Loading...
HomeMy WebLinkAboutDRAFT Middle Hoh River Resiliency Plan 071821 THIS PAGE INTENTIONALLY LEFT BLANK TABLE OF CONTENTS Introduction 1 Problem Statement 1 Plan Goals 1 Plan Overview 2 Plan Stakeholders 2 Existing Conditions 3 Landscape Setting 3 Geology & Geomorphology 3 Landslides 5 Sediment Sources 6 Channel Migration Zone 8 Riparian and Floodplain Forests 20 Methods & Study Area 20 Results & Discussion 23 Hydrology & Hydraulics 27 Hydrology and Boundary Conditions 28 Mesh Development and Roughness Categories 30 Modeled Infrastructure 30 Calibration 30 Results 31 Aquatic Habitat 33 Methods 34 Results 38 Oxbow Canyon Reach 52 Willoughby Creek Reach 52 Morgans Crossing Reach 54 Spruce Canyon Reach 55 Huelsdonk-South Fork Reach 55 Transportation 57 Trends & Anticipated Changes 59 Climate Change 59 Sediment Sources 60 Forests 61 Aquatic Habitat 62 Invasive Species Trends 62 Desired Future Conditions 63 Restoration Corridor 63 Long-Term Desired Conditions (include overarching goals and specific objectives) 64 Intermediate-Term Desired Conditions 64 Short-Term Desired Conditions 65 Local Capacity 65 Phase II Approach 66 References 67 Introduction 1 Problem Statement 1 Plan Goals 1 Plan Overview 2 Plan Stakeholders 2 Existing Conditions 3 Landscape Setting 3 Geology & Geomorphology 3 Landslides 5 Sediment Sources 6 Channel Migration Zone 8 Riparian and Floodplain Forests 20 Methods & Study Area 20 Results & Discussion 23 Hydrology & Hydraulics 27 Hydrology and Boundary Conditions 28 Mesh Development and Roughness Categories 30 Modeled Infrastructure 30 Calibration 30 Results 31 Aquatic Habitat 33 Methods 34 Results 38 Oxbow Canyon Reach 52 Willoughby Creek Reach 52 Morgans Crossing Reach 54 Spruce Canyon Reach 55 Huelsdonk-South Fork Reach 55 Transportation 57 Trends & Anticipated Changes 59 Climate Change 59 Sediment Sources 60 Forests 61 Aquatic Habitat 62 Invasive Species Trends 62 Desired Future Conditions 63 Restoration Corridor 63 Long-Term Desired Conditions (include overarching goals and specific objectives) 64 Intermediate-Term Desired Conditions 64 Short-Term Desired Conditions 65 Local Capacity 65 Phase II Approach 66 References 67 LIST OF TABLES Table 1. Inside the CMZ. Forest type and height class for 2014. 26 Table 2. Outside the CMZ. Forest type and height class 27 Table 3. Estimated peak flows at each gage in the model domain. 28 Table 4. Modeled Discharge Values at Inflow Locations 29 Table 5. Calibrated Manning's n roughness values for each roughness category. 31 Table 6. Average depth by reach for the modeled peak floods. 31 Table 7. Average velocity by reach for the modeled peak floods. 31 Table 8. Pool frequency standards for functioning rivers developed by NMFS (1996). 36 Table 9. Overview of mainstem habitats surveyed in the Middle Hoh River Study Area by reach. The count of observed units is n; length (feet) and area (square feet) are totals for the channels. Edge length (feet) and slow water area (square feet) are totals for all edges. Braid and side-channel habitats were surveyed as time allowed and were not a full census of habitats. Na indicates the habitat did not exist; ns indicates the habitat did exist but was not surveyed due to time constraints. 41 Table 10. Pool counts, reach width (feet), total pool area (square feet), the percent of area in pools, pool frequency (pools per mile), and pool spacing (channel widths per pool) (Montgomery et al. 1995) for all channel types and reaches for the Hoh River mainstem. Braid and side-channel habitats were surveyed as time allowed and were not a full census of habitats. Na indicates the habitat did not exist; ns indicates the habitat did exist but was not surveyed. 45 Table 11. Counts (n) and total areas (square feet) large wood jams (LWJ) located in main channel and braids, side channels, and islands, on banks, bars, or islands, and wet or dry. Jams were delineated using aerial imagery collected in March 2021. 46 Table 12. Counts of large wood jams (n), frequency (LWJ count/mile), total LWJ area (square feet), and mean and median LWJ size (area in square feet; 50th, and 90th percentiles, Jam50, Jam90). 48 Table 13. Identified road segments within the Middle Hoh CMZ, Resiliency Corridor and FEMA 100-yr floodplain. 57 Table 14. The magnitude of future peak flows 2070-2099 projected as result of warming climate (A1B scenario). 60 Table 1. Inside the CMZ. Forest type and height class for 2014. 26 Table 2. Outside the CMZ. Forest type and height class 27 Table 3. Estimated peak flows at each gage in the model domain. 28 Table 4. Modeled Discharge Values at Inflow Locations 29 Table 5. Calibrated Manning's n roughness values for each roughness category. 31 Table 6. Average depth by reach for the modeled peak floods. 31 Table 7. Average velocity by reach for the modeled peak floods. 31 Table 8. Pool frequency standards for functioning rivers developed by NMFS (1996). 36 Table 9. Overview of mainstem habitats surveyed in the Middle Hoh River Study Area by reach. The count of observed units is n; length (feet) and area (square feet) are totals for the channels. Edge length (feet) and slow water area (square feet) are totals for all edges. Braid and side-channel habitats were surveyed as time allowed and were not a full census of habitats. Na indicates the habitat did not exist; ns indicates the habitat did exist but was not surveyed due to time constraints. 41 Table 10. Pool counts, reach width (feet), total pool area (square feet), the percent of area in pools, pool frequency (pools per mile), and pool spacing (channel widths per pool) (Montgomery et al. 1995) for all channel types and reaches for the Hoh River mainstem. Braid and side-channel habitats were surveyed as time allowed and were not a full census of habitats. Na indicates the habitat did not exist; ns indicates the habitat did exist but was not surveyed. 45 Table 11. Counts (n) and total areas (square feet) large wood jams (LWJ) located in main channel and braids, side channels, and islands, on banks, bars, or islands, and wet or dry. Jams were delineated using aerial imagery collected in March 2021. 46 Table 12. Counts of large wood jams (n), frequency (LWJ count/mile), total LWJ area (square feet), and mean and median LWJ size (area in square feet; 50th, and 90th percentiles, Jam50, Jam90). 48 Table 13. Identified road segments within the Middle Hoh CMZ, Resiliency Corridor and FEMA 100-yr floodplain. 57 Table 14. The magnitude of future peak flows 2070-2099 projected as result of warming climate (A1B scenario). 60 LIST OF FIGURES Figure 1. Exposure of Pleistocene glacial outwash (gravel) and glaciolacustrine (clay) deposits near Elk Creek on the right bank of the main stem channel at RM 19. 4 Figure 2. Ice River Glacier (2010) terminus with exposed pro-glacial and steep lateral moraine sediments (left) and Blue Glacier (2009) lateral moraine and unstable hillside contributing to rockfall on glacier (right) Courtesy of NPS. 6 Figure 3a. Photo showing relatively stable snag in river where bank erosion recruited a large tree. The snag is deflecting flow away from the bank and helped to slow down bank erosion. Flow is from right to left (Oct. 1 2020 at RM 18.8). 12 Figure 3b. Photos illustrating importance of large trees in providing wood to the river that is capable of altering channel hydraulics and slowing bank erosion (top photo of stable 9 ft diameter Sitka Spruce with 30 ft rootwad in the Queets River). Bottom photo showing high terrace with young trees (industrial tree farm) where trees reaching channel are quickly transported downstream and don’t add the hydraulic roughness to the channel needed to slow erosion (Abbe and Brooks 2011). 13 Figure 3d. Example of where recruitment of large trees to the South Fork Hoh and channel response. resulted in not only stopping bank erosion but building new floodplain along toe of eroding bank. From 1990 to 2006 the river migrated 106 ft into an area of mature timber. Once in the channel the timber halted erosion and by 2013 the active channel had moved back to the south (Abbe et al. 2016). 15 Figure 3e. Example of where erosion proceeded along the South Fork Hoh due to a lack of large trees. This site is located downstream of the previous figure. Between 1990 and 2006 the river migrated 153 ft (9.6 ft/yr) into a clear-cut. Bank erosion triggered a landslide that extended 550 ft into the adjacent valley margin. The landslide headscarp retreated at a rate of 24 ft/yr from 1990 to 2013 (Abbe et al. 2016). 16 Figure 4. Comparison of Bureau of Reclamation 2004 mapping to updated mapping conditions and revised methodologies. Image at left is 2004 channel location mapping (colored ribbons) and channel migration zone (black hatched zone). Image at right is updated channel mapping with historic migration zone (red), geomorphic migration zone (dark orange), erosion hazard area (light orange) and geotechnical setback (yellow), all included in the updated CMZ limit (black boundary). Note that the current channel is outside of the 2004 delineated channel migration zone. 19 Figure 5. First return 2014 LiDAR DEM of forest vegetation heights 21 Figure 6. Middle Hoh River riparian forest type cover and height classes. 22 Figure 7. Hoh River riparian forest mosaic of cover types. October 1st, 2020 near RM 20.6 22 Figure 8. Forest typing and channel migration zone (CMZ). 23 Figure 9. Young deciduous (red alder and willow) and older mixed conifer deciduous (red alder and Sitka spruce) forest types. October 1, 2020 near RM 20.6 24 Figure 10. Mature mixed (Black cottonwood, red alder, Sitka spruce) and conifer forest types (Sitka spruce, Douglas fir). October 1, 2020 near RM 20.7 25 Figure 11. Inside CMZ. Riparian Forest Type (Acres) and Height Class (Feet/Acres) for 2014 26 Figure 12. Outside CMZ. Riparian Forest Type (Acres) and Height Class (Feet/Acres)for 2014. 27 Figure 13. Locations of model inflow and outflow locations illustrated over the model domain. 29 Figure 14. Instantaneous flow (cfs) data at USGS Gage 12041200 – Hoh River at Highway 101 near Forks, WA recorded and reported by USGS. Surveys were conducted between September 28th and October 1st, 2020. 35 Figure 15. Example of jams delineated in the Morgans Crossing Reach (RM 21.9), shown at a 1:600 scale using aerial photography collected in March 2021 (NV5 GeoSpatial 2021). Blue polygons depict wood in wetted channel (time of photo) and red polygons wood outside wetted channel but within ordinary high water (bankfull) channel. Flow is from right to left. 37 Figure 16. Instantaneous flow (cfs) data at USGS Gage 12041200 – Hoh River at Highway 101 near Forks, WA recorded and reported by USGS. Aerial imagery was collected on March 20th, 2021. 38 Figure 17. The percent of wetted habitat area by habitat unit, channel type, and reach. Braid and side-channel habitats were surveyed as time allowed and were not a full census of habitats, ns indicates the habitat type was present but not surveyed. No braid or side-channel habitat was wetted at the time of survey in the Oxbow Canyon Reach. 49 Figure 18. Edge length (feet) by edge type, channel type, and reach. Braid and side-channel habitats were surveyed as time allowed and were not a full census of habitats, ns indicates the habitat type was present but not surveyed. No braid or side-channel habitat was wetted at the time of survey in the Oxbow Canyon Reach. 50 Figure 19. Average slow water edge area (feet) by edge type, and reach. Braid and side-channel habitats were surveyed as time allowed and were not a full census of habitats, ns indicates the habitat type was present but not surveyed. No braid or side-channel habitat was wetted at the time of survey in the Oxbow Canyon Reach. 50 Figure 20. Pool counts by pool type, channel type, and reach. Braid and side-channel habitats were surveyed as time allowed and were not a full census of habitats, ns indicates the habitat type was present but not surveyed. No braid or side-channel habitat was wetted at the time of survey in the Oxbow Canyon Reach. 51 Figure 21. Pool counts by pool forming feature, channel type, and reach. Channel feature includes meanders and confluences. Braid and side-channel habitats were surveyed as time allowed and were not a full census of habitats, ns indicates the habitat type was present but not surveyed. No braid or side-channel habitat was wetted at the time of survey in the Oxbow Canyon Reach. 51 Figure 22. Mean pool frequency (pools per mile) versus wood frequency (LWJ count per mile) of wetted jams pool for reaches in the Hoh River study area with a 95% confidence interval shown. The relationship was not significant, p-value=0.40. 52 Figure 1. Exposure of Pleistocene glacial outwash (gravel) and glaciolacustrine (clay) deposits near Elk Creek on the right bank of the main stem channel at RM 19. 4 Figure 2. Ice River Glacier (2010) terminus with exposed pro-glacial and steep lateral moraine sediments (left) and Blue Glacier (2009) lateral moraine and unstable hillside contributing to rockfall on glacier (right) Courtesy of NPS. 6 Figure 3a. Photo showing relatively stable snag in river where bank erosion recruited a large tree. The snag is deflecting flow away from the bank and helped to slow down bank erosion. Flow is from right to left (Oct. 1 2020 at RM 18.8). 12 Figure 3b. Photos illustrating importance of large trees in providing wood to the river that is capable of altering channel hydraulics and slowing bank erosion (top photo of stable 9 ft diameter Sitka Spruce with 30 ft rootwad in the Queets River). Bottom photo showing high terrace with young trees (industrial tree farm) where trees reaching channel are quickly transported downstream and don’t add the hydraulic roughness to the channel needed to slow erosion (Abbe and Brooks 2011). 13 Figure 3d. Example of where recruitment of large trees to the South Fork Hoh and channel response. resulted in not only stopping bank erosion but building new floodplain along toe of eroding bank. From 1990 to 2006 the river migrated 106 ft into an area of mature timber. Once in the channel the timber halted erosion and by 2013 the active channel had moved back to the south (Abbe et al. 2016). 15 Figure 3e. Example of where erosion proceeded along the South Fork Hoh due to a lack of large trees. This site is located downstream of the previous figure. Between 1990 and 2006 the river migrated 153 ft (9.6 ft/yr) into a clear-cut. Bank erosion triggered a landslide that extended 550 ft into the adjacent valley margin. The landslide headscarp retreated at a rate of 24 ft/yr from 1990 to 2013 (Abbe et al. 2016). 16 Figure 4. Comparison of Bureau of Reclamation 2004 mapping to updated mapping conditions and revised methodologies. Image at left is 2004 channel location mapping (colored ribbons) and channel migration zone (black hatched zone). Image at right is updated channel mapping with historic migration zone (red), geomorphic migration zone (dark orange), erosion hazard area (light orange) and geotechnical setback (yellow), all included in the updated CMZ limit (black boundary). Note that the current channel is outside of the 2004 delineated channel migration zone. 19 Figure 5. First return 2014 LiDAR DEM of forest vegetation heights 21 Figure 6. Middle Hoh River riparian forest type cover and height classes. 22 Figure 7. Hoh River riparian forest mosaic of cover types. October 1st, 2020 near RM 20.6 22 Figure 8. Forest typing and channel migration zone (CMZ). 23 Figure 9. Young deciduous (red alder and willow) and older mixed conifer deciduous (red alder and Sitka spruce) forest types. October 1, 2020 near RM 20.6 24 Figure 10. Mature mixed (Black cottonwood, red alder, Sitka spruce) and conifer forest types (Sitka spruce, Douglas fir). October 1, 2020 near RM 20.7 25 Figure 11. Inside CMZ. Riparian Forest Type (Acres) and Height Class (Feet/Acres) for 2014 26 Figure 12. Outside CMZ. Riparian Forest Type (Acres) and Height Class (Feet/Acres)for 2014. 27 Figure 13. Locations of model inflow and outflow locations illustrated over the model domain. 29 Figure 14. Instantaneous flow (cfs) data at USGS Gage 12041200 – Hoh River at Highway 101 near Forks, WA recorded and reported by USGS. Surveys were conducted between September 28th and October 1st, 2020. 35 Figure 15. Example of jams delineated in the Morgans Crossing Reach (RM 21.9), shown at a 1:600 scale using aerial photography collected in March 2021 (NV5 GeoSpatial 2021). Blue polygons depict wood in wetted channel (time of photo) and red polygons wood outside wetted channel but within ordinary high water (bankfull) channel. Flow is from right to left. 37 Figure 16. Instantaneous flow (cfs) data at USGS Gage 12041200 – Hoh River at Highway 101 near Forks, WA recorded and reported by USGS. Aerial imagery was collected on March 20th, 2021. 38 Figure 17. The percent of wetted habitat area by habitat unit, channel type, and reach. Braid and side-channel habitats were surveyed as time allowed and were not a full census of habitats, ns indicates the habitat type was present but not surveyed. No braid or side-channel habitat was wetted at the time of survey in the Oxbow Canyon Reach. 49 Figure 18. Edge length (feet) by edge type, channel type, and reach. Braid and side-channel habitats were surveyed as time allowed and were not a full census of habitats, ns indicates the habitat type was present but not surveyed. No braid or side-channel habitat was wetted at the time of survey in the Oxbow Canyon Reach. 50 Figure 19. Average slow water edge area (feet) by edge type, and reach. Braid and side-channel habitats were surveyed as time allowed and were not a full census of habitats, ns indicates the habitat type was present but not surveyed. No braid or side-channel habitat was wetted at the time of survey in the Oxbow Canyon Reach. 50 Figure 20. Pool counts by pool type, channel type, and reach. Braid and side-channel habitats were surveyed as time allowed and were not a full census of habitats, ns indicates the habitat type was present but not surveyed. No braid or side-channel habitat was wetted at the time of survey in the Oxbow Canyon Reach. 51 Figure 21. Pool counts by pool forming feature, channel type, and reach. Channel feature includes meanders and confluences. Braid and side-channel habitats were surveyed as time allowed and were not a full census of habitats, ns indicates the habitat type was present but not surveyed. No braid or side-channel habitat was wetted at the time of survey in the Oxbow Canyon Reach. 51 Figure 22. Mean pool frequency (pools per mile) versus wood frequency (LWJ count per mile) of wetted jams pool for reaches in the Hoh River study area with a 95% confidence interval shown. The relationship was not significant, p-value=0.40. 