Burns et al., 2011). The purpose of this document was to establish a statewide database of previously mapped landslide-related features. The features included in this database include landslides, debris flows or alluvial fans, and colluvium or talus. The primary sources of this historical landslide information are published geologic reports and geologic hazard studies by the USGS, DOGAMI, and to a lesser extent regional studies by the U.S. National Forests and thesis studies in the state. The landslide database from this study, which is compiled in GIS format, was used to overlay landslides or landslide-related features on the Geologic Hazard Map in Figure 3.

The SLIDO database does not show any large landslide hazards along the proposed alignment. The largest landslide area along the alignment is described as “Talus”, and is discussed in more detail later. A large landslide area is mapped east of Mitchell Point, and landslides are mapped in the steep-sided drainages including Lindsey, Warren, and Perham Creek. Note that this database does not include several small landslides that are likely to exist in the vicinity, given the steep topography and wet climate. In addition, any mass movement that originated from a slope above the proposed trail alignment and ended up on the highway would pose a hazard to the trail and users on it.

Special Paper 34. A second DOGAMI statewide landslide database was created that includes data specifically on landslides that were triggered by three severe storm events in 1996 and 1997. Each storm event received a Federal “major disaster” declaration (Hofmeister, 2000). A storm in February

December 2014 10 Cornforth Consultants, Inc.

1996 led to a Federal disaster declaration in 27 counties, a storm in November 1996 led to this declaration in three counties, and a storm in December 1996/January 1997 led to this declaration in 14 counties. These storms triggered landslides throughout the state of Oregon, except for the flatter southeast portion of the state. The database for this report includes 9,582 slide locations. Information on these landslides was obtained primarily from National Forest System offices, the Federal Emergency Management Agency, the Bureau of Land Management districts, the Oregon Department of Forestry, the Oregon Department of Transportation, Oregon Emergency Management, county public works departments, and major metropolitan public works departments.

Figure 3 shows a generalized geologic hazard map and locations along the proposed alignment where landslides and rockfalls were mapped that were triggered by three severe storm events in 1996 and 1997. Within the proposed alignment, events included flooding at Viento Creek that closed part of the highway, debris flow/flooding at Starvation Creek, and a landslide at Mitchell Point.

Bulletin 91. In Bulletin 91, Beaulieu (1977) discusses existing and potential future landslide hazards.

As discussed previously, Beaulieu (1977) described two types large mass movements including deep bedrock slides and earthflow/slump topography. The deep bedrock slides are characterized by large, down-dropped blocks of bedrock that are both active and inactive, and are controlled by faults, joints, and weak interbeds. Potential geologic hazards related to these slide blocks include continued sliding, variable foundation strength, variable cut slope stability, and poor drainage. The potential for development or land use changes in these areas is highly variable. Large slide blocks were recognized by Tolan et al (1984) and have been documented in the Columbia Gorge at locations such as Rooster Rock, just east of Vista House. In heavily forested areas with irregular topography sculpted by catastrophic floods, these slide blocks are difficult to recognize.

Slump/earthflows consist of moderately-sloping terrain with irregularities of slope, drainage, and soil distribution. Recent movement can be shown by tension cracks, bowed trees, and other signs of slope instability. Potential geologic hazards related to slump/earthflows include on-going movement and creep, low cut slope stability (slides can easily be destabilized), variable foundation strength, and poor drainage. The potential for development or land use changes in these areas is highly variable, and changes to vegetation or slope angles can reactivate or accelerate sliding.

Beaulieu (1977) notes that areas less than 10 to 20 acres in size are not shown on his geologic hazard maps, and it is likely that several unmapped landslides could be present in the project vicinity.

Delineation of these slide areas and potential hazards to the proposed alignment would require detailed geologic mapping, subsurface investigations, and instrumentation to evaluate ongoing movements.

3.5.4 Debris Flows / Flooding

Debris flows are fast-moving landslides that occur in a wide variety of environments throughout the world. They are particularly dangerous to life and property because they move quickly, destroy objects in their paths, and often strike without warning. Debris flow hazards exist in the vicinity of the project alignment. A damaging debris flow occurred in the Columbia River gorge near the town of Dodson, Oregon, (approximately 15 miles west of the proposed project alignment) during a rainfall and snowmelt event in February 1996. The debris flow was deposited in a number of separate

December 2014 11 Cornforth Consultants, Inc.

events over a period of 2 days. The flows originated high on the palisade and traveled through the steep canyon. Boulders, mud, and debris from these and several nearby debris flows were deposited across the east-bound lanes of I-84, and blocked the highway for 5 days.

Smaller alluvial fans deposited by debris flows in the Columbia River Gorge are present but are not indicated on geologic hazard mapping. Their presence can be inferred, however, at the mouths of all torrential flood channels indicated on geologic hazards maps (Beaulieu, 1977). Alluvial fan deposits are subject to torrential flooding, extremely rapid erosion, and also deposition. Numerous stream channels cross the proposed alignment and several of these could be subjected to debris flows, flooding and alluvial fan deposition.

IMS 22 – Rapidly Moving Landslide Hazard Mapping. DOGAMI published another useful document (IMS-22) entitled “Potential Rapidly Moving Landslide Hazards in Western Oregon” (Hofmeister, 2002). This study was developed as the result of a cooperative effort among DOGAMI, the Oregon Department of Forestry, the Oregon Department of Land Conservation and Development, the Earth Systems Institute, and other landslide researchers. In this study, rapidly moving potential landslides (also called debris flows) are identified based on GIS modeling, which primarily used slope steepness and slope curvature data to identify locations where landslides could occur and the extent of predicted debris flow movement. This modeling was calibrated with results from targeted field explorations and known historical earthquakes. This study only focused on rapidly moving landslides and does not attempt to predict the locations of any other types of landslides.

Areas mapped as having the potential for rapidly moving landslides/debris flows based on the IMS- 22 data are not shown in Figure 3, because according to IMS-22 the proposed alignment is entirely within high rapidly moving/debris flow/flood hazards due to steep topography.

Bulletin 91. Beaulieu (1977) listed several creeks as having torrential flooding potential. These are areas with high probability for floods characterized by rapidly-flowing water with high channel and stream-bank erosion potential in narrow canyons with little or no floodplain. In general these restricted to short, high-gradient streams flowing through steep terrain or high relief into the Columbia River. The four locations in Segment D include Lindsey Creek (152+00), Warren Creek (178+00), Cabin Creek (202+00), and Starvation Creek (220+00).

Alluvial fans have been deposited by repeated debris flow events at the mouths of numerous drainages along the proposed alignment, including Lindsey, Unnamed, Warren, Starvation, Viento, Perham and Mitchell. These were observed during the field reconnaissance and also are observable on aerial photographs and LIDAR imagery. In addition, several alluvial fans that indicate past debris flows and floods are likely present but were submerged under water when Bonneville Dam was constructed.

3.5.5 Talus

Based on a review of the SLIDO-2 hazard mapping, the proposed alignment would cross talus deposits for approximately ½ mile. The proposed alignment would cross thick, rocky talus deposits at the base of Shellrock Mountain between Stations 65+00 and 92+00. The talus slope on Shellrock Mountain averages close to 34 degrees above horizontal. However, locally steeper areas up to 40 or

December 2014 12 Cornforth Consultants, Inc.

more degrees were observed. It is anticipated that at this slope angle, the mode of travel for most rocks down the slope would be initially by rolling, but after a certain velocity is achieved bouncing and rolling would occur (Ritchie, 1963).

Based on site observations, the talus deposits near the alignment in the vicinity of Shellrock Mountain consist of angular, fine-grained diorite blocks with typical maximum dimensions of 12 to 18 inches. However, in some areas such as between Stations 76+00 to 80+00, larger clasts up to 2 or 3 feet in diameter were observed. The typical size of talus clasts is controlled by joint spacing in the rock mass at the source, which is variable in the source rock.

Large rockslides and infinite-slope debris flows/debris slides were also observed in these thick talus deposits. The infinite slope failures have a failure plane that is parallel to the slope angle, and occur at shallow depths.

Oversteepening of the talus would result in unstable slopes subject to failure and rockfalls. The danger of rockfalls, rolling rocks, and unstable cuts in talus can be mitigated by retaining walls and rockfall attenuator fences.