52 LIST OF MAPS Map 1 Project Reach Map 2 Geologic Map Map 3 Landslide and Slope Stability Map Book Map 4 Channel Migration Zone Map Book Map 5 Riparian Forest Map Book Map 6 Existing Conditions Hydraulic Model Results Map 7 Aquatic Habitat Map Book LIST OF APPENDICES Appendix A Aquatic Habitat Field Data Appendix B Middle Hoh River Action Plan THIS PAGE LEFT INTENTIONAL BLANK INTRODUCTION Problem Statement River valleys are the most ecologically rich and complex areas within a watershed. River valleys also are attractive areas for people, offering productive soils for agriculture, level ground for homes and roads. The principal processes that define river valleys, flooding, and erosion, are crucial in forming and sustaining habitat but can pose significant hazards to infrastructure and property. There are no formal programs for comprehensive management of river valleys. Management typically reflects the diversity of landownership and infrastructure found in a valley; thus, actions of different entities can frequently conflict with one another. The lack of communication between responsible entities can lead to unintended impacts to ecosystems and human communities. The Middle Hoh Resiliency Plan is to provide a unifying structure to better manage this important river. The Hoh River has provided subsistence for the Hoh Tribe and a diverse wildlife population for thousands of years. The native peoples’ deep respect for the river and its natural rhythm of flooding and shifting channel banks was key to sustaining a balanced human and wildlife ecosystem for millennia. Over the last century, the watershed has undergone changes that it had never been subjected to either naturally or by its original inhabitants, including local and broader changes as well such as flow and temperature changes occurring due to a warming climate. Human changes in the basin such as the loss of large trees and long rock revetments have altered natural processes that create and sustain aquatic and riparian habitat, thus negatively impacting salmonid populations. Changes have also resulted in accelerated rates of bank erosion and property loss, elevated flood risks, and costly road repairs. Road washouts threaten access for local residents as well as for visitors to Olympic National Park (ONP); thus, directly impacting the economy of this important local and scenic valley. Road repairs are expensive and directly destroy salmon habitat; channel margins are some of the most important zones for juvenile and adult salmon where they expect to find a submerged forest to hide within and feed on a complex food web. The chronic threat of erosion puts the community’s homes and roads at risk - and it is only going to become more severe in the coming decades. The realization that the river community is not currently as resilient as it can be prompted Jefferson County to develop a resiliency plan for the Middle Hoh River, extending from the ONP boundary to Oxbow Canyon (Map 1). This plan will better define flood and erosion risks to residents, infrastructure, and habitat, and with that information, outline meaningful measures that can be taken to reduce flood and erosion risks to landowners and the infrastructure network while allowing the valley bottom to achieve its historic level of salmon and wildlife productivity. A multi-disciplinary team of locals, scientists, conservation groups, county officials and Tribal representatives have developed this plan working together with the greater community, state and federal agencies and the Hoh Tribe to develop, prioritize and implement actions that are mutually beneficial to the community and wildlife. This resiliency plan will also open additional opportunities for funding as well as reducing regulatory costs and timelines that have impacted past and current projects. The Plan establishes an initial template that is intended to be periodically updated with new information. The Plan and the leadership team of stakeholders will hopefully be sustained permanently, like commitments to maintain infrastructure. Most importantly the Plan is intended to bring the community and numerous entities responsible for managing resources in the Middle Hoh River valley together under one tent. Plan Goals The primary goal of this plan is to produce a scientific foundation for improving community and ecosystem resiliency for people and wildlife in the Middle Hoh River, particularly for salmonids, extending from now and into the future. This plan has been collaboratively developed with the local community, property owners and agencies responsible for land management and infrastructure. The goal is that through the collaborative process, this plan will result in enthusiastic support and a commitment to implementation from all stakeholders, to better serve the community and the environment into the future. The Resiliency Plan has established a structure that has brought together the people who live, work and manage resources within the river valley and provide the following: an understanding of the processes and conditions that define the rivers ecology and impact the human community, a source of information and experts (e.g., natural resources, hazards, regulations, infrastructure), a network for communication regarding management actions, a safe structure for debate that is respectful of different perspectives and encourages transparency in decision making that impacts natural resources and people, works to develop a shared vision for one of the most unique and intact river ecosystems in the United States, provides guidance for protecting and restoring critical salmonid habitat as well as helping the local community prosper. Plan Overview This plan is separated into 6 broad topics covering existing conditions in the project reach, trends and anticipated response, desired conditions, resiliency opportunities, local capacity to perform work, and the action plan. These topics have been organized to describe the current condition, how is it likely to change, describe the desired condition, identify opportunities to move toward desired conditions, evaluate how much of the work can be completed with local labor and materials, develop overarching strategies for improving resiliency where opportunities exist. A description of existing conditions is provided to establish baseline conditions in the project reach and identify opportunities for improving resiliency. System trends describe how the existing baseline conditions have evolved in recent history and anticipates future system responses accounting for climate change under a no-action scenario (i.e., no proactive measures to improve system resiliency). Plan Stakeholders A project steering committee was convened monthly during the development of this plan to provide a forum for collaboration with the Hoh Tribe, various government agencies, non-government organizations, local landowners and non-local river users. This diverse group provided information and data resources, coordinated outreach to the community and contributed meaningful and critical feedback on assessments completed, proposed actions, and interim drafts of this plan. Agencies and organizations represented in the steering committee included: 10,000 Years Institute Coast Salmon Partnership Hoh River Trust Hoh Tribe Olympic National Park Olympic Natural Resources Center North Pacific Lead Entity Pacific Coast Salmon Coalition The Nature Conservancy Trout Unlimited United States Forest Service Washington Department of Fish and Wildlife (WDFW) Washington Department of Natural Resources (DNR) Wild Salmon Center EXISTING CONDITIONS Landscape Setting The Hoh River is located on the West Coast of Washington State, originates from glaciers high on Mount Olympus and descends 7,980-feet (ft) over 57-miles (mi) before entering the Pacific Ocean. The watershed encompasses 298 mi2, the upper 57% of the basin lies within Olympic National Park (ONP) and is considered to be pristine temperature rainforest and home to some of the largest old growth trees in the United States (McMillan and Starr 2008; NPCLE 2020). Downstream of ONP the Hoh basin is a mix of public and private land under various uses. The watershed receives an average annual precipitation of 155-inches (in) falling as rain and snow depending on time of year and elevation, however there is a large precipitation gradient from 240-in above the glaciers to 93-in at the coast (Lieb and Perry 2005). The cobble-gravel bed river flows generally east to west gaining flow from several major non-glacial tributaries draining primarily industrial timber lands. River flows fluctuate throughout the year with the highest monthly averages occurring in December and January and minimum averages in September, with a sustained freshet typically extending into June (Lieb and Perry 2005). The Middle Hoh project reach extends from the ONP boundary to Highway 101, covering 15-river miles (RM) with an average slope of 0.3% (Map 1). The upstream end of the reach begins as the river transitions from the confined steep upper watershed draining ONP to a broad valley with wide floodplains and lower slope. The valley is approximately 1800-ft wide on average and has Pleistocene glacial terraces along the margins composed of outwash, drift and glaciolacustrine deposits. The 15-mi project reach is divided into 5 reaches, adopted from previous work completed by the Bureau of Reclamation (Piety et. al. 2004) and are representative of distinct geomorphic reaches. Geology & Geomorphology The geologic and glacial history of the Olympic Mountains are primary drivers influencing conditions in the Hoh River basin. The Olympic Mountain Range was and continues to be uplifted resulting from the Juan de Fuca plate subducting under the North American plate, creating a mélange of oceanic sedimentary rocks accreted onto the North American plate at the convergent boundary. These rocks are uplifted, folded, fractured and experience low-grade metamorphism, weakening the rocks and making them highly erodible. These rocks underly the watershed and are exposed primarily at the eastern, highest portion of the watershed in steep mountainous terrain (Map 2). This bedrock is exposed in a few locations within the study reach on the valley floor, but the predominant geologic units are comprised of unconsolidated/poorly-consolidated glacial deposits and post-glacial alluvium/colluvium. Where bedrock is present, it has significant influence on both lateral channel migration and vertical gradient control. Bedrock controls are present in two locations within the study reach: Spruce Canyon (RM 26) and Oxbow Canyon (RM 17). During the Pleistocene from 54,000 – 18,000-yr B.P. alpine glaciers advanced and retreated up and down the Hoh valley at least 4 times (Thackray 2001). Glacial processes both widened the valley, scoured the valley, and then deposited a variety of complex glacial sediments within the valley based on evolving depositional environments. Glacial deposits in the valley include moraines, tills, alluvial outwash, lacustrine, sub-glacial, ice-marginal, and ablation drifts. Deposits are present as low to moderately high banks along the margins of the active geomorphic migration zone. Of note is the presence of thick clay deposits in the channel banks and valley margins in the vicinity of Elk Creek near RM 19 and Owl Creek near RM 27, just upstream of Oxbow and Spruce Canyons, respectively (Figure 1). Oxbow Canyon formed as the Hoh cut through a mapped terminal moraine, formed at the end of an advancing alpine glacier. Once the glacier retreated a pro-glacial lake formed behind the moraine as meltwater accumulated, depositing the clay layers observed in the channel banks today. Once the river breached the moraine it carved Oxbow Canyon and drained the lake, which were subsequently buried with coarse outwash deposits. While not formally mapped, we interpret a similar course of events transpired to form Spruce Canyon, leaving behind the abundant and thick clay deposits found upstream to Canyon Creek. Figure 1. Exposure of Pleistocene glacial outwash (gravel) and glaciolacustrine (clay) deposits near Elk Creek on the right bank of the main stem channel at RM 19. More recent Holocene post-glacial deposits are comprised of unconsolidated alluvial sediments that range from sand and gravels to floodplain silts. Also included in this geologic group is alluvial fan deposits, colluvium, and recent mass-wasting deposits. The active geomorphic migration zone is comprised primarily of post-glacial fluvial deposits, forming low banks and is very erodible and is the predominant channel bank material within the study reach. The Middle Hoh River has a variety of channel morphologies ranging from confined (no floodplain) single thread boulder-bedrock channels (Oxbow and Spruce Canyons) to different unconfined (floodplain) alluvial channels that include meandering channel reaches, wide braiding (large gravel bars) channel reaches and multi-channel island dominated reaches. Through the latter half of 1900s the alluvial reaches of the Middle Hoh River have been predominantly an irregular, sinuous single-thread channel with occasional branched channels (anastomosing) or braided channels. Depending on temporal changes or disturbances to the system, a wandering channel morphology may evolve between either single-thread or anabranching channel forms over time or may locally develop other channel forms (such as braiding). The wandering channel type is less predictable over time than the meandering channel forms because of higher erosion rates and a relatively high potential for avulsions (new channel courses that take completely new path through the floodplain). The high sediment supply to the Middle Hoh promotes channel migration. Landslides Landslides are common in the Hoh watershed. The stability of slopes is controlled by the underlying geology, topography, vegetation, rainfall duration/intensity and natural and human disturbances. Natural disturbances include fire and windstorms that alter forest cover. Human disturbances such as road, drainage and forest clearing can destabilize slopes. Natural variability across the watershed creates the potential for 3 types of mass wasting: rapid shallow landslides, deep-seated rotational landslides and debris flows. Rapid shallow landslides are translational (parallel) slope failures occurring on a relatively planar surface, typically associated with a zone of weakness in the soil (e.g., limit of root depth) or contact with bedrock and are the most common mass wasting process in the watershed. Deep seated rotational landslides occur on a curved failure plane within the slope, with the displaced material rotating during failure, leaving a head-scarp at the up-gradient end and rotated material at the toe. Deep seated slides are typically associated with a deep impermeable layer such as the glacial lacustrine clay deposits common in the Hoh Valley. These deep clay layers are commonly exposed at the toe of hillslopes along the river and its tributaries. The third type are debris flows resemble wet concrete in consistency and can be very intense where they build up mass and momentum. Debris flows happen when water pore pressure exceeds the soil strength and gravitational forces trigger a landslide that develops into a debris flow, usually in steep headwater channels or hillslopes. If there is sufficient large wood in the path of the debris flow it can slow or halt the flow by limiting the flow from increasing in mass and momentum. Where flows encounter little resistance, they quickly gain mass and momentum and scour out channels all the way down to the Hoh valley where they deposit the sediment and wood mobilized. As part of the resiliency plan an inventory of landslides was completed that includes previous work (Parks 1999, McHenry 2001), as well as identification of new landslides not included in previous work. The primary source of previously identified landslides was from Parks (1999), and a hill shade generated from the 2013 LiDAR was the primary data used to identify new, previously unidentified landslides. Three licensed geologists in the state of Washington reviewed the LiDAR hill shade for surface forms indicating mass wasting had occurred, delineating the initiation point and runout to the failed material. A total of 202 landslides encompassing 730-ac were delineated by Parks (1999), with an additional 19 landslides mapped by NSD covering 123-ac (Map 3). Most of the rotational and translational failures occur along the contact between Quaternary alluvial deposits and unconsolidated Pleistocene glacial deposits at the valley margins. NSD mapped several landslides in the vicinity of Spruce Canyon and upstream to the confluence with the South Fork Hoh River that initiated in fine grained glaciolacustrine deposits not mapped previously (Map 2). These deposits are prone to deep seated slope failure (McHenry 2001) and are understood to have been deposited in a proglacial lake impounded by a terminal moraine that crossed the valley near Spruce Canyon. Meltwater from the receding glaciers would have backed up behind the moraine, depositing the thinly bedded clays and silts. In addition to mapping landslides that have occurred over time in the Hoh valley, the results of a predictive model of shallow rapid slope stability are included on the landslide inventory maps (Map 3). The model does not predict the potential for deep-seated landslides or debris flows. The data are from DNR and classify slope stability based on slope gradient and curvature, landslide inventories, geology, precipitation, etc. into zones of high, medium and low probability for failure. The models used were SMORPH (Shaw and Johnson 1995) and SHALSTAB (Montgomery and Dietrich 1994, the terrain surface used for the analysis was a 10-meter DEM with an accuracy of +/- 35-ft. The results show almost all the steep high terrain forming the valley margins are in the medium to high probability zones, as well as the same boundary between Quaternary alluvial deposits and Pleistocene glacial deposits where NSD mapped several landslides (Map 3). These results highlight the inherent instability of the watershed, exacerbated by forest clearing in places, and high potential for mass wasting. Direct delivery of failed slope sediments to the channel can result in large sediment pulses propagating downstream or remain chronic high sediment load contributors to the channel. Sediment Sources The sediment supply in the upper watershed of the Hoh River is virtually limitless given the high relief and weak rock and glacial deposits. The native forests have played a major role in limiting wide scale erosion. This sediment is generated from glacial processes, rainfall, slope failures, road disturbance and hillslope surface erosion. The glaciers descending Mount Olympus are the predominant source of sediment to the Hoh River as they grind away at the valley floor and margins. As these glaciers recede over the next century, sediment once buttressed by ice against the valley walls will become exposed on steep unvegetated slopes. These deposits, prone to shallow rapid failure, have the potential to deliver copious amounts of additional sediment to the already sediment laden pro-glacial Hoh River. Additionally, rockfall and rockslides similarly deliver sediment to the river and have been increasing as the ice buttressing the over-steepened valley walls is removed (Lyons, E. 2003) (Figure 2). All these glacier derived sediments can be delivered to the river gradually over time or as a sediment pulse or wedge that will propagate downstream. Recent studies have concluded that these processes are occurring at an accelerated rate due to the changing climate (Riedel et. al. 2015 and East et. al. 2016), increasing the sediment supply downstream and initiating channel widening and taking on a more braided form (East et. al/ 2016). // Figure 2. Ice River Glacier (2010) terminus with exposed pro-glacial and steep lateral moraine sediments (left) and Blue Glacier (2009) lateral moraine and unstable hillside contributing to rockfall on glacier (right) Courtesy of NPS. Landslide susceptibility mapping by the DNR of the contributing basin shows that mass wasting potential in the watershed is high (Map 3). While upper watershed tributaries continuously transport sediment to the river system, significant volumes of sediment delivered to the channel from mass wasting processes tend to occur episodically and are typically associated with high intensity precipitation events. Estimates of sediment production and transport were taken from published studies where available for each of the primary sources identified. Nelson (1986) estimated that 630,000 tons of suspended sediment are transported by the Hoh River (at the mouth) annually, and that 60% of that load is from the portion of the watershed draining ONP. Because the Middle Hoh reach begins at this same location, we have used 60% of the suspended sediment load reported by Nelson (1986) at the mouth, or 378,000-tons/yr, as an appropriate average annual suspended sediment load entering the project reach. This value, or 810-tons/yr/km2 when scaled to the drainage area, is in-line with other findings that coastal Pacific Northwest rivers have sediment loads between 100 – 500-tons/yr/km2 and rivers draining the glaciated Mt. Rainier are on the order of 1,000- tons/yr/km2 (Czuba et. al. 2012). Scaling the suspended sediment load to the downstream end of the project reach at Oxbow Park yields a load of 815-tons/yr/km2, greater than the load per unit area within ONP due to human disturbances within the Middle Hoh reach. Because only the suspended sediment load was measured (Nelson 1986), estimates for bedload are needed to determine a total sediment load (dissolved loads are not accounted for in this assessment). The fraction of the total sediment load constituting bedload varies as a function of eroding surface characteristics, slope steepness, discharge magnitude and underlying geology and typically fall the range of 