Massive shallow earthflows and debris slides that extend upslope to the bedrock source areas (900 hundred feet above the Railroad Track) have also developed in the talus deposits on the flanks of Wind Mountain, across the river from Shellrock Mountain. Wind Mountain is identical in morphology and geology as Shellrock Mountain.

3.5.6 Rockfall

Beaulieu’s (1977) geologic hazard maps in Bulletin 91 show areas of potential rockfall hazard in the vicinity, based on steep slopes and areas underlain by CRB or intrusive rocks. Beaulieu (1977) essentially shows the entire alignment as being susceptible to steep-slope mass movement, with 30% to 50%, and >50% slopes with hazards including local debris flows, rockfalls, rockslides and severe erosion. The locations are a function of rock type and structure (i.e. joint density and orientation), soil thickness, vegetative cover, and ground disturbance.

ODOT Rockfall Hazard Study. The Oregon DOT conducted an evaluation and analyses of rockfall hazards along the HCRH trail between Wyeth and Hood River (ODOT, 2013). The study examined segments of the proposed alignment that are located immediately adjacent to existing rock or talus slopes, where trail design will require consideration of rock slope performance and hazard. The study included field assessment, rockfall modeling, analysis of rockfall risk, and provides scoping-level, conceptual rockfall mitigation alternatives and cost estimates for nine hazardous slopes.

During the geologic site reconnaissance, rockfall damage was observed in the catchment fence and to the bin walls between 75+00 and 78+00. Rockfall hazards can be minimized by controlling blasting, scaling loose rocks from cliffs, constructing ditches or fences to contain rockfalls, covering slopes with draped or anchored mesh, and grouting areas of loose rock.

December 2014 13 Cornforth Consultants, Inc.

4. FIELD EXPLORATIONS

4.1 General

The field exploration program for Segment D included drilling 4 borings and excavating 18 test pits.

The borings and 9 of the test pits were observed by Cornforth Consultants, Inc. (CCI). Nine of the test pits were observed by Allegiance Group, LLC (Allegiance) of Vancouver, Washington under subcontract to CH2M Hill. Boring and test pit locations were established using a hand held GPS unit, and are provided in Appendix C. Appendix C also includes foundation data sheets that show the locations of explorations and a summary of subsurface conditions encountered. Summary logs and foundation data sheets for explorations performed by Allegiance are included in a separate report attached as Appendix D.

4.2 Test Pits

Test pits LINTP-1 through LINTP-7, WWTP-1 and WWTP-2 were excavated between June 16 and 18, 2014 using a John Deere 120 track-mounted excavator operated by Stratus Corporation of Gaston, Oregon. Test pits WETP-1 through WETP-9 were observed by Allegiance. The test pits were excavated to a maximum depth of 12 feet below ground surface (BGS) or until refusal was encountered. Representative soil samples were collected from the test pits for classification and laboratory testing. Summary test pit logs which include materials encountered, groundwater observations, and sample locations are shown in Appendix C. Representative photographs of each test pit are included after each summary log.

4.3 Borings

Borings LINB-1 and LINB-2 were drilled between June 26 and July 1, 2014 by Western States Soil Conservation, Inc. of Hubbard, Oregon using a CME 45 skid-mounted drill rig. Each boring was advanced to a depth of 40 feet using mud-rotary drilling methods through the overburden and HQ3 wireline coring techniques into the underlying rock. HQ3 drilling methods were also used on occasion to advance through overburden containing large cobbles and boulders. Both borings were abandoned with bentonite chips at the completion of drilling.

Borings WWB-1 and WEB-1 were drilled between July 10 and 14, 2014 by Western States Soil Conservation, Inc. of Hubbard, Oregon using a CME 80 track-mounted drill rig. WWB-1 and WEB- 1 were advanced to 85 feet and 76.5 feet, respectively, using mud-rotary drilling methods. A 1-inch PVC stand pipe piezometer was installed in WWB-1 at a depth of 75.5 feet and finished with a flush mount monument at the ground surface. WEB-1 was backfilled with bentonite chips.

Standard Penetration Tests (SPTs) were performed in the overburden at 2.5-foot intervals to a depth of 10 feet and at 5 foot intervals thereafter to evaluate the density of the overburden and collect soil samples. Summary Boring Logs describing the drilling methods, materials encountered, depths of samples, and SPT blow counts are shown in Appendix C. Photographs of rock core recovered during drilling are included after each summary log.