1 – 15% of the total load, but can be as high as 87% or greater (Babinski 2005). Overall, the percentage of the total load as bedload increases away from the equator and for terrains sculpted by previous glaciations (Babinski 2005). Due to this uncertainty, estimates of bedload were calculated as a percentage of the total load between 1 – 30%, or between 381,780 – 491,400-tons/yr at the ONP boundary. Parks (1999) found that sediment loading within the Middle Hoh from hillslope fine sediment production to be 1,739-tons/yr and 2,147-tons/yr contributing from roads, Logan et. al. (1991) estimated 94,350-tons/yr of coarse sediment loading from mass wasting and 16,405 tons/yr contributing from tributaries (Czuba et. al. 2012) within the Middle Hoh. To estimate sediment production within ONP from these same sources we determined the tons/yr/mi2 of sediment produced in the Middle Hoh and scaled to an appropriate value for within ONP due to the difference in landcover and use. We estimated hillslope fine sediment production within ONP would be 50% that outside of ONP, sediment contribution from roads to be 1% relative to outside ONP, coarse sediment loading from mass wasting to be 90% and from tributaries to be 75% of that outside ONP per unit area. The resultant estimated sediment production within ONP from hillslope fine sediment is 1,840-tons/yr, from roads is 45-tons/yr, 179,820-tons/yr from mass wasting and 26,055-tons/yr from tributary inputs. / Summing these sources yields 207,762-tons/yr of sediment production within ONP, however this does not account for the contribution of sediment generated from the many glaciers feeding the Hoh River. Simply subtracting the total sediment production estimated within ONP (207,762-tons/yr) from the total sediment load exiting ONP (381,780 – 491,400-tons/yr depending on 1 - 3% bedload) yields an estimated sediment production between 174,018 – 283,638-tons/yr from glaciers, accounting for 46 – 58% of the total sediment load entering the reach. At the downstream end of the project reach the relative contribution of sediment from glaciers diminishes to between 40 – 54%. This approach, equating the fluvial transported sediment load to a sediment production rate, neglects any change in sediment storage within the reach (i.e. bank erosion reducing sediment storage, aggradation increasing storage). Based on these findings glacial processes are the dominant sediment producer within ONP. As the climate warms and the glaciers recede, sediment production should be expected to increase in the short-term as sediments previously shielded and buttressed by ice are now exposed to erosive forces. As the glaciers recede and sediment production increases the average total sediment load will increase in-kind, likely with a larger proportion of the total load as bedload. Channel Migration Zone This chapter summarizes the methods and analyses utilized to develop a Channel Migration Zone (CMZ) map for the Middle Hoh River. The Hoh River has provided subsistence for the Hoh Tribe and countless species for thousands of years. Respect for the river and its natural processes of flooding and erosion was key to sustaining the ecosystem for millennia. Over the last century the watershed and river valley have experienced changes it had never been subjected to naturally. These local and broader changes to the landscape affected flow, channel banks, floodplains, vegetation, sediment, climate, and land use. This new landscape condition affects physical processes that create and sustain aquatic and riparian habitat essential for healthy salmonid populations. The new condition also results in risks to human communities with regards to flooding, channel incision and bank erosion. This assessment is intended to provide the best available science to better define fluvial-related risks and opportunities to residents, infrastructure and habitat. Regulatory Framework The purpose of mapping the CMZ for the Middle Hoh was necessitated by both riverine hazard planning and ecosystem recovery planning. The intent of this study is to provide a technical background document and set of maps to help guide decision makers in adopting a CMZ and developing comprehensive flood hazard and ecological planning. CMZ mapping is supported within several regulatory codes, specifically: Shorelines Management Act (SMA): “Applicable shoreline master programs should include provisions to limit development and shoreline modifications that would result in interference with the process of channel migration that may cause significant adverse impacts to property or public improvements and or result in a net loss of ecological functions associated with the rivers and streams” (Chapter 173-26 WAC, 58). Growth Management Act (GMA): Critical Areas - RCW. 36.70A.030(5). ESA: Limit 12 of the 4(d) Rule requires the delineation of a CMZ. National Flood Insurance Act (NFIA) supports the delineation of a CMZ to manage flood hazards and reduce flood damages. Department of Ecology encourages development of a “meander belt” delineation in Comprehensive Flood Hazard Management Plans. The State further establishes recommendations on how to delineate the historic and hazard areas in [WAC 173-26-221(2)(c)(iv)(C)(3)(b)]: “For management purposes, the extent of likely migration along a stream reach can be identified using evidence of active stream channel movement over the past one hundred years. Evidence of active movement can be provided from historic and current aerial photos and maps and may require field analysis of specific channel and valley bottom characteristics in some cases. A time frame of one hundred years was chosen because aerial photos, maps and field evidence can be used to evaluate movement in this time frame.” In addition to the State recommendation of using the 100-year migration potential, FEMA also recommends developing channel migration zones that predict 100-year migration potential. The CMZ can be used to plan for or assess public safety (risk to life/property), economic costs (cost – benefit), and ecological function (salmon recovery). The CMZ can also be used for regulatory purposes (CAO and SMP) to lessen future risk. The State describes this concept in [WAC 173-26-221(2)(c)(iv)(C)(3)(b)]: “Scientific examination as well as experience has demonstrated that interference with this natural process often has unintended consequences for human users of the river and its valley such as increased or changed flood, sedimentation and erosion patterns. It also has adverse effects on fish and wildlife through loss of critical habitat for river and riparian dependent species. Failing to recognize the process often leads to damage to, or loss of, structures and threats to life safety.” Methods The Washington State Department of Ecology Publication (#03-06-027) “A Framework for Delineating Channel Migration Zones” was the general guideline used to develop the CMZ. The erosion and avulsion hazards analysis “takes into account trends in channel movement, context of disturbance history and changes in boundary conditions, as well as topography, bank erodibility, hydrology, sediment supply and woody debris loading” (Rapp et al, 2003). To the extent possible, the CMZ mapping for this analysis builds upon existing knowledge and information. This included previous mapping efforts and studies. In particular, the Geomorphic Assessment of Hoh River in Washington State (Bureau of Reclamation, 2004), LiDAR topography, geologic mapping, other geomorphic studies, air photo analysis, interviews, anecdotal accounts, and reconnaissance-level field investigation provided the foundation of information for this analysis. Definitions The definitions used in this analysis build on the definitions in regulatory and guidance documents but have been modified to fit the conditions and processes specific to the Middle Hoh River. The definitions used are: Historic Migration Zone (HMZ): The HMZ is a composite of the historic locations of the river as determined from historic information and interpretation, over some length of historic record. Active Geomorphic Migration Zone (GMZ): The GMZ is “the geographic area where a stream or river has been and will be susceptible to channel erosion and/or channel occupation” (Rapp et al, 2003); the GMZ considers a time period greater than the historic photo record. Erosion Hazard Area (EHA): The EHA in this analysis refers to the hazards resulting from the natural process of lateral migration of a channel through bank erosion that occurs from channel expansion, channel meandering, channel course changes, or channel bank and fluvially related slope failures. The EHA is a predicted horizontal channel migration potential area. Lateral erosion is not necessarily limited to the floodplain or areas inundated during the 100-year flood event. Avulsion Hazard Area (AHZ): The AHZ describes an area susceptible to multiple alluvial hazards associated with rapid channel course changes or temporary channelization of flow that in addition to having flooding hazards, instantaneously becomes an erosion hazard. Geotechnical Hazard Area (GHA): The GHA describes an area of potential slope instability driven by channel migration processes. For this study the fluvial-related geotechnical hazards are shown as an overlay on the maps. Alluvial Fan Hazard Area (AFH): The AFH describes an area where alluvial fan processes have occurred based on landform interpretation. Alluvial fan hazards include flooding, scour, erosion, avulsion, deposition, and debris impacts. For this study, AFHs are shown as an overlay on the maps. Channel Migration Zone (CMZ): The cumulative areas of the HMZ, GMZ, EHA, AHA, and GHA is the CMZ. The State of Washington, in [WAC 173-26-221(2)(c)(iv)(3)(b)] describes this concept further as: “The dynamic physical processes of rivers, including the movement of water, sediment and wood, cause the river channel in some areas to move laterally, or "migrate," over time. This is a natural process in response to gravity and topography and allows the river to release energy and distribute its sediment load.” A thorough review of previous data and literature was conducted as part of the development of this plan. The intent was to build extensively from the existing information. From this review, data gaps and updates were identified, and additional analyses were conducted to supplement and update the existing information. The Geomorphic Assessment of Hoh River in Washington State (Bureau of Reclamation, 2004) was identified as a key document. A synthesis of the key findings from this study is provided below. The Bureau of Reclamation study was a detailed geomorphic analysis of the Middle Hoh River between Oxbow Canyon (RM 17) to Mount Tom Creek (RM 40). The purpose of the analysis was to improve the understanding of channel processes and human impacts on the channel conditions. The analysis included channel migration mapping, hydrologic analysis, woody debris conditions, sediment estimates, and management considerations. Key findings were: The frequency and magnitude of floods has been increasing over the past 50 years Channel migration and erosion rates increased with increased flood magnitude Channel planform changes, particularly from avulsion, caused the highest localized erosion rates in the photo record (1939 – 2002) Human impacts, particularly from loss of mature riparian woody vegetation and stable instream logjams, increased erosion rates in the study area Sediment supply and transport appeared balanced based on geomorphic indicators. Field and Desktop Observations Channel migration is lateral movement of the channel and was observed to occur in two primary ways on the Middle Hoh River: 1) progressive lateral erosion of channel banks and valley margins, and 2) relict channel or topographic capture and channel realignment (avulsions). Dynamic rates of channel migration from lateral erosion were observed in field and desktop analysis. Evidence of regular avulsions was also apparent in both desktop and field observations. Mapping and interpreting lateral erosion and avulsion potential included analysis of different physical conditions that tend to drive these processes. Primary drivers and conditions observed influencing channel migration evaluated were: valley geologic composition valley vegetation sediment deposition trends channel forms. Observed Valley Geologic Composition Field observations of geologic composition revealed that post-glacial alluvial and glacial geology (low- moderate bank) groups are the predominant bank composition within the geomorphic migration zone. A recent maximum annual lateral erosion rate of 75-ft/yr was measured in post-glacial alluvial group sediments (low bank, pasture) at one location in the reach. Anecdotally, maximum annual erosion rates are reported to be double or triple that rate. The glacial group sediments were observed to have more variable erosion rates depending upon the geologic composition which ranged from lacustrine silts/clays to course outwash. The lacustrine silts appear to have lower erosion rates than the outwash but had higher potential for slope instability than the outwash deposits. The outwash deposits likely have erosion rates similar to the erosion rates of the post-glacial alluvial group deposits. The bedrock was metamorphosed marine silt/sandstone and was relatively erosion resistant over the planning-horizon of this study. For the mapping analysis, we considered three geologic groups for erosion potential mapping: Post-glacial Alluvial Group and Glacial Group (Low-Moderate Bank Height) Glacial Group (High Bank) Bedrock Group While short term erosion rates can be close to 100-ft/yr, longer term rates are influenced by the duration of erosion in any one place. Given rapid avulsion and lateral migration processes result in relatively short durations of channel occupation in a single place, we estimated lower long-term erosion potential rates. Observed Valley Vegetation Groups Mature Conifer Dominant Stands Immature/pioneering Trees and Shrubs Herbaceous Immature/pioneering trees and shrubs were observed to be the dominant riparian condition within the geomorphic migration zone. Where large, mature trees were observed in riparian areas, erosion rates appeared lower. Mature forest’s deep, dense root mass creates increased soil cohesion and roughness which reduce erodibility. We observed that where large trees were recruited to the channel, they were relatively stable and influenced channel migration processes (Figure 3a, 3b). In contrast, immature/pioneering trees and shrub zones are recruited regularly to the channel and were observed to provide little erosion resistance and had shallow rooting that was often undercut by lateral erosion (Figure 3b). Immature/pioneering trees and shrubs recruited to the channel appear to have transported downstream and did not directly influence channel migration processes. Herbaceous vegetation offered the least erosion resistance and bank cohesion due to limited root depths. Erosion rates are expected to be higher in the herbaceous and immature/pioneering trees and shrubs riparian areas (Figure 3c, Abbe and Brooks 2011). Abbe et al. 2016 presented two examples illustrating the influence of tree size on channel migration on the South Fork Hoh, one showing how the recruitment of large trees retarded erosion and initiated new floodplain development (Figure 3d) and one showing that the lack of large trees failed to limit erosion (Figure 3e). Figure 3a. Photo showing relatively stable snag in river where bank erosion recruited a large tree. The snag is deflecting flow away from the bank and helped to slow down bank erosion. Flow is from right to left (Oct. 1 2020 at RM 18.8). / Figure 3b. Photos illustrating importance of large trees in providing wood to the river that is capable of altering channel hydraulics and slowing bank erosion (top photo of stable 9 ft diameter Sitka Spruce with 30 ft rootwad in the Queets River). Bottom photo showing high terrace with young trees (industrial tree farm) where trees reaching channel are quickly transported downstream and don’t add the hydraulic roughness to the channel needed to slow erosion (Abbe and Brooks 2011). / Figure 3c. Comparison of erosion rates in the Queets and Hoh Rivers as a function of tree size. Riparian areas with trees greater than 21-in (0.53m) erode at less than half the rate of areas with trees less than 21-in (Abbe et al. 2003, Abbe and Brooks 2011). / Figure 3d. Example of where recruitment of large trees to the South Fork Hoh and channel response. resulted in not only stopping bank erosion but building new floodplain along toe of eroding bank. From 1990 to 2006 the river migrated 106 ft into an area of mature timber. Once in the channel the timber halted erosion and by 2013 the active channel had moved back to the south (Abbe et al. 2016). / Figure 3e. Example of where erosion proceeded along the South Fork Hoh due to a lack of large trees. This site is located downstream of the previous figure. Between 1990 and 2006 the river migrated 153 ft (9.6 ft/yr) into a clear-cut. Bank erosion triggered a landslide that extended 550 ft into the adjacent valley margin. The landslide headscarp retreated at a rate of 24 ft/yr from 1990 to 2013 (Abbe et al. 2016). Hazards and Risks Hazards: To understand the potential impacts related to fluvial process hazards in the project reach, it is necessary to define hazard and risk. Hazard is defined as the source of danger. For this analysis, the channel migration associated hazards are: Erosion and scour hazards Flooding, debris impacts, and deposition hazards Landslide/slope instability hazards Risk: Risk is defined as an integration of the probability of an occurrence of a hazard combined with the potential effects, or consequences, if the hazard does occur. Therefore, frequency and effect are captured by discussion of risk rather than hazard. A chart illustrating different levels of relative risk based on probability of occurrence and resulting consequences is presented below. / Mapping for this analysis presents areas in which erosion, scour, debris impacts, and scour related to fluvial process may occur within a 100-year planning horizon. No distinction of probability of occurrence is provided in this analysis, therefore relative risk levels have not been assigned. It is our opinion that specific risk assessments would be best informed by a more detailed analysis. Mapping Results The mapping assessment provides the following hazard information and is provided in Map 4: Historic Migration Zone (HMZ) Mapping: The HMZ mapping in this analysis shows channel locations occurring between 1939 and 2018. Mapping of channel locations conducted by the Bureau of Reclamation consisted of the visible wetted channel area. The 2017-18 channel mapping consisted of the visible “high flow” channel (wetted channel plus unvegetated bar areas) and low flow wetted channel. GPS mapping of the October 2020 channel thalweg was also included. Not all channel locations during this historic air photo interval were captured by the air photo record, so the true HMZ may extend beyond the mapped HMZ area. The HMZ is an area with erosion and avulsion hazards that has a high probability of occurrence. Active Geomorphic Migration Zone (GMZ) Mapping: The GMZ mapping was based on LiDAR interpretation, REM surface interpretation, FEMA flood mapping. The GMZ extends beyond the HMZ and includes an area where channel activity has recently occurred and at certain flows or with changes in channel bed elevation or wood loading, may be reactivated and experience fluvial processes. The GMZ is an area with erosion and avulsion hazards that has a moderate to high probability of occurrence. 100-Year Erosion Hazard Area (EHA) Mapping: The EHA refers to an area in which erosion hazards may be realized at some point in the future, but within the 100-year planning horizon. The probability of erosion hazards being realized within the designated EHA have not been determined for this analysis, and therefore risk has not been evaluated. Structures and infrastructure impacted by erosion hazards may be well above river water surface elevations or a long distance from today’s current channel location. Included in the erosion hazard areas are landslides that are triggered by lateral channel migration and expand upgradient from the channel margin (see Geotechnical Hazard Area). Professional opinion of erosion rates was applied. This opinion considered bank height, geologic composition (younger alluvium, older alluvium, glacial diamicts, or bedrock), vegetation type and age. Geology type was inferred by mapped geology, LiDAR, and field observations. The most erodible banks are low height, younger alluvium, with grass vegetation. Short-term erosion rates in this bank type may be on the order of 100-ft/yr, perhaps higher. Erosion rates over a 100-year time frame must also consider the duration time that the river channel is present at a bank. In most cases on the Hoh, channel migration processes limit the duration of time a channel is actively eroding a bank to a few years before the channel migrates to a new location. The frequency that a channel returns to a bank over the assessment time frame must also be considered. For this analysis, the 100-year erosion rate for low banks with younger alluvium and immature vegetation was a minimum of 500 feet. Bedrock banks were assumed to have a 100-year erosion rate of zero feet. The EHA is an area with erosion and avulsion hazards that has a low probability of occurrence. Avulsion Hazard Area (AHZ) Mapping: The AHZ describes an area within the HMZ and GMZ that is susceptible to rapid channel course changes or temporary channelization of flow that would become an erosion hazard. While erosion hazards are often realized incrementally and avoidance or retreat is an option to manage risk, avulsion hazards may be realized abruptly and potentially catastrophically, thereby potentially increasing consequences and decreasing viability of avoidance or retreat management strategies. Avulsions are common on the Hoh River, thus exacerbating risk. Once a relict channel is reclaimed in an avulsion, rapid lateral erosion and channel widening of this channel results. Over the course of the 100-year planning horizon of this delineation, the GMZ can be considered synonymous with the AHZ, as any of the relic channels defining the GMZ could pose an avulsion risk over time as the channel migrates. Therefore, the AHZ is not specifically identified in the delineation. Geotechnical Hazard Setback Area (GHA) Mapping: The GHA mapped area has potential slope instability driven by channel migration processes. For this study the fluvial-related geotechnical hazards are considered if the 100-year EHA contacts steep slopes. The GHA is shown as an overlay on the maps. Site specific geotechnical conditions should be considered for risk analyses within the GHA. The GHA is an area with erosion and landslide hazards that has a low to high probability of occurrence based on the location of the channel relative to the mapped GHA. Alluvial Fan Hazard Areas (AFH) Mapping: The AFH defines an area where a sharp decrease in channel slope creates a wedge of sediment to deposit as stream energy is quickly lost, creating a characteristic fan shape in plan view. Alluvial fans are common at the margins of valley bottoms and are highly dynamic environments prone to rapid channel avulsions and aggradation. The AFH has the potential to be an erosion hazard, aggradation hazard, flooding hazard and/or debris impact hazard or any combination thereof and may change over time as the fan evolves. The AFH is an area with erosion, aggradation, flooding, and debris impact hazards that has a high probability of occurrence based on the high energy of the environment. Channel Migration Zone (CMZ) Mapping: The cumulative area and hazards associated with the HMZ, GMZ, EHA and GHA areas define the extent of the CMZ. Summary and Conclusion The Middle Hoh is a dynamic landscape that has undergone historic disturbances that have influenced natural processes, and faces a future of additional changes that will further alter natural processes. Historic clearing of the riparian forests altered the rate of channel migration in the reach, making the river more dynamic and less predictable. Future climate changes will alter the flow regime, sediment supply and transport capacity of the river, with the potential to fundamentally change the character of the river. Given the uncertainty in what changes will occur and when they will be realized we recommend planning conservatively, as the past is becoming less and less a predictor of what to is to come. An example of this is show in Figure 4, where the river had migrated beyond the 2004 CMZ delineation in 2006, continuing to erode through 2018 expanding 100-ft past the boundary. As the river continues to adjust to future changes, we can expect the river to occupy and flood areas that have been safe during modern history. Planning ahead of time and developing adaptative management strategies are key to mitigating risks resulting from future channel migration and increased peak flows. Figure 4. Comparison of Bureau of Reclamation 2004 mapping to updated mapping conditions and revised methodologies. Image at left is 2004 channel location mapping (colored ribbons) and channel migration zone (black hatched zone). Image at right is updated channel mapping with historic migration zone (red), geomorphic migration zone (dark orange), erosion hazard area (light orange) and geotechnical setback (yellow), all included in the updated CMZ limit (black boundary). Note that the current channel is outside of the 2004 delineated channel migration zone. Limitations Limitations of this analysis include: Mapping scale – temporal and spatial: The mapping efforts for this analysis were done on a reach scale. Site-specific investigation may be necessary. Site-specific investigations and validation of mapping was conducted only sparingly because of a limitation of time and resources. Temporal changes, disturbances, or recovering conditions, such as riparian development or wood jams, are variables that will impact channel response. Predicting the local response and impacts over the time scale of the study is not considered feasible because the unpredictable nature, time scale, and extent of change. For example, predicting the location of the channel, size and duration of a wood jam, and riparian conditions at a site even in 5-years is felt to have a low degree of assurance. Analysis looking at a smaller time context and a more site-specific or sub-reach perspective should include such conditions and potential channel response. New information and technology – new interpretations: Improvements in technology, new information and geologic interpretations may allow for a higher confidence in mapping or reinterpretation of mapped hazards. Maps should be updated as new information and technology becomes available. Hazards that exist but were not mapped for this assessment: The hazards mapped for this analysis were limited to lateral erosion and avulsion potential on the Hoh River. Other natural hazards in the valley exist, including but not limited to flooding, landslides, tectonic deformation. Riparian and Floodplain Forests The Hoh River valley floodplains and terraces are the home of the extraordinary Olympic rain forest. Although the Middle Hoh River floodplain is predominantly second and third growth forest and recently recruited red alder. There are also some remnant patches of mature forest hundreds of years old. Just up-river within Olympic National Park are stands of riparian forest four to seven hundred years in age (Fonda 1974; Kramer et al., 2020). The Olympic rain forest provides few limits to forest growth and old growth trees reach diameters of 2-3+ meters and up to 80-90 meters in height (Fonda 1974; Kramer et al., 2020). The structural size and complexity of the Hoh River riparian forest provides the river with in-channel large wood that the river hydraulically organizes into stable wood jams forming the structural skeleton of river channel splits, mid-channel islands and patchwork floodplains (Montgomery and Abbe 2006). Compositionally the young riparian forest consists of early successional willows (Salix L.), red alder (Alnus rubra), and black cottonwood (Populus balsamifera trichocarpa T. & G.). As the young pioneer forest matures (25-50 years) conifers colonize the floodplain–primarily Sitka spruce (Picea sitchensis)–with minor elements of western hemlock (Tsuga heterophylla), Douglas fir (Pseudotsuga menziesii ) and western red cedar (Thuja plicata). Later colonizers include deciduous trees–vine maple (Acer circinatum) and big leaf maple (Acer macrophyllum). As the riparian forest matures over 50-100 years, Sitka spruce and black cottonwood reach key member size diameters of 100 cm or more. Circa 100 years, and older, the riparian forest becomes the source pool for recruitment of large trees to the river channel a phenomenon called the floodplain large wood cycle (Collins et al. 2012). The floodplain large wood cycle is the natural process that generates Hoh River channel complexity, mid-channel islands, pools, salmonid aquatic habitat and overall valley complexity and habitat diversity. The foundation of the Hoh River aquatic and riparian ecosystem are the large trees and the ecological processes that generate these trees. The interaction of abundant precipitation, large trees, and river and gravel/cobble channel bed together produce the diversity of habitats that are the home of Hoh River salmon, bear, cougar, osprey, eagle, and a myriad of other creatures. The goal of the Middle Hoh River riparian forest study was to characterize the current riparian forest composition and structure to inform development of a forest conservation and restoration and conservation strategy. Methods & Study Area Forest Mapping & Characterization Middle Hoh River study area includes river floodplains and terraces along the riparian corridor from Olympic National Park to Highway 101. Riparian forest mapping was conducted in a GIS (ArcMap) environment using 2015 NAIP imagery and a 2014 first return LiDAR digital surface model (DSM). The first return LiDAR DEM provides a mapped measurement of vegetation height (Figure 5). The NAIP color imagery allows mapping of coarse scale monotypic forest stands. Forests were typed in three cover classes and four height classes (Figures 6, 7 & 8). Cover classes included: (1) > 75% cover deciduous trees (willow, red alder, black cottonwood, big leaf maple, and vine maple; (2) > 75% cover conifers (Sitka spruce, Douglas fir, western hemlock and western red cedar); and cleared land (pasture, development); and (3) mixed conifer deciduous (<75% deciduous & <75% conifers). Five height classes were measured with the LiDAR DEM: <5-ft, <25-ft, 25-75-ft, 75-125-ft, and >125-ft. l Park / Figure 5. First return 2014 LiDAR DEM of forest vegetation heights / Figure 6. Middle Hoh River riparian forest type cover and height classes. / Figure 7. Hoh River riparian forest mosaic of cover types. October 1st, 2020 near RM 20.6 / Figure 8. Forest typing and channel migration zone (CMZ). Results & Discussion The Middle Hoh River riparian forest was mapped within the CMZ and outside the CMZ to capture the natural channel disturbance dynamics within the CMZ as compared with older terrace and adjacent hillslope surface processes (Figure 6). The Hoh River CMZ riparian forest is a mosaic of patches of young pioneer red alder flats to mature mixed conifer deciduous forest stands (Figures 9, 10 & 11; Table 1). Channel migration through erosion and sedimentation generates new surfaces that are colonized by willows, red alder and black cottonwood. As the forest matures, within 25 years, conifers that will grow to maturity over hundreds of years, colonize. Deciduous forest (red alder, willow, black cottonwood) make up 41% of the CMZ cover types, all of which are <125 feet in height, descriptive of young to mature red alder floodplain forests less than 80 years (Table 1). Mixed conifer deciduous type comprises 32% of the total CMZ forest cover (Table 1). Together the deciduous and mixed cover types total 73% of the entire CMZ forest cover. While coniferous cover type comprises 19% of the total CMZ forest (Table 1). The height structure of the riparian forest generally reflects age, with younger trees being younger and taller trees older. The >125 foot coniferous and mixed classes comprise 24% of the total forest cover, while <75 foot classes comprise 44%. This height distribution is reflective of a younger CMZ red alder dominated deciduous forest that has patches of mature conifer and mixed conifer deciduous stands dispersed throughout. Within the CMZ channel disturbance is the primary control of riparian forest composition and structure. While large key member size trees and stable wood jams generate mid-channel islands and stable floodplain surfaces, refuge sites for mature floodplain forests to develop. The forest composition and structure outside the CMZ, by contrast to the CMZ forest, is controlled primarily by land use–clearing, timber operations and development. Wind also plays a role in structuring mature forest stands. Outside the CMZ the forest cover is dominated by Mixed (44%) and Coniferous (33%) cover types (Figure 12; Table 2). Deciduous forest makes up only 16% of the total non-CMZ cover as compared to CMZ’s 41%. Forest stands >125 feet comprise 25% of the total forest cover indicating forest stands reaching a century in age. Comparing CMZ to outside CMZ forests is a bit like comparing apples and oranges. The primary drivers of forest types and structure are different, channel migration and presence of large wood within the CMZ and land-use practices outside the CMZ. Regarding forest conservation and restoration consideration of these process drivers is critical to a successful strategy. / Figure 9. Young deciduous (red alder and willow) and older mixed conifer deciduous (red alder and Sitka spruce) forest types. October 1, 2020 near RM 20.6 / Figure 10. Mature mixed (Black cottonwood, red alder, Sitka spruce) and conifer forest types (Sitka spruce, Douglas fir). October 1, 2020 near RM 20.7 / Figure 11. Inside CMZ. Riparian Forest Type (Acres) and Height Class (Feet/Acres) for 2014 Table 1. Inside the CMZ. Forest type and height class for 2014. HEIGHT (FT) CLEARED (ACRES) CONIFEROUS (ACRES) DECIDUOUS (ACRES) MIXED (ACRES) GRAND TOTAL (ACRES)  <25 165 16 76  257  25-75  116 476 142 734  75-125  28 362 334 725  >125  272  260 532  Grand Total 165 432 914 737 2248   / Figure 12. Outside CMZ. Riparian Forest Type (Acres) and Height Class (Feet/Acres)for 2014. Table 2. Outside the CMZ. Forest type and height class HEIGHT (FT) CLEARED (ACRES) CONIFEROUS (ACRES) DECIDUOUS (ACRES) MIXED (ACRES) GRAND TOTAL (ACRES)  <25 222 175 67 44 507  25-75  162 287 311 760  75-125  277 91 632 1000  >125  376 24 339 739  Grand Total 222 990 468 1325 3006   Hydrology & Hydraulics A hydraulic model of the Middle Hoh River was developed using Hydronia’s RiverFlow-2D Plus GPU and Aquaveo SMS v13.0 computer software. RiverFlow-2D is a two-dimensional finite volume computer model that provides depth averaged hydraulic parameters at centroids within a triangular mesh model domain. The model domain encompasses 15-RM, with the upper boundaries on the Hoh River and the South Fork Hoh River upstream of the National Park boundary, and the lower boundary just downstream of the Highway 101 bridge. Hydrology and Boundary Conditions To develop estimated inflows for the Middle Hoh River model, NSD performed a hydrologic analysis of the region. This analysis focused on two USGS gages: USGS gage #12041200, Hoh River at US Highway 101 near Forks, WA (henceforth referred to as the Highway 101 gage) and USGS gage #12041000, Hoh River near Forks, WA (henceforth referred to as the upstream gage). The downstream gage’s period of record is 1960-2021, but the upstream gage was retired in 1964, though it has data all the way back to 1921. Since the model domain is so large, both gages were analyzed to help determine model hydrology. NSD performed a flow frequency analysis on both gages using the Log-Pearson schedule 3b methodology (USGS 1981), the results of which are shown in Table 3. Table 3. Estimated peak flows at each gage in the model domain. USGS GAGE DRAINAGE AREA (SQ MI) 1-YR FLOOD (CFS) 2-YR FLOOD (CFS) 10-YR FLOOD (CFS) 100-YR FLOOD (CFS)  # 12041200 (Highway 101 gage) 253 12,260 32,700 52,300 73,600  # 12041000 (Upstream gage) 208 8,190 19,000 30,800 46,000   The model was run for a series of three representative flow scenarios – the 1-year, 10-year, and 100-year floods – to evaluate hydraulic parameters at the project site. There were 10 inflow locations to this model – their locations are illustrated in Figure 13. The magnitude of all modeled flows at each of 10 inflow locations is shown in Table 4. The mainstem inflow was determined using the peak flow analysis of the upstream gage, scaled by drainage area to the mainstem inflow location using the weighted USGS scaling method presented in Mastin et al 2016. Discharge at the remaining inflow locations was determined using Streamstats where available, and when Streamstats was not available the discharge was calculated using the USGS regional regression equations. Because only the 9 largest tributaries were explicitly modeled, the flow contributed by several smaller tributaries as well as seepage and hillslope runoff within the model domain needed to be accounted for for the model discharge at the downstream end to approximate the 10-year and 100-year peak flows estimated from the downstream gage. This was achieved by adding a small amount of additional discharge to each inflow location proportional to its drainage area. For the 1-year flood, the inflows were reduced in proportion to their drainage area to achieve the same goal. For each inflow location, the model inflow which has been adjusted to account for hillslope runoff is shown Figure 13, Table 4. The models were run in a steady-state (constant discharge) and have only one outflow boundary located just downstream of the Highway 101 bridge and the downstream gage. The outflow boundary condition was set to a constant water surface elevation (WSE) corresponding to each flow event. These WSEs were based on analysis of the rating curve of the Highway 101 gage. Downstream WSE was set to 175.03, 185.53, and 189.81 for the 1-year flood, 10-year flood, and 100-year flood respectively. / Figure 13. Locations of model inflow and outflow locations illustrated over the model domain. Table 4. Modeled Discharge Values at Inflow Locations INFLOW LOCATION DRAINAGE AREA (SQ MI)  1-YR  10-YR  100-YR  Hoh River 124.9 7,700 27,200 35,900  SF Hoh River 53.8 4,080 13,900 17,700  Winfield Creek 11.6 60 2,100 3,950  Clear Creek 6.9 20 480 880  Maple Creek 9.5 100 1,480 2,620  Owl Creek 2.4 200 2,200 3,760  Hell Roaring Creek 5.5 30 1,020 1,920  Alder Creek 8.3 90 1,670 3,050  Tower Creek 4.5 - 1,140 1,920  Spruce Creek 4.5 - 1,140 1,920  Total  12,280 52,330 73,620  Highway 101 Gage Estimate  12,260 52,300 73,600   Mesh Development and Roughness Categories The topographic data was developed by combining information from LiDAR data collected in 2012 and 2013, as well as bathymetry data that was provided by Wild Fish Conservancy from a 2013 survey. The model mesh was created with fine mesh spacing in the main channel and in floodplain channels, with expanded mesh spacing in less topographically complex areas further from the stream. Hydraulic resistance is characterized in the model by polygons representing differing surface roughness types such as mainstem channel, forest, or bare earth. The full list of roughness categories and their associated Manning’s n values is included in the calibration section in Table 5. Due to the large domain of this model, roughness categories were assigned in as automated a fashion as possible. First, to attain a base of vegetation roughness, NSD subtracted the bare earth LiDAR from the highest hit LiDAR to attain a 15 ft x 15-ft raster of vegetation height split into 5 categories: <1’, 1-4’, 4’-6’, 6’-12’, and >12’. To map roads onto this vegetation base, the lines in the DNR roads shapefile were given a 10’ width and overlaid. The locations of the roads were spot-checked against aerial imagery to confirm their accuracy; when discrepancies were identified, the road locations were edited to reflect aerial imagery. The low flow channel was then delineated using a wetted perimeter polygon derived from the combined surface, and on top of that the tributaries, side channels, and logjam roughness