December 2014 14 Cornforth Consultants, Inc.

5. LABORATORY TESTING

5.1 General

Soil and rock samples recovered during explorations were visually examined in our laboratory to confirm and refine field classifications. Atterberg Limits, natural moisture content, and gradation analyses were performed on select soil samples to further refine field classification. Unconfined compression strength tests were performed on select rock samples. Laboratory testing was performed by WFLHD. Laboratory test results are provided in Appendix E. Results of laboratory testing have been incorporated into the summary boring and test pit logs.

December 2014 15 Cornforth Consultants, Inc.

6. SUBSURFACE CONDITIONS

6.1 Trail Reaches

Segment D was subdivided into reaches with similar geologic conditions, topographic features, and/or proposed improvements. The following sections describe the subsurface conditions encountered during the field investigation for the individual trail reaches.

6.1.1 Lindsey Creek Drainage – Segment D (151+50 to 155+00)

Borings LINB-1 and LINB-2 were advanced 40 feet BGS on either side of the Lindsey Creek box culvert. LINB-1 encountered medium dense to dense, brown and gray, angular, sandy gravel with some silt (Fill) from ground surface to approximately 12 feet; medium dense to dense, brown and gray, subangular to subrounded, sandy gravel with some silt and cobbles (Vegetated Talus) from approximately 12 to 25 feet; and hard (R4), moderately to slightly weathered, very closely to closely jointed, basalt (Columbia River Basalt) from approximately 25 to 41.8 feet. Boring LINB-2 encountered similar conditions as LINB-1 with Fill from ground surface to approximately 9 feet, Vegetated Talus from approximately 9 to 25.7 feet and Columbia River Basalt from approximately

25.7 to 41.5 feet. Depth to groundwater in the borings was not determined due to the use of mud-rotary drilling and HQ3 wireline coring methods. Borings were backfilled with bentonite chips after drilling operations were completed.

Test pit LINTP-1 was excavated approximately 150 feet east of LINB-2 at approximate station 154+10 to a total depth of about 12 feet. Material encountered consisted of medium dense, brown, subangular to subrounded, cobble to boulder-sized rock fragments in a matrix of sandy gravel with some silt (Fill/Reworked Vegetated Talus) from ground surface to approximately 7 feet; and brown, subangular to subrounded, cobble to boulder-sized fragments in a matrix of sandy gravel with some silt (Vegetated Talus) from approximately 7 feet to 12 feet. Approximately 3-inches of old asphalt pavement was encountered at 4 feet. Two 2.5-inch old communication lines were encountered at approximately 6 feet in the southeast corner of the test pit. The communications lines were surrounded by approximately 18-inches of fine to medium sand backfill. No seepage was encountered in the test pit.

6.1.2 Dog Hair Forest – Segment D (155+00 to 172+00)

Test pits LINTP-2 through LINTP-7 were excavated in the southern ditch line of Interstate 84 approximately 150 feet to 350 feet apart, starting just east of Lindsey Creek (approximate station 156+10) and extending east to a BPA access road (approximate station 170+00). Test pits LINTP-2 through LINTP-6 generally encountered medium dense, brown, subangular to subrounded, cobble to boulder-sized rock fragments in a matrix of sandy gravel with some silt (Vegetated Talus). LINTP-3 encountered 8 feet of vegetated talus overlying medium hard (R3) to hard (R4), gray to red-brown, moderately to slightly weathered, basalt (Columbia River Basalt). LINTP-2, LINTP-3, LINTP-5 and LINTP-6 encountered two 2.5-inch old communications lines at depths ranging from approximately 3 to 5 feet. LINTP-4 encountered a 10-inch metal culvert at approximately 5 feet on the north end of the test pit, which was surrounded by approximately 3 feet of fine to medium sand backfill. Test pit LINTP-7 encountered medium dense, brown, angular, sandy gravel with some silt (Fill/Reworked

December 2014 16 Cornforth Consultants, Inc.

Vegetated Talus) from ground surface to approximately 2.5 feet; loose, light brown, silt with some sand and gravel (Fill) from approximately 2.4 feet to 8 feet; and brown, subangular to subrounded, cobble to boulder-sized fragments in a matrix of sandy gravel with some…