areas were manually delineated using 2013 aerial imagery. Modeled Infrastructure NSD identified 5 road crossings that are directly engaged by the Middle Hoh River or by the modeled tributaries: 3 near the Clear Creek confluence, 1 upstream of the Clear Creek confluence, and 1 on Spruce Creek before the confluence with the mainstem. At these locations, NSD modeled the road crossings using the Riverflow2D culvert tool, which accounts for culvert shape, size, material, slope, and inlet configuration using the FHWA procedure (Norman et al. 1985). Details on the three culverts near Clear Creek were provided by the 10,000 Years Institute from field observation; details on the two remaining culverts were estimated based on the surrounding LiDAR data. Calibration The 2D model was calibrated by comparing the modeled 100-yr inundation area to the digitized FEMA base flood map, effective June 7, 2019, and adjusting the model to bring it as close to the FEMA floodplain as possible. The main tools available to calibrate the 2D model were adjusting the hydraulic roughness parameters using the Manning roughness coefficient (n-value) for various land cover types. Initial values were set for the roughness coefficients based on recommended values from previous studies and professional judgement. After the initial run, results were compared with the FEMA base flood and minor adjustments were made to optimize calibration. Final calibrated roughness values are shown in Table 5. Map 6 includes a series of maps comparing the calibrated model to the FEMA floodplain. The areas in which the two do not agree can be explained by differences in tributary modeling – for instance at small tributaries where the model does not include a specific inflow, the FEMA floodplain shows more inundation than the model, and conversely where tributary inflows were inflated to account for surrounding drainage area, the FEMA floodplain shows less inundation than the model. The focus in calibration was on flooding related to mainstem flow. Table 5. Calibrated Manning's n roughness values for each roughness category. CATEGORY MANNING'S N VALUE  Veg <1' 0.04  Vegetation 1'-4' 0.07  Vegetation 4'-6' 0.08  Vegetation 6'-12' 0.08  Vegetation >12' 0.1  Road 0.016  Building 0.5  Active Channel 0.035  Tributary 0.045  Revetment 0.1   Results Maps of hydraulic model results for depth, velocity, and shear stress during the 1-yr, the 10-yr, and the 100-yr floods are included in Map 6. The depth and velocity results were spatially separated based on location in the floodplain vs. the active channel and then summarized for each of 5 reaches within the model domain. The resulting depth and velocity averages are shown in Table 6 and 7, respectively. Table 6. Average depth by reach for the modeled peak floods. AREA FLOOD EVENT MEAN DEPTH (FT)    HUELSDONK-SOUTH FORK REACH SPRUCE CANYON REACH MORGANS CROSSING REACH WILLOUGHBY CREEK REACH OXBOW CANYON REACH  Channel 1-yr 4.4 7.3 5.1 4.6 8.5   10-yr 6.9 15.8 8.7 8.9 19.8   100-yr 7.9 19.0 9.9 11.5 24.2  Floodplain 1-yr 2.6 3.9 3.1 3.1 5.0   10-yr 3.7 6.6 3.6 5.2 10.0   100-yr 4.3 8.2 4.3 7.6 12.2   Table 7. Average velocity by reach for the modeled peak floods. AREA FLOOD EVENT MEAN VELOCITY (FPS)    HUELSDONK-SOUTH FORK REACH SPRUCE CANYON REACH MORGANS CROSSING REACH WILLOUGHBY CREEK REACH OXBOW CANYON REACH  Channel 1-yr 5.1 8.0 5.6 4.8 7.4   10-yr 6.9 12.3 7.7 5.9 12.4   100-yr 7.4 13.2 8.3 5.9 13.8  Floodplain 1-yr 2.9 4.0 3.2 3.2 3.8   10-yr 3.6 3.7 2.8 2.3 4.7   100-yr 3.9 3.7 3.1 2.3 4.9   The average depth in the mainstem ranges from 4.4-8.5-ft in the 1-yr event, 6.9 - 19.8-ft in the 10-yr event, and 7.9 - 24.2-ft in the 100-yr event. The range of average depth at high flows is large due to the geomorphic differences between the reaches: the mainstem is tightly confined in Oxbow Canyon reach and part of Spruce Canyon reach, resulting in high depths compared to the other reaches like Huelsdonk-South Fork reach, where flow is spread out across the flood plain The average velocity in the mainstem ranges from 4.8 - 8.0-feet per second (fps) in the 1-yr event, 5.9 - 12.4-fps during the 10-yr event, and 5.9 - 13.8-fps in the 100-yr event. Floodplain velocities remain in the range of 2.3 - 4.9-fps regardless of flood event recurrence interval. As with depth, the highest mainstem velocities are in the confined reaches while the lower velocities are associated with reaches where flow can overtop the banks and engage the floodplain. Oxbow Canyon Reach Flow conditions in the Oxbow Canyon Reach are the deepest on average for the entire 15-mi project reach due to near complete confinement of flow in the channel up to the 100-yr flood (Table 6). Velocities rival those found in Spruce Canyon with the greatest average velocities in Spruce Canyon during the 1-yr flood, transitioning to Oxbow Canyon as flows approach and exceed the 10-yr flood. A small floodplain is present on the right bank downstream of RM 17, and backwater eddy at RM 16 on the left bank that provide high flow refugia within the canyon during floods exceeding the 5 – 10-yr flood (Map 6). Willoughby Creek Reach Upstream of the Oxbow Canyon Reach the floodplain abruptly widens and flows begin to spread out as the channel traverses the Willoughby Creek Reach. At the downstream end of the reach, as the channel enters Oxbow Canyon, a backwater forms that rapidly floods the adjacent floodplain and diminishes instream velocities. The influence of this backwater propagates upstream with increasing flow, extends over a mile upstream during the 100-yr flood, and influences instream and off-channel flow velocities and depths (Map 6). The effect of this backwater also explains the lack in an increase the average channel velocity from the 10-yr to the 100-yr flood (Table 6). During the 1-yr flood side channels are activated in the Elk Creek Floodplain area as well as 2.1-mi long side channel following the valley margin on the left bank floodplain from Peterson’s Bottom to Elk Creek. Velocities in the main stem channel are greatest where the channel is a single thread and diminish rapidly as the channel bifurcates in more and more channels (Map 6). Morgans Crossing Reach The Morgans Crossing Reach extends from Peterson Bottom to the downstream end of Spruce Canyon and is naturally unconfined with a broad floodplain through the entire reach. Downstream of Rock Creek the channel bifurcates into multiple threads during the 1-yr flood and floodplain side channels are activated in the Lindner and Clear Creek floodplains. A backwater extends approximately 2000-ft up an abandoned main stem channel at the Rock Creek confluence, upstream of which flow is confined to the main stem channel to the upstream end of the reach. As flows increase to the 10-yr flood, water exits the channel banks and there is broad inundation of floodplains throughout the reach (Map 6). Between RM 21 – 22.6 over-bank flooding crosses the Upper Hoh Road near Lindner Creek Lane during the 10-yr flood and continues downstream north of the road following a historic side channel down to Lindner Creek, crossing the road back to the channel next to the Peak 6 Adventure Store. When flow approaches the 100-yr flood the length of the Upper Hoh Road inundated increases and extends up to the Morgans Crossing floodplain below the Tower Creek confluence. Spruce Canyon Reach The shortest reach in the Middle Hoh is the Spruce Canyon Reach, starting below the Owl Creek confluence at Spruce Island, ending at the downstream end of Spruce Canyon. Flows in the Spruce canyon are similar in nature that that in Oxbow Canyon, the biggest difference being flows are in general deeper in Oxbow Canyon. The channel is confined to the bedrock canyon throughout the reach, minus the split around Spruce Island at the upstream end of the reach. Overall, the Spruce Canyon Reach has the highest average velocity in the Middle Hoh for more frequent high flows less that the 10-yr flood. Near the 10-yr flood overbank flow enters a gravel pit on the right bank floodplain at the upstream end of the reach, exiting the other side and continuing downstream on the floodplain to the Spruce Canyon inlet (Map 6). Huelsdonk-South Fork Reach At the upstream end of the Middle Hoh is the Huelsdonk-South Fork Reach, extending from Spruce Island to ONP. Average channel flow depths in this reach are the lowest of all reaches and are the shallowest or close to on the floodplains (Table 6). Velocities are consistently second lowest for all flows, behind the Willoughby Creek Reach (dampened velocity due to backwater above Oxbow Canyon). The main stem channel flows unconfined through the reach, meandering across the valley bottom at low flows. As flows increase historic main stem flow paths begin to activate forming large islands that remain dry during the 1-yr flood, short of a few side channels traversing the islands (Map 6). With increasing flow, the islands and adjacent floodplains become submerged but does not flood the Upper Hoh Road. The channel has however expanded into private property in the Brandabury and Fletcher Ranch areas, eroding land and damaging structures in the process. The main stem channel through this reach has been actively avulsing and migrating in recent history, frequently adjusting the nature of hydraulics in the reach. The topographic surface available for this modeling does not represent the current condition, however predicted flow characteristics in the channel and floodplain area anticipated to be similar. The location of the main stem channel will dictate where high velocity areas are located, and recent avulsions and bank erosion will expand the area inundated and increase flow depths were occurring. Aquatic Habitat The Hoh River is located on the West Coast of Washington where it originates from the glaciers of Mount Olympus and runs through the Olympic National Park, through the foothills and out into a wide floodplain where it eventually enters the Pacific Ocean. The upper 65% of the Hoh River Basin area is located within the ONP and is considered to be pristine temperature rainforest and home to some of the largest old growth trees in the United States (McMillan and Starr 2008; NPCLE 2020). Downstream of the park, the mainstem Hoh is still considered to provide quality fish habitat, but stressors from infrastructure encroachment, logging, and climate change are present. The Hoh River hosts five salmonid species, including populations of spring and fall Chinook Oncorhynchus tshawytscha, fall coho O. kisutch, and winter steelhead O. mykiss (McHenry et al. 1996). The Hoh River salmon stocks are not listed under the endangered species act (ESA) and are considered to be among the last healthy wild populations in the contiguous United States, however recent declines in adult returns have been documented (Huntington et al. 2004; Hoh Tribe 2016; Cram et al. 2018). The runs of interest differ in their timing and use of riverine habitats, but Chinook, coho, and steelhead all rely on the mainstem Hoh for both rearing and spawning (McHenry 2001; McMillan and Starr 2008). Salmon spawning occurs in the mainstem throughout the year, but the timing depends on the species. Spring Chinook generally spawn in August and September, fall Chinook and coho spawn from October through December, and winter steelhead spawn from December through July (McMillan and Starr 2008). These species also utilize the mainstem habitat differently as juveniles. Most spring and fall Chinook out-migrate at age-0, while coho and steelhead may remain in the river for a full year. Coho and steelhead are also documented to use more lateral mainstem (braids and side channels) and tributary habitats during rearing than Chinook, which primarily rely on main channels of the mainstem (McMillan and Starr 2008; Quinn 2018). However, all species are documented to use braid and side channels for spawning. Adequate freshwater migration, spawning, incubation, and rearing habitat is important for all these species (Quinn 2018). All three species require suitable sized gravel and cobble for spawning and rearing with low levels of fine sediment. Winter rainfalls and the resulting flooding can exacerbate erosion and landslides, scour and damage redds, and increase infiltration of fine sediment reducing oxygen levels available to embryos, which in turn decreases the chance of survival to emergence (Quinn 2018). Furthermore, overwinter rearing habitat is particularly important for juvenile coho and steelhead due to their extended stay in freshwater compared to Chinook. Lack of adequate summer flows, water quality and temperatures, channel complexity, slow water habitat, and off-channel and floodplain connectivity can contribute to low salmonid rearing and spawning success (Ames et al. 2000; Hoh Tribe 2016). While the status of salmon stocks in the Hoh River are generally considered to be healthy, recent declines have prompted cause for evaluation (Smith 2003; Hoh Tribe 2016). The Middle Hoh has been impacted by the loss riparian old growth forest and reduction of large wood inputs, erosion, landslides and increased fine sediment pulses, and decreased water quality (Parks 1999; Hoh Tribe 2016). Additionally, this Middle Hoh Study Area has observed dramatic channel migration in recent years, with movements of over 60 feet a year, and local landowners and the county are concerned about the loss of property and road infrastructure, as well as the impacts on salmon and instream habitat quality (Chadd 1997; Piety 2004; Hoh Tribe 2016). The goal of this habitat assessment is to characterize current mainstem aquatic habitat conditions regarding salmonid use to support the development of a resiliency plan that addresses flood hazards and benefits salmon in the Middle Hoh River Study Area from RM 16 to RM 31. Methods Low flow habitat surveys on the Hoh River mainstem were completed by Cramer Fish Sciences (CFS) from upstream of the Highway 101 Bridge (RM 15-16) to the National Park Boundary (RM 30-31) from September 28th to October 1st. Average daily flows during the surveys ranged from 1,614 to 2,652 cubic feet per second (cfs) at the Highway 101 bridge (USGS Gage 12041200; Figure 14. All main channel watercourses were surveyed; and braid and side channels were surveyed as time allowed (Leopold and Wolman 1957; Peterson and Reid 1984). Wetted side channels and braids not surveyed were captured with photos and their diversion point from the main channel was captured by GPS coordinates./ Figure 14. Instantaneous flow (cfs) data at USGS Gage 12041200 – Hoh River at Highway 101 near Forks, WA recorded and reported by USGS. Surveys were conducted between September 28th and October 1st, 2020. Mainstem surveys were completed moving downstream by boat. Reaches were determined by geomorphic characteristics (Piety et al. 2004). Channel type was recorded as main, braid, or side channel; main channels of the mainstem were identified as channels with the most flow; braids were identified as one or more anastomosing channels connected to the main channel, and side channels identified as channels within the mainstem floodplain separated by islands with stable woody vegetation (Leopold and Wolman 1957; Peterson and Reid 1984). Habitat units were identified as pools (non-turbulent), riffles (fast-turbulent), or glides (fast non-turbulent) (Bisson et al. 1982; Beechie et al. 2005; CHaMP 2016). Pool type (e.g., plunge, scour, dam) and pool-forming feature were recorded for pool units (Bisson et al. 1982). Pool frequency was calculated as the pools per mile of channel length and pool spacing (channel widths per pool) was calculated using the Montgomery et al. (1995) formula as the channel length per wetted channel width divided by the number of pools. Pool frequency targets were identified using NMFS (1996) (Table 8). Lengths and wetted widths were recorded in meters using a laser rangefinder. All metric units were converted to English units in processing. The GPS coordinates of the top and the bottom of each dominant unit were recorded, GPS units were also used to record tracks of the channels surveyed. Recorded data outputs are available in Appendix A. Table 8. Pool frequency standards for functioning rivers developed by NMFS (1996). CHANNEL WIDTH (FEET) # POOLS/MILE  5 184  10 96  15 70  20 56  25 47  50 26  75 23  100 18   For each riverbank of each unit, the percent of length occupied by each edge type at the wetted edge was estimated. Edge types were recorded as bank edge (natural or hydro-modified) or bar edge (Hayman et al. 1996). Hydro-modified banks were identified as banks where modifications were visually observed, such as levees or riprap, however the quality was not recorded. If banks were modified but the modification could not be identified, they were categorized as “Hydro-modified unknown.” Bar edges were assumed to be naturally formed (Hayman et al. 1996). The width of slow water was recorded if present for each edge type. Slow water area was defined as present when a boundary between the edge type and mid-channel was visible as a current shear line; slow water edge width was assumed to be a minimum of 1-m wide for all banks. Large wood jams (LWJs) were mapped using aerial photography collected in March 2021 (NV5 GeoSpatial 2021; Figure 15). The average daily flow during collection was 3,366 cfs at the Highway 101 bridge (USGS Gage 12041200; Figure 16). We digitized all jams visible within the bankfull channel, including wood visible in the water, on gravel bars, and on vegetated islands (Beechie et al. 2017). All jams composed of three or more wood pieces 7.6 m in length by 0.3 m in diameter were manually delineated using QGIS at a zoom level of 1:600 or higher, and the full perimeter of the jam was recorded (Leif et al. 2004). The location of the jam in the mainstem (main channel and braids, side channel, or islands), the channel location (bar, bank, or island), and whether the jam was wet or dry were recorded for each jam. Jam locations were also recorded during field surveys, however, considerable channel movement led to changes in jam positions and areas, therefore all reported jam metrics are from the aerial imagery analysis. / Figure 15. Example of jams delineated in the Morgans Crossing Reach (RM 21.9), shown at a 1:600 scale using aerial photography collected in March 2021 (NV5 GeoSpatial 2021). Blue polygons depict wood in wetted channel (time of photo) and red polygons wood outside wetted channel but within ordinary high water (bankfull) channel. Flow is from right to left. / Figure 16. Instantaneous flow (cfs) data at USGS Gage 12041200 – Hoh River at Highway 101 near Forks, WA recorded and reported by USGS. Aerial imagery was collected on March 20th, 2021. Results A total of 28.8 miles of mainstem habitat were surveyed in the Middle Hoh River Study Area from RM 16, at the Oxbow Campground to RM 31 at the National Park boundary (Table 9). This included 15.8 miles of main channel, 8.0 miles of braided channel, and 4.9 miles of side-channel habitat. Additional braid and side-channel habitats wetted at the time of the survey were not surveyed because of time constraints. Wetted braids and side channels were present in all reaches except in the Oxbow Canyon Reach (Map 7). Braid and side-channel habitat was the most abundant in the Huelsdonk-South Fork Reach, followed by the Morgans Crossing Reach (Table 9). Overall, glides were the dominant habitat unit with 48% of the surveyed wetted channel area observed in glide units followed by riffles and rapids which made up 31% of the surveyed wetted channel area, and pools and backwater units, which accounted for the remaining 21% of wetted channel area (Figure 17). It should be noted, that because we were unable to measure depths and calculate residual depths, pools were identified by channel and large wood features or changes in surface flow and hydrology (i.e., the presence of slow water), and therefore pools may have been classified as glides and underestimated if no visual clues were present. Riffles and rapids were the most frequently observed units, with eight units per mile, followed by pools and backwater units (six units per mile), and glides (five units per mile). Of the reaches surveyed, the Huelsdonk-South Fork Reach had the highest total number of pools, total area in pools, and the highest pool frequency (pools per mile), but also had the most channel length surveyed (Table 10). The Spruce Canyon Reach had the fewest pools and lowest pool frequency (pools per mile), but surveys were largely limited to the main channel in this reach. Trench, scour, and backwater pool units were observed (Figure 20). Trench pools were formed by bedrock and canyon features and were primarily observed in the Oxbow Canyon Reach (Figure 21). Meanders were the primary pool forming feature identified in main channel habitats, followed by bedrock features and large wood jams and pieces (Figure 21). Large wood frequency was not significantly correlated to pool frequency across the reaches but there was a slight trend indicating that there was a positive relationship between LWJ frequency and pool frequency (Figure 22). The majority of bank habitat in the Hoh mainstem was made up of bar edges (55% of edge length) and natural bank edges (42% of edge length) (Figure 18; Table 9). However, hydro-modifications including rip rap, roads, and other bank armoring and residential bank modifications were documented in all reaches except the Oxbow Canyon (Figure 18). Rip rap and bank armoring were most prevalent in the Huelsdonk-South Fork Reach, with 1.3 miles of hydro-modified banks present over ten locations, followed by the Willoughby Creek Reach with 0.8 miles of hydro-modified banks present over five locations (Figure 18; Table 9). Slow water edge area is important mainstem habitat for juvenile salmonids (Beamer et al. 2005). We observed bar edges and natural bank edges to have the greatest associated average slow water edge width (4.1 feet) and total area, followed by hydro-modified bank edges (3.7 feet) (Figure 19). Wood was mapped using aerial imagery collected six months after the low flow habitat surveys, during which three flow events over 20,000 cfs occurred leading to multiple changes in the channel form. Wood values are reported as they were observed in the March 2021 aerial imagery, however, the field data recorded in the low flow 2020 habitat surveys suggest trends in wood distribution were consistent between the surveys. The majority of the 2,573,637 square feet of mapped LWJ area was located on main and braided channel gravel bars (64%), followed by LWJs attached to main and braided channel banks (24%) (Table 11). Side channel and island jams accounted for the remaining 12% of the total mapped jam area (Table 11). The majority of the LWJs mapped were dry, however LWJs with some wetted area accounted for most of the mapped area (Table 12). The Huelsdonk-South Fork Reach contained the most LWJs and total overall, wetted, and dry LWJ area, but the Morgans Crossing Reach had the highest LWJ frequency (LWJ/mile) (Table 12). The mean and median LWJ sizes were above the full study area value (5,066 square feet and 2,813 square feet, respectively) in the Huelsdonk-South Fork, Spruce Canyon, and Morgans Crossing reaches (Table 12). Jam sizes were smaller than the study area average in the Willoughby Creek and Oxbow Canyon reaches, likely due to the fact that majority of large wood transport from the ONP would be racked in the upstream reaches and the riparian area within the study area is dominated by deciduous and second-growth conifer forests (Welber et al. 2013; Kramer and Wohl 2017). The reaches of the Middle Hoh Study Area generally represent high quality salmon habitat, with abundant braid and side-channel habitat, a large number of pools and slow water glides, and frequent large wood (17.6 jams/mile). Pool frequencies were below reported targets for both main channels (over 100 feet in wetted width), for which the National Marine Fisheries Service (NMFS) considers 18 pools per mile to be adequate, and braids and side channels (wetted width of 25-50 feet), for which NMFS has a target of 47 pools per mile (NMFS 1996; Table 8), however, there was a large amount of slow water edge habitat observed, both with wood cover and without, and a large number of slow water glide habitats that would provide rearing and holding habitat for juvenile and adult migrating salmon, respectively. The Huelsdonk-South Fork Reach, Morgans Crossing Reach, and Willoughby Creek Reach contained the most braid and side-channel habitat, though the confinement of the Spruce Canyon Reach and Oxbow Canyon Reach was driven by natural canyon features and not infrastructure encroachment. Hydro-modified banks were infrequent throughout the study area as a whole. However, the Upper Hoh Road flanks the right bank of the mainstem and is directly adjacent to the river in the Willoughby Creek Reach, Morgans Crossing Reach, and Huelsdonk-South Fork Reach. Rocky bank armoring and rip rap was documented along the road without vegetative cover and presented an opportunity for enhancement and roughening to improve the habitat. Additionally, the mainstem is eroding multiple residential properties in the Huelsdonk-South Fork Reach, which could be candidates for bank roughening and planting to provide short-term bank protection and reduce erosion of fine sediment. Substrate was visually characterized throughout the study area and was largely gravel and cobble dominated, and numerous high quality spawning habitats were present. However, substrate was not characterized at the unit level and spawning habitats were not mapped as part of this survey. Several landslides and eroding banks were observed throughout the reaches, which could lead to fine sediment pulses during heavy rainfall events (Hoh Tribe 2016). However, fines were not visible as the dominant sediment in any of the reaches at the time of survey. Restoration efforts focused on enhancement of hydro-modified banks, establishing coniferous vegetation in the floodplain, and protecting existing coniferous forests would best benefit salmon habitat and the long-term resiliency of the river in order to help stabilize the channel, reduce erosion and fine sediment deposition, and to provide sources of pool forming large wood (Fausch and Northcote 1992; Beechie and Sibley 1997). These efforts would be especially beneficial in the Willoughby Creek, Morgans Crossing, and Huelsdonk South Fork reaches. Table 9. Overview of mainstem habitats surveyed in the Middle Hoh River Study Area by reach. The count of observed units is n; length (feet) and area (square feet) are totals for the channels. Edge length (feet) and slow water area (square feet) are totals for all edges. Braid and side-channel habitats were surveyed as time allowed and were not a full census of habitats. Na indicates the habitat did not exist; ns indicates the habitat did exist but was not surveyed due to time constraints. REACH CHANNEL TYPE UNIT TYPE TOTAL BAR EDGE HYDRO-MODIFIED BANK EDGE NATURAL BANK EDGE EDGES     N LENGTH AREA EDGE LENGTH SLOW WATER AREA EDGE LENGTH SLOW WATER AREA EDGE LENGTH SLOW WATER AREA EDGE LENGTH SLOW WATER AREA  Oxbow Canyon Main Channel Glide 5 5,023 733,819 2,395 7,858 0 0 7,651 31,990 10,046 39,848    Pool 8 3,271 420,589 2,500 11,582 0 0 4,042 16,641 6,542 28,223    Rapid 1 574 170,473 172 565 0 0 976 3,202 1,148 3,767    Riffle 4 1,099 179,703 1,099 4,575 0 0 1,099 3,606 2,198 8,181    Total 18 9,967 1,504,585 6,166 24,579 0 0 13,768 55,440 19,934 80,019   Braid Glide na na na na na na na na na na na    Pool na na na na na na na na na na na    Rapid na na na na na na na na na na na    Riffle na na na na na na na na na na na    Total na na na na na na na na na na na   Side Channel Glide na na na na na na na na na na na    Pool na na na na na na na na na na na    Rapid na na na na na na na na na na na    Riffle na na na na na na na na na na na    Total na na na na na na na na na na na   Total 18 9,967 1,504,585 6,166 24,579 0 0 13,768 55,440 19,934 80,019  Willoughby Creek Main Channel Glide 7 9,150 1,773,440 11,929 39,137 1,944 6,378 4,428 14,527 18,301 60,041    Pool 5 2,287 284,765 1,827 5,995 0 0 2,746 9,009 4,573 15,005    Rapid 1 328 26,910 328 1,076 0 0 328 1,076 656 2,153    Riffle 7 5,531 973,176 6,347 27,640 0 0 4,716 15,472 11,063 43,112    Total 20 17,297 3,058,291 20,431 73,848 1,944 6,378 12,218 40,085 34,593 120,310   Braid Glide 13 3,784 269,695 4,672 44,060 0 0 2,896 38,235 7,568 82,294    Pool 20 3,281 140,569 4,281 14,044 356 1,167 1,925 6,317 6,562 21,528    Riffle 26 3,192 161,506 5,098 16,726 476 1,563 809 2,846 6,384 21,135    Total 59 10,257 571,770 14,050 74,830 832 2,730 5,631 47,398 20,513 124,957   Side Channel Glide 3 843 68,229 912 2,992 774 2,540 0 0 1,686 5,533    Pool 3 1,020 87,661 1,020 3,348 568 1,862 453 1,485 2,041 6,695    Riffle 2 709 60,797 1,417 4,737 0 0 0 0 1,417 4,737    Total 8 2,572 216,687 3,350 11,077 1,342 4,402 453 1,485 5,144 16,964   Total 87 30,125 3,846,748 37,831 159,754 4,118 13,510 18,302 88,968 60,251 262,232  Morgans Crossing Main Channel Glide 12 14,403 2,839,300 15,001 65,495 1,181 4,650 12,624 42,491 28,806 112,636    Pool 5 2,746 365,961 2,746 17,082 787 2,583 1,959 6,426 5,492 26,092    Riffle 13 7,427 1,486,321 9,232 49,071 0 0 5,621 22,734 14,853 71,805    Backwater 2 787 34,914 787 2,583 0 0 787 2,583 1,575 5,167    Total 32 25,363 4,726,496 27,767 134,232 1,969 7,233 20,991 74,234 50,726 215,700   Braid Glide 20 3,946 233,661 6,577 27,128 0 0 1,314 4,312 7,891 31,439    Pool 19 2,276 130,114 3,059 10,036 0 0 1,492 4,896 4,551 14,932    Riffle 26 3,168 143,422 5,385 19,733 0 0 952 3,834 6,337 23,567    Total 65 9,389 507,197 15,020 56,896 0 0 3,759 13,042 18,779 69,938   Side Channel Glide 5 1,457 79,232 1,511 4,958 0 0 1,404 4,605 2,915 9,563    Pool 12 2,733 210,829 2,358 8,515 0 0 3,107 10,193 5,465 18,707    Riffle 13 1,423 62,827 1,445 4,741 0 0 1,401 4,596 2,846 9,337    Backwater 1 66 850 66 215 0 0 66 215 131 431    Total 31 5,678 353,739 5,380 18,429 0 0 5,977 19,609 11,357 38,037   Total 128 40,431 5,587,431 48,167 209,557 1,969 7,233 30,726 106,885 80,862 323,675  Spruce Canyon Main Channel Glide 2 2,014 424,981 1,260 4,133 0 0 2,769 14,122 4,029 18,256    Pool 4 3,268 487,390 102 1,001 377 1,238 6,056 25,876 6,535 28,115    Rapid 2 1,883 254,674 1,693 5,553 0 0 2,074 8,266 3,766 13,819    Riffle 2 1,886 298,833 492 3,229 0 0 3,281 20,828 3,773 24,057    Total 10 9,052 1,465,878 3,546 13,917 377 1,238 14,180 69,092 18,104 84,247   Braid Glide 0 0 0 0 0 0 0 0 0 0 0    Pool 0 0 0 0 0 0 0 0 0 0 0    Rapid 1 472 48,050 472 3,100 0 0 472 1,550 945 4,650    Riffle 0 0 0 0 0 0 0 0 0 0 0    Total 1 472 48,050 472 3,100 0 0 472 1,550 945 4,650   Side Channel Glide ns ns ns ns ns ns ns ns ns ns ns    Pool ns ns ns ns ns ns ns ns ns ns ns    Rapid ns ns ns ns ns ns ns ns ns ns ns    Riffle ns ns ns ns ns ns ns ns ns ns ns    Total ns ns ns ns ns ns ns ns ns ns ns   Total 11 9,524 1,513,928 4,019 17,017 377 1,238 14,653 70,642 19,049 88,897  Huelsdonk-South Fork Main Channel Glide 10 1,545,778 471,153 8,192 46,036 741 2,433 7,707 35,033 16,640 83,502    Pool 13 1,084,475 330,548 5,924 26,750 1,182 3,877 5,748 35,806 12,854 66,434    Riffle 13 956,417 291,516 7,458 31,685 1,733 5,684 4,812 17,660 14,003 55,029    Total 36 3,586,670 1,093,217 21,574 104,471 3,656 11,994 18,267 88,500 43,497 204,965   Braid Glide 36 251,175 76,558 9,801 42,468 151 495 3,723 15,454 13,675 58,417    Pool 44 241,536 73,620 6,353 39,715 0 0 4,033 13,982 10,386 53,698    Riffle 71 431,849 131,628 16,640 59,936 0 0 3,377 13,239 20,017 73,175    Backwater 2 4,821 1,469 433 1,421 0 0 112 366 545 1,787    Total 153 929,381 283,275 33,227 143,540 151 495 11,245 43,041 44,623 187,076   Side Channel Glide 34 276,367 84,237 5,447 31,108 115 377 7,640 36,303 13,201 67,789    Pool 40 325,878 99,328 4,800 19,928 276 2,809 6,974 25,399 12,049 48,136    Riffle 35 277,716 84,648 5,387 22,450 0 0 4,781 16,744 10,168 39,194    Backwater 3 1,209 369 0 0 0 0 313 1,027 313 1,027   Table 10. Pool counts, reach width (feet), total pool area (square feet), the percent of area in pools, pool frequency (pools per mile), and pool spacing (channel widths per pool) (Montgomery et al. 1995) for all channel types and reaches for the Hoh River mainstem. Braid and side-channel habitats were surveyed as time allowed and were not a full census of habitats. Na indicates the habitat did not exist; ns indicates the habitat did exist but was not surveyed. REACH CHANNEL TYPE TOTAL POOLS REACH WIDTH POOL AREA % POOL AREA POOLS/ MILE CHANNEL WIDTHS/ POOL  Oxbow Canyon Main Channel 8 147 420,589 28 4.2 8.5   Braid na na na na na na   Side Channel na na na na na na   Total 8 147 128,196 28 4.2 8.5  Willoughby Creek Main Channel 5 159 284,765 9 1.5 21.8   Braid 20 41 140,569 25 10.3 12.5   Side Channel 3 77 87,661 40 6.2 11.2   Total 28 71 512,995 13 4.9 15.1  Morgans Crossing Main Channel 5 167 365,961 8 1 30.4   Braid 19 43 130,114 26 10.7 11.5   Side Channel 12 49 210,829 60 11.2 9.6   Total 36 75 706,904 13 4.7 14.9  Spruce Canyon Main Channel 4 156 487,390 33 2.3 14.5   Braid na 102 na na na na   Side Channel na - na na na na   Total 4 151 487,390 32 2.2 15.8  Huelsdonk-South Fork Main Channel 13 159 1,084,475 30 3.2 10.5   Braid 44 33 241,536 26 10.4 15.4   Side Channel 40 39 325,878 37 11.8 11.5   Total 97 50 1,651,889 31 8.3 12.7  Grand Total 173 65 3,779,766 21 6.0 13.6   Table 11. Counts (n) and total areas (square feet) large wood jams (LWJ) located in main channel and braids, side channels, and islands, on banks, bars, or islands, and wet or dry. Jams were delineated using aerial imagery collected in March 2021. REACH CHANNEL BANK BAR ISLAND    DRY WETTED TOTAL DRY WETTED TOTAL DRY    N AREA N AREA N AREA N AREA N AREA N AREA N AREA  Oxbow Canyon Main channel 0 0 2 4,541 2 4,541 0 0 0 0 0 0 0 0   Total 0 0 2 4,541 2 4,541 0 0 0 0 0 0 0 0  Willoughby Creek Main channel 4 20,431 9 19,249 13 39,680 61 168,310 36 208,936 97 377,247 0 0   Side channel 1 1,319 2 6,566 3 7,884 0 0 0 0 0 0 0 0   Island 0 0 0 0 0 0 0 0 0 0 0 0 1 1,995   Total 5 21,750 11 25,814 16 47,564 61 168,310 36 208,936 97 377,247 1 1,995  Morgans Crossing Main channel 15 60,096 15 123,304 30 183,400 70 279,962 32 292,139 102 572,101 0 0   Side channel 0 0 1 1,683 1 1,683 9 23,415 9 38,972 18 62,387 0 0   Island 0 0 0 0 0 0 0 0 0 0 0 0 3 11,167   Total 15 60,096 16 124,987 31 185,083 79 303,377 41 331,111 120 634,488 3 11,167  Spruce Canyon Main channel 2 7,524 3 78,130 5 85,654 5 12,577 1 6,194 6 18,771 0 0   Side channel 0 0 3 30,422 3 30,422 0 0 1 6,546 1 6,546 0 0   Total 2 7,524 6 108,552 8 116,076 5 12,577 2 12,740 7 25,317 0 0   Huelsdonk-South Fork Main channel 7 19,599 19 280,629 26 300,228 93 353,388 50 322,393 143 675,781 0 0   Side channel 6 16,306 15 81,220 21 97,526 17 51,667 5 20,019 22 71,686 0 0   Island 0 0 0 0 0 0 0 0 0 0 0 0 11 24,938   Total 13 35,905 34 361,849 47 397,754 110 405,055 55 342,412 165 747,468 11 24,938  Total 35 125,275 69 625,742 104 751,018 255 889,319 134 895,200 389 1,784,519 15 38,100   Table 12. Counts of large wood jams (n), frequency (LWJ count/mile), total LWJ area (square feet), and mean and median LWJ size (area in square feet; 50th, and 90th percentiles, Jam50, Jam90). REACH WET/DRY N LW FREQUENCY TOTAL AREA (FT2) MEAN AREA (FT2) JAM50 JAM90  Oxbow Canyon Dry 0 0.0 0 0 0 0   Wet 2 1.1 4,541 2,270 2,270 2,648   Total 2 1.1 4,541 2,270 2,270 2,648  Willoughby Creek Dry 67 11.7 192,055 2,866 1,995 5,248   Wet 47 8.2 234,751 4,995 2,648 12,129   Total 114 20.0 426,805 3,744 2,283 7,539  Morgans Crossing Dry 97 12.7 374,641 3,862 2,647 7,467   Wet 57 7.4 456,098 8,002 5,067 19,639   Total 154 20.1 830,739 5,394 3,135 10,703  Spruce Canyon Dry 7 3.9 20,101 2,872 1,998 4,921   Wet 8 4.4 121,292 15,161 7,922 29,950   Total 15 8.3 141,393 9,426 5,527 15,219  Huelsdonk-South Fork Dry 134 11.4 465,898 3,477 2,641 5,469   Wet 89 7.6 704,261 7,913 3,620 21,724   Total 223 19.0 1,170,159 5,247 2,975 10,347  Hoh MS Study Area Dry 305 10.6 1,052,695 3,451 2,483 6,071   Wet 203 7.1 1,520,942 7,492 3,951 18,876   Total 508 17.6 2,573,637 5,066 2,813 10,202   / Figure 17. The percent of wetted habitat area by habitat unit, channel type, and reach. Braid and side-channel habitats were surveyed as time allowed and were not a full census of habitats, ns indicates the habitat type was present but not surveyed. No braid or side-channel habitat was wetted at the time of survey in the Oxbow Canyon Reach. / Figure 18. Edge length (feet) by edge type, channel type, and reach. Braid and side-channel habitats were surveyed as time allowed and were not a full census of habitats, ns indicates the habitat type was present but not surveyed. No braid or side-channel habitat was wetted at the time of survey in the Oxbow Canyon Reach. / Figure 19. Average slow water edge area (feet) by edge type, and reach. Braid and side-channel habitats were surveyed as time allowed and were not a full census of habitats, ns indicates the habitat type was present but not surveyed. No braid or side-channel habitat was wetted at the time of survey in the Oxbow Canyon Reach. / Figure 20. Pool counts by pool type, channel type, and reach. Braid and side-channel habitats were surveyed as time allowed and were not a full census of habitats, ns indicates the habitat type was present but not surveyed. No braid or side-channel habitat was wetted at the time of survey in the Oxbow Canyon Reach. / Figure 21. Pool counts by pool forming feature, channel type, and reach. Channel feature includes meanders and confluences. Braid and side-channel habitats were surveyed as time allowed and were not a full census of habitats, ns indicates the habitat type was present but not surveyed. No braid or side-channel habitat was wetted at the time of survey in the Oxbow Canyon Reach. / Figure 22. Mean pool frequency (pools per mile) versus wood frequency (LWJ count per mile) of wetted jams pool for reaches in the Hoh River study area with a 95% confidence interval shown. The relationship was not significant, p-value=0.40. Oxbow Canyon Reach The Oxbow Canyon Reach is the downstream extent of the study area and runs from to the Oxbow Creek Campground near RM 16 to downstream of RM 17.5 at the confluence of Winfield Creek and the mainstem Hoh River (Map 7). The reach is heavily confined from a natural canyon that prevents lateral channel migration. No braided channels were wetted at the time of survey and no side channels were present in this reach (Table 9). Glides were the dominant habitat unit in this reach, and accounted for 49% of the wetted channel area, followed by pools, which accounted for 28% of the wetted channel area (Figure 17). One rapid and four riffles were also observed, accounting for the remaining 23% of surveyed wetted channel area. Natural bank edges made up 69% of the edge length in this reach, while bar edges made up the remaining 31%. No hydro-modified bank edges were present in this reach (Figure 18). The average slow water edge width for both natural bank edges and bar edges was 4.1 feet (Figure 19). The Oxbow Canyon Reach had the lowest main channel pool spacing (8.5 channel widths per pool) and highest main channel pool frequency (4.2 pools per mile) of the reaches surveyed, with eight pools observed in the 1.9 miles of channel length (Table 10). There were seven trench pools and one scour pool documented in this reach, all of which were formed by bedrock features (Figure 20; Figure 21). Wood was infrequent in this reach, with only two wetted jams and no dry jams mapped (Table 11). Large wood jams were relatively small in this reach, with an average jam size of 2,270 square feet (Table 12). However, given the confinement and bedrock present in this reach, opportunities for natural wood racking and accumulation are minimal. Willoughby Creek Reach The Willoughby Creek Reach extends from RM 17.5 at the Winfield Creek confluence with the mainstem up to upstream of Willoughby Creek near RM 20.5 (Map 7). The Upper Hoh Road flanks the right bank of the main channel at the upper extent of the reach. The Willoughby Creek Reach is a relatively unconfined sinuous gravel- cobble reach. Near the middle of the reach, the main channel demonstrates substantial migration, and has shifted over the last 30 years away from a large meander bend on the left bank to a more linear channel form. The lower extent of the reach is more naturally confined and exhibits less migration. There were 3.3 miles of main channels, 1.9 miles of braided channel, and 0.5 miles of side-channel habitat surveyed in this reach. Additional braid and side-channel habitats were wetted at the time of survey but were not surveyed due to time constraints. Glides were the dominant main channel unit in in this reach and accounted for 58% of the wetted channel area (Figure 18; Table 9). Riffles and rapids made up 32% and 1% of the wetted main channel area, respectively. Pools were infrequent in the main channel and only accounted for 9% of the total wetted main channel area. Pools were more frequent in braid and side-channel habitats and made up 25% and 40% of wetted channel area, respectively (Figure 18). Glides made up 47% and 31% of braid and side-channel wetted habitat area, respectively, and riffles made up the remaining 28% of channel area for both braids and side channels. Mainstem edges were primarily composed of bar edges, which made up 63% of the surveyed edge length. Bank edges made up the other 37% of main channel length, of which 18% were hydro-modified (Table 9). The Willoughby Creek Reach had the second highest total length of hydro-modified bank edges after the Huelsdonk-South Fork Reach of the reaches surveyed (Figure 18). Two areas on the river right bank were noted to have substantial erosion and trees actively sloughing into the channel. Most of the slow water edge area observed was associated with bar edges, however, slow water edge width was the greatest in natural bank edges, with an average slow water edge width of 4.1 feet, followed by bar edges, with an average edge width of 3.6 feet (Table 9). Hydro-modified bank edges had an average slow water edge area of 3.3 feet (Figure 19). Total percent pool area and pool frequency were low and pool spacing was high in this reach compared to other reaches, with only 13% pool area, 1.5 pools per mile, and 21.8 channel widths per pool (Table 10). Twenty-eight pools were documented in the 5.7 miles of channel length surveyed in this reach. Large wood jams and pieces were the primary pool forming feature, followed by channel features, including meanders and confluences (Figure 21). The Willoughby Creek Reach contained the highest LW frequency for wet jams and second highest overall. However, the median jam size was smaller than other reaches at 2,685 square feet, and the reach had the third lowest total jam area (Table 12). The majority of jams were located in the main channel on gravel bars (97 jams comprising 377,247 square feet), while only 16 bank-attached jams were mapped totaling 47,564 square feet of area (Table 11). Jams in this reach were smaller than the upstream Huelsdonk-South Fork and Morgans Crossing reaches, which suggests the larger old growth conifers entering the system from the ONP are being deposited in the upstream reaches and the wood inputs to the Willoughby Reach are primarily from the deciduous and second-growth conifer trees located in the Middle Hoh Study Area riparian (Welber et al. 2013; Kramer and Wohl 2017). Pool frequency and total pool area was low compared to other reaches and to NMFS targets (NMFS 1996; Table 8). However, slow water estimates that include glide habitat and slow water edge areas suggest that there is abundant habitat in this reach for juvenile rearing and adult holding for coho, steelhead, and spring and fall Chinook (Table 9). Additionally, there were numerous off channel habitats for rearing coho and steelhead and spawning coho, steelhead, and Chinook in this reach. Restoration efforts in this reach should target the 0.8 miles of road that runs along the right bank of the mainstem, which would benefit from enhancement actions such as raising the road prism and protecting the bank toe with large wood structures to create a more natural hillslope form and increase bank attached wood jam area to create slow water main channel habitat and provide shading and predator protection. Additional actions aimed at protecting riparian forests with the long-term goal of establishing forests similar to those in the ONP and planting meander bends with conifers to increase roughness would also benefit this reach. Morgans Crossing Reach The Morgans Crossing Reach extends from RM 20.5 upstream of the Willoughby Creek confluence with the mainstem to RM 25.5 (Map 7). The Upper Hoh Road runs adjacent to this reach and confines the floodplain on the right bank. The road runs directly adjacent to the main channel right bank at the downstream and upstream end of this reach and flanks the right bank of a side channel in the middle of the reach. A substantial amount of channel migration has occurred in the middle of this reach, with the channel straightening and moving away from the road on the right bank. Most of the sinuosity in this reach occurs in the middle extent, with minimal braid and side-channel habitat main channel meandering occurring in the upstream and downstream end of the reach. The Morgans Crossing Reach was the longest reach surveyed. A total of 7.7 miles of wetted mainstem channel length was mapped in the reach, including 4.8 miles of main channel, 1.8 miles of braided channel, and 1.1 miles of side-channel length (Table 9). Additional braids and side channels were wetted at the time of survey but were not surveyed due to time constraints (Map 7). Glides accounted for approximately 60% of the total wetted channel area, followed by riffles which made up 31%, pools which made up 9%, and backwater units that made up 1% (Figure 17; Table 9). This reach had the lowest main channel percent pool area, and second lowest total pool channel area. Pools accounted for more wetted channel area in braids and side channels than in main channels in this reach. In braids, glides accounted for 46% of wetted channel area, riffles accounted for 28% of wetted channel area, and pools accounted for 26% of wetted channel area. In side channels, pools accounted for 60% of the wetted channel area, followed by glides which made up 22% of wetted channel area, and riffles which made up 18% of wetted channel area (Figure 17). Approximately 60% of the total edge length was bar edges, 38% was natural bank edges, and 2% of banks were hydro-modified (Figure 18). Bank hydro-modifications, including rip rap and residential development, were only documented in main channels in this reach, however, a road also runs along an un-surveyed side channel and confines lateral movement on the right bank (Figure 18; Map 7). The average slow water edge width 4.1 feet for bar edges, 3.9 feet for hydro modified bank edges, and 3.6 feet for natural bank edges (Figure 19). The Morgans Crossing Reach had the second highest total mainstem (main channels, braids, and side channels combined) pool area of the reaches surveyed, after the Huelsdonk-South Fork Reach (Table 10). While the percent pool area and pool frequency (pools per mile) in the main channel were the lowest of the reaches surveyed, in braids and side channels, percent pool area and pool frequency were amongst the highest of the reaches (Figure 17). Main channel pools were primarily formed by channel features, while large wood primarily drove pool formation in braids (Figure 21). Side-channel pools were formed by both large wood and channel features. The Morgans Crossing Reach had the highest overall (wet and dry) LWJ frequency and second highest total LWJ area (Table 12). Main channel and side channel LWJs were abundant in this reach (Table 11). There were 120 jams located on gravel bars, comprising 634,488 square feet of total jam area, and 31 jams located attached to banks, comprising 184,083 square feet of jam area (Table 11). This reach had the second largest mean jam size of the reaches surveyed (Table 12). Total pool area was low in the main channel of Morgans Crossing Reach but there was abundant slow water habitat in side channels for juvenile rearing and for coho, steelhead, and spring and fall Chinook spawning. The Upper Hoh Road runs directly adjacent to this reach in multiple locations which presents an opportunity to enhance bank habitat. Bank enhancement actions such as raising the road prism, planting, and placing large wood benefit salmon, by roughening banks and providing cover, and help to protect infrastructure. Additionally, three locations along the right bank were noted to have severe erosion and landslide risk and may need further evaluation or stabilization efforts. Floodplain plantings of conifers in the larger alder groves on older bars and meander bends could speed up riparian succession and provide bank stabilization. This reach would also benefit from continued floodplain forest protection to encourage the establishment of mature conifer forests, similar to those present in the ONP, to provide long-term large wood inputs to create large functional jams and stabilize the channel. Spruce Canyon Reach The Spruce Canyon Reach runs from approximately RM 25.5 to RM 27 and is the shortest reach in the Middle Hoh Study Area (Map 7). The left bank is confined by the valley wall, and the right bank is confined by intermittent bedrock and boulders. This reach experiences little channel migration, and the channel is largely confined to the main channel with little braid or side-channel habitat present, apart from one side channel in the upper extent. We surveyed 1.7 miles of main channel and 0.1 miles of braided channel length in the Spruce Canyon Reach (Table 9). One additional small side-channel complex was wetted at the time of survey but was not surveyed due to time constraints (Map 7). Pools accounted for approximately 33% of the surveyed wetted channel area, followed by glides, riffles, and rapids, which accounted for 29%, 20% and 17% of the surveyed wetted channel area, respectively (Figure 17). One rapid spanned 472 feet and made up all of the surveyed braided channel. Natural bank edges accounted for 78% of the total edge length surveyed in this reach and bar edges made up 20% (Figure 18). One section of rip rap was present on river right that made up the other 2% of the edge length. Natural bank edges had an average slow water edge width of 5.4 feet, followed by bar edges which had an average slow water width of 5.2 feet (Figure 19). The rip rap bank section had a slow water edge width of 3.3 feet. Total LWJ area and frequency were low in this reach, with only 15 jams mapped, however, the mean and median wetted jam size in this reach was the highest of the reaches surveyed (Table 12). This was caused by the presence of six large bank-attached wetted LWJs at upstream end of the reach, upstream of the canyon. Minimal gravel bar habitat was present downstream of other natural LWJ accumulation (Table 11). Four scour pools were observed in the Spruce Canyon Reach, including three pools formed by channel features and one formed by bedrock (Figure 20; Figure 21; Table 10). The main channel of the Spruce Canyon Reach had the highest percent pool area, though pool frequency and pool spacing were lower and higher than other reaches, respectively (Figure 17). This reach is largely confined by natural features and presents little opportunity for restoration. Given the presence of bedrock throughout the reach, there is not much of an opportunity for wood recruitment, however, management actions to protect the floodplain forest in the upper extent of the reach, upstream of the canyon to establish large conifers would increase long-term large wood inputs. Huelsdonk-South Fork Reach The Huelsdonk-South Fork Reach extends from RM 27 to RM 30.5 below the confluence of the South Fork Hoh River and National Park Boundary (Map 7). The Huelsdonk-South Fork Reach is a meandering and unconfined reach with abundant braid and side-channel habitat. The Upper Hoh Road runs along the main channel Hoh in two places in this reach on right bank. Multiple residential properties exist on the left bank and various armoring techniques were present that attempt to confine lateral movement and reduce erosion. The reach has experienced substantial lateral channel movement in the recent years and the mainstem migration zone has widened considerably. We surveyed 4.1 miles of main channel, 4.2 miles of braided channel, and 3.4 miles of side channels in the Huelsdonk-South Fork Reach (Table 9). All wetted braided channels connected to the main channel at the time of survey were surveyed, however, one large side-channel complex was wetted at the time of survey but was not surveyed due to time constraints (Map 7). In main channels of the mainstem, glides were the predominant habitat type and accounted for 43% of the total wetted habitat area, followed by pools and riffles which accounted for 30% and 27%, respectively (Figure 17). In braided channels, riffles accounted for 46% of the wetted channel area, followed by glides, pools, and backwaters, which made up 27%, 26%, and 1% of the wetted habitat area, respectively. In side channels, pools accounted for 37% of the total wetted channel area, followed by riffles and glides, which accounted for the other 32% and 31% of the wetted channel area, respectively. Edge habitat in the Huelsdonk-South Fork Reach was primarily made up of natural banks and bar edges, however, the Huelsdonk-South Fork Reach had the most hydro-modified edge length of the reaches surveyed (Figure 18; Table 9). Right bank modifications included rip rap along the Upper Hoh Road, and left bank modifications included bank armoring to protect residential property and an active erosion zone along the Huelsdonk Property. Hydro-modified banks mostly occurred along the main channel of the mainstem. The average slow water edge width associated with bar edges was 4.3 feet (Figure 19). Natural bank edges and hydro-modified bank edges had average slow water edge widths of 4.1 and 3.9 feet, respectively. The Huelsdonk-South Fork Reach had the most total pool area (325,878 square feet) of the reaches surveyed (Table 10). The total percent main channel pool area in this reach was the highest of the reaches surveyed, however, main channel pool frequency was low with only 3.2 pools per mile (Figure 18; Table 10). Pools were abundant in braids and side channels; 44 and 40 pools were documented in braided channels and side channels, respectively. Channel features, including meanders and confluences, were main pool forming features in main channel pools, but large wood jams and pieces drove pool formation in braids and side channels (Figure 21). There were 223 LWJs mapped in the Huelsdonk-South Fork Reach, accounting for 1,170,159 square feet of jam area, which was the largest total area of the mapped reaches (Table 12). The reach also contained the largest amount side channel and island total LWJ area (Table 11). The Huelsdonk-South Fork Reach contained the largest jam mapped in the study area; however, the median jam size was smaller than the Spruce Canyon and Morgans Crossing reaches (Table 12). The large total number and area of jams in this reach is likely due to the proximity to the ONP, which provides a source of large old growth conifers to the river, as well as the abundant gravel bar and roughened bank habitat present to rack large wood accumulations (Table 11). While pool frequency was low in this reach, pool area was high, and the abundance of braid and side-channel habitat provides abundant off-channel slow water habitat for juvenile salmon rearing and refuge. Cobbles and gravels were the dominant substrate throughout the reach, which suggests adequate spawning habitat for coho, Chinook, and steelhead. The Huelsdonk-South Fork Reach contains the most private property in the channel migration zone of the reaches surveyed and had multiple zones of erosion that are concerning both for their impacts on private property and as sources of sediment and infrastructure debris into the river. Restoration actions, such as raising the road prism, planting, and placing large wood, aimed at enhancing the 0.4 miles of the Upper Hoh Road, which runs adjacent to the mainstem of this reach in two locations, would provide cover and create slow water habitat for salmon as well as protect infrastructure. Additionally, floodplain fencing and other bank stabilization methods, including planting, could serve to prevent further erosion in the short-term. Numerous alder forested islands and meander bends were present in this reach, which present an opportunity for conifer planting to speed up riparian succession and establish stable riparian forests. Lastly, this reach appears to have the most floodplain potential and relic side channels could be identified for reoccupation to reduce pressure on priority banks, however, further assessment would be needed to assess the feasibility of this restoration action. Due to the more moderated flows in side channels they have more stable substrate. They also tend to have a higher frequency of pools and habitat diversity. For these reasons side channels can provide preferred spawning in large river valleys. Since they are also associated with floodplain wetlands and slower velocities during peak flows, side channels are crucial rearing habitat. Logjams can be the dominate mechanism in forming and sustaining side channels (Abbe and Montgomery 1996, 2003). The presence of large trees in riparian/floodplain forests is critical in the formation of logjams (Abbe and Montgomery 1996, Collins et al. 2012). These linkages demonstrate that critical salmonid habitat (side channels) in large alluvial rivers is closely linked to riparian forests and wood Transportation A network of roads traverse the Middle Hoh River valley, ranging from major roads like the Upper Hoh Road running along the northern side of the river, to unimproved and decommissioned logging roads up along the hillsides. This road system brings regional and international tourists to ONP and local businesses, provides access for local landowners and recreational users and is used for logging and mining operations. The development of this road network is the result of cumulative additions over decades, some of which did not consider the many hazards and natural processes present in the Middle Hoh. When considering the future resiliency of the Middle Hoh one must recognize and integrate the road network into future planning efforts. The current alignment of several road segments have chronic maintenance issues and impair natural processes and habitat formation and creation, limiting the resiliency of the transportation network that is relied upon while degrading habitat conditions. The purpose of this assessment is to inventory segments of roads that are in conflict with providing and improving resiliency in the project reach, and to identify opportunities for redesign that would improve resiliency that are consistent with the local community needs and desires. Road segments were identified within the resiliency corridor, delineated CMZ and the FEMA 100-yr floodplain in GIS and are summarized in Table 13. Table 13. Identified road segments within the Middle Hoh CMZ, Resiliency Corridor and FEMA 100-yr floodplain. ROAD SEGMENTS LENGTH IN CMZ (MILES) LENGTH IN RESTORATION OCRRIDOR (MILES) LENGTH IN FEMA SFHA (MI)  Unnamed 7.80 3.54 2.19  Upper Hoh Road and spurs 8.58 5.62 2.46  Old Milwaukee Rd 3.53 2.19 1.95  FR-H-1088 and spurs 1.33 0.04 -  FR-H-1009 and spurs 1.91 0.65 0.51  FR-H-1011 and spur 0.57 - -  Lewis Ponds and spur 0.58 0.14 0.09  FR-H-1062 and spurs 0.11 - -  FR-H-3700 0.06 0.00 -  FR-H-1060 0.04 - -  FR-H-3900 0.02 - -   Road Before Allen's Cutoff 0.21 0.21 -  Pit Road 0.01 - -  Hoh Oxbow Boat Launch 0.02 0.02 0.02  Minnie Peterson Campground 0.14 0.14 0.14  Grand Total 24.92 12.56 7.36   These summary tables indicate there are a number of road segments that are currently at risk from flooding and/or erosion and that limit natural processes on the Middle Hoh. Strategies for addressing these risks and their impact on the environment should consider the location of the segment relative to the resiliency corridor; segments within the resiliency corridor should be prioritized for relocation as they pose the greatest risk to both the human and ecologic communities. Segments within the CMZ and/or FEMA 100-yr floodplain that are outside of the resiliency corridor could consider other alternatives to limit risk to the road, however planned relocation to a safer position on the landscape provides the greatest resiliency as local protection typically requires maintenance and can fail. Opportunities for improving resiliency within the Middle Hoh through road redesign and relocation are limited as much of the valley bottom is within the CMZ and/or FEMA 100-yr floodplain and the surrounding hillsides are naturally unstable (Map 3). As such, many alternatives need to be considered and assessed in more detail to identify preferred alignments. During meetings with the Middle Hoh steering committee and local landowners several ideas have been brought forward that illustrate the complexities and “out-of-the-box” thinking needed to consider current and future risks, local and visitor access and safety, all while restoring, preserving and protecting ecologic function. Some examples include: Move visitor access to the Middle Hoh to the Hoh Mainline on the south side of the valley, crossing the river at Spruce Canyon where the CMZ narrows with a new bridge. Once over the new bridge on the north side of the river the route climbs out of the valley bottom to mid-valley slope elevations, heading east until gradually descending back to the valley floor and joining the Upper Hoh Road east of Canyon Creek. This alternative would improve existing roads where possible to create the new alignment, allow decommissioning segments of road that are within the CMZ and/or FEMA 100-yr floodplain that have been bypassed by the new road, and provide an alternative route in and out of the valley for emergency purposes. The required bridging of the Hoh for this option would detract from ecologic resiliency as the span would need to be within the resiliency corridor, there would be impacts to upland ecosystems establishing and improving the new alignment, and the cost would be very high. Relocate those segments of the Upper Hoh Road that are within the CMZ and/or FEMA 100-yr floodplain away from the river, out of the valley bottom where necessary. New road alignments traverse terrain with a low probability for slope failure if they ascend up the valley hillsides, subsequently descending to rejoin the Upper Hoh Road once the relocated segment has been bypassed. Establish alternative transportation methods for visitors to access the Middle Hoh and ONP. Creation of a new parking area near highway 101 and creation of a shuttle bus service taking passengers on the Upper Hoh Road would reduce traffic and improve safety. Building recreational trails along the river corridor and/or through the hillsides leading up to ONP, using a similar parking area near highway 101 would serve a similar purpose. A bike lane could be added to the Upper Hoh Road providing yet another alternative method to access the Middle Hoh and ONP. As new technologies come online in the coming decades, additional alternatives for moving visitors up the Middle Hoh valley will be realized that should be considered as well. Any transportation method that does not rely on contact with or disturbance of the ground surface would provide the least impact and provide the greatest resiliency, if not cost prohibitive. Low-altitude aircraft requiring minimal infrastructure of take-off and landing and tunnels are examples of technologies that are currently not developed sufficient to be cost effective but may be in the future and should remain under consideration as implementation timelines allow. A more complete evaluation and assessment of alternatives to improve transportation resiliency is needed to identify a preferred solution. Alternatives need to incorporate the needs and desires of the local community, account for potential hazards associated with the alternative while improving ecologic and community resiliency in the Middle Hoh. The current national trend of National Parks moving toward reservation and timed entry systems for day use may change transportation needs in the future, potentially changing the number of visitors to the area and where they look to recreate. Future anticipated changes in precipitation patterns need to be included in planning as they will alter the frequency and locations of hazards posing risk to any alternative route under consideration. These include more extensive flooding and accelerated erosion adjacent to the river and more frequent slope failures on steep valley margins. Site specific investigations can resolve these concerns once a preferred alternative has been identified. TRENDS & ANTICIPATED CHANGES Broadly speaking the existing condition of the Middle Hoh River is a system that is largely intact, relative to the many others that have experienced significant encroachment and manipulation. The Middle Hoh also has the benefit of the Upper Hoh River being in the largely undisturbed and protected ONP. These factors have given the Middle Hoh River resiliency from the other disturbances that have impacted habitat conditions, primarily logging of the floodplain forests. Even with the loss of the old-growth forest within most of the Middle Hoh, large wood persists in the channel largely due to inputs from the ONP. Projecting future conditions based on the best available science and observed indicators suggests changes to the Middle Hoh are coming and should be planned for accordingly. Climate Change Scientific studies in the Pacific Northwest region have concluded that the frequency and magnitude of peak flows will increase over the next 100 years while summer/fall low flows will diminish (Hamlet et al. 2013; Mauger et al. 2016; Warner, Mass, and Salathé 2015; Hamlet and Lettenmaier 2007; Elsner et al. 2010, Isaak et al. 2017). Impacts from the warming climate on flooding are projected to occur in the coming decades, with increases in peak flows modeled for the 2040s (inclusive of 2030-2059; Hamlet et al. 2013), 2050s (inclusive of 2040-2069; Mauger et al. 2016), and beyond. Numerous studies for western Washington have indicated increases in peak flows as early as the 2020s (inclusive of 2010-2039; Elsner et al. 2010, Mantua, Tohver, and Hamlet 2010). The warming climate effects are therefore relevant to the consideration of all geomorphic and hydraulic processes related to flow duration, frequency, and magnitude. For the Hoh River watershed, changes in rainfall intensity, duration, and seasonality can be expected. High rainfall in the watershed currently occurs primarily during the fall and winter when Pacific cyclones cause prolonged, orographically enhanced precipitation. These storms can last for several days and are often the cause of flooding in the Pacific Northwest. Snowfall in the higher elevations can absorb some of this precipitation if it falls as snow or if the snowpack has capacity to store rainfall. If the associated precipitation rates are associated with rapid rises in freezing level associated with warm marine weather fronts from the central Pacific or rain falls in areas with less snowpack storage, the hydrologic volumes and delivery rates can compound flooding and influence channel responses. Looking to the future, the impact of the warming climate on river dynamics in the project area was analyzed by projecting peak flow estimates into the. The discharge estimates used for climate change floods are based on the work done by the University of Washington Climate Impacts Group project (Tohver et al. 2014) and are included in Table 14. Projections for increased peak flow magnitudes with the warming climate are driven by projections for more intense precipitation, particularly in the fall and winter months. Higher precipitation rates will increase the magnitude of annual peak and average winter flows and contribute to increases in mass-wasting and sediment delivery to the river network, both upstream and within the project area. These increases in flow will expand flooding extent in the Middle Hoh and increased sediment loads may aggrade the channel bed, further exacerbating flooding and water level rise. The increase in annual peak flow magnitude will also contribute to increases in channel migration rates and erosion. The result if this increased erosion will be to further limit the availability of off-channel habitats as they will be re-worked by the river more frequently, impacts to infrastructure will be more frequent and of greater magnitude, and private property loss should be expected to continue. Table 14. The magnitude of future peak flows 2070-2099 projected as result of warming climate (A1B scenario). RECURRENCE INTERVAL PRESENT DISCHARGE ESTIMATE (CFS) PERCENT INCREASE DUE TO CLIMATE CHANGE FUTURE (2070-2099) DISCHARGE ESTIMATE (CFS)  1-year 12,280 14 – 34% 13,975 – 16,427  10-year 52,300 14 – 34% 58,574 – 68,850  100-year 73,600 14 – 34% 82,140 – 96,551   The warming climate will also result in lower base flows in the late summer and fall. This will diminish the wetted channel area and depths that will decrease available habitat. Warmer air temperatures will also result in warmer water in the river, further stressing fish (Mantua et al., 2010; Isaak et al. 2017, Winkowski 2020) and making riparian shade and in-stream cover even more important for salmonids. While these changes were not quantitatively evaluated by the scope of this study, they further underscore the importance of increasing the ecological resiliency of the project area. In practice, this could be accomplished by increasing stream shade, the quantity of off-channel habitats and deep pools, as well as increased surface water/ground water interactions through increased floodplain connectivity and added roughness so that the predicted increases in stream temperature can be partially offset by temperature decreases. Sediment Sources Sediment transport can be divided into two general categories: Supply Limited Trends (Incision and straighter channel with lower erosion rates) Transport Limited Trends (Deposition and braided channel/wander channel forms with higher erosion rates) Channel forms observed in the air photo record suggest that past century sediment transport conditions and trends are neither predominantly supply nor transport limited. The BOR hypothesizes that sediment transport conditions are generally balanced. Field conditions identified that while the sediment supply is robust, the Hoh River has the ability to transport this supply over time. Field observations indicated that some localized areas are experiencing net deposition (transport limited) conditions, whereas observations at other areas indicated that the river can be locally supply limited. The long-term trend since deglaciation has been one of net supply limited conditions which has driven incision and resulted in stranded alluvial terraces. Vertical changes in the channel can influence channel migration. Primary drivers in vertical changes in the channel are sediment supply, roughness, and hydrology. Within the study reach, the net, long-term Holocene trend was incision and evidence of the downward vertical movement of the channel is supported by the presence of numerous terraces within the valley. The incision is controlled in part by the bedrock-lined canyons. Over time, the channel slowly incises into the bedrock grade controls and the channel lowers. Periods of aggradation (deposition) or vertical stability have likely occurred within the overall long-term incision trend. During these periods, channel migration areas would be expected to widen. Based on the stream gradient and active erosion observed within the bedrock- controlled areas, the long-term incision trend is likely continuing. However, this trend is very slow and channel migration processes over the 100-year planning horizon are likely more influenced by periods of aggradation and stability. Both weather variability and climate change can alter hydrology and sediment transport conditions of the watershed. Periods of both increases and decreases of sediment recruitment volumes delivered to the channel are anticipated in the future. Event driven sediment delivery increases occur from hydrologic events, slope instability, and glacial retreat. Event driven sediment delivery to the fluvial system can create pulses of sediment that may take years to transport through the project reach. Once in the fluvial system, the sediment is temporarily stored in the channel and on the floodplain and periodically remobilized through lateral erosion processes that also create temporal increases in sediment load and sediment pulses that migrate downstream. Conversely, longer term stabilization of some sediment supply is possible with vegetation establishment. This is true within the channel migration zone as well as in the sediments exposed as glaciers retreat. The variability of event driven processes and uncertainty of future hydrologic conditions makes predictive estimates of long-term sediment supply speculative. The trends point to a shift to increased sediment loads as the climate warms, driven by more mass wasting and glacial retreat exposing and releasing more sediment available. The fate of this sediment will be a function of a number of variables that are dynamic in nature, therefore, planning for intermediate and longer-term should consider a wide range of variability in sediment loading and transport. However, the overarching trend in the short to intermediate-term is for higher sediment loads that could elicit a channel response in the Middle Hoh. Indicators that would suggest sediment storage is increasing include an increase in the rate of avulsions, channel migration, channel braiding and an increase in flood elevations and extent. Forests As previously noted, climate change is increasing temperatures, reducing snowpack and changing west side Olympic Peninsula precipitation patterns resulting in more winter flooding, and lower summer stream flows (Halofsky et al. 2011). Changing temperature regimes and low snowpack have resulted in extreme drought years such as 2015, when the Paradise Fire, a rare extensive coastal temperate rain forest fire, occurred in the Upper Queets River basin burning more than four square miles of rainforest (2,796 acres) (Current Fire Status - Olympic National Park (U.S. National Park Service) (nps.gov)). Climate changes impacting precipitation, snowpack and temperatures are projected to increase through the 21st century with increasing wild fire potential within the temperate rain forests of the Olympic Peninsula (Perry et al. 2015; Halofsky et al. 2011). Riparian forests are naturally resilient to adjacent forest fires due to their higher levels of humidity and therefore provide floodplain side channel networks thermal refugia in a changing climate. In addition to climate increased fire frequency, channel migration rates are projected to increase that threaten to erode the riparian forests. Removal of much of the Middle Hoh River riparian forest in the 20th century has reduced the area of mature floodplain forest and associated side channel network–the area of mature side channel aquatic habitat. Mature riparian forests are the source of large conifer riparian trees–keystones of mature riparian and aquatic ecosystems. Since riparian forest removal, red alder floodplains have replaced the mature conifer deciduous patchwork floodplain forest. The loss of large keystone coniferous trees has further resulted in loss of large wood jams, in-channel habitat, and forested islands. Wood jam islands are stable refugia for establishment of riparian conifers. Alder dominated floodplains without wood jam hard points, together with climate driven increases in channel migration, portrays a future floodplain forest continuously dominated by red alder. Without large wood jams very few floodplain conifers will become established as the channel migrates freely across the valley bottom. Climate increased channel migration will in turn require larger trees to create stable wood jams. In effect the current and future river and riparian forest ecosystem has shifted from an historic state of mid-channel mature forested islands; anastomosing primary and secondary channel network and patchwork of mature floodplain forests; to a simplified more braided floodplain state dominated by young red alder and characterized by a lack of stable large wood. This scenario will result in a future Middle Hoh River–likely on the order of hundreds of years to recover–in which large wood generated patchwork channel and floodplain forest processes are lost resulting in continuous degradation and loss of both in-channel salmonid habitat and riparian forest fish and wildlife habitat. Aquatic Habitat Aquatic habitat in the Middle Hoh River will change and redistribute within the reach as the channel and riparian corridor respond to changes in the flow and sediment regimes. As a result of the changing climate the channel is anticipated to have a higher and increasing channel migration and avulsion rates due to changes in hydrology. As these processes become more dominant it will lead to a wider channel, as mature forests are lacking along the channel banks that could slow erosion. This wider channel, coupled with higher sediment loads associated with receding glaciers, will increase the potential for channel aggradation and a more braided planform. The cumulative effect of these changes will be a more simplified channel, with fewer and more shallow pools and unstable off-channel habitat as the channel freely migrates and avulses through the floodplain. The transition of the riparian forests to younger and more deciduous dominant will reduce local LWD inputs as the few remaining small patches of mature timber are eroded. Fewer large trees in the channel will further reduce the number of forced pools with cover, smaller and more few stable logjams in the channel and an overall more simplified channel lacking hydraulic complexity. All these changes will negatively impact habitat conditions in the Middle Hoh for salmonids at the most critical times in their entire life history: rearing and spawning. Destabilized floodplains frequently eroding cannot provide the off channel refugia of more mature floodplains with established forests and side channel networks connected to hyporheic flow, all needed for rearing juveniles. Diminished instream complexity reduces hydraulic complexity and slow water refugia in the channel and the loss of large trees reduces instream cover, creating an unhospitable main stem channel for rearing juveniles. Fewer and more shallow pools will further reduce juvenile refugia as well as holding spots for migrating adults. Spawning adults will also find it harder to locate preferred spawning areas as sediment will tend to segregate by size less as hydraulic complexity diminishes and riffles are buried in sediment. Cumulatively these impacts to aquatic habitat will most impact rearing juveniles as off-channel and instream habitats become increasingly degraded. Invasive Species Trends Insert text DESIRED FUTURE CONDITIONS A key element to the development of this resiliency plan is to identify desired future conditions, to provide a target through which we can develop a series of steps to achieve. Incorporating our understanding of the reach with projected and anticipated changes provides an opportunity to improve resiliency in the Middle Hoh. Of particular interest are how restoration of riverine processes, improvements to roads or changes in landuse and public access could improve resiliency for the community and environment. Short-, intermediate- and long-term desired future conditions statements for each of these categories are provided that include overarching goals and specific objectives to meet these conditions. Restoration Corridor The core of our approach to the development of resiliency planning for the reaches of the Middle Hoh is the delineation of a resiliency corridor, which estimates the minimum spatial footprint required to sustain habitat-forming processes along the river corridor. The processes that form and sustain aquatic habitat in a western Washington river system – lateral channel migration, channel avulsion (i.e., rapid channel relocation), flow splitting, overbank inundation – are dynamic and require both sufficient space and a healthy riparian forest to function (Collins et al., 2003; 2012). NSD utilized an examination of the geomorphic, hydraulic, and aquatic habitat parameters to define the Resiliency Corridor for the study reach. The resiliency corridor was defined as the entire valley bottom throughout the project reach based on following: The 100-year floodplain which includes areas frequently subject to flood inundation, The channel migration zone, Floodplain topography and geomorphology, including areas with clear evidence of historical alluvial channels (GMZ), The meander bend amplitude and frequency. The boundary of the resiliency corridor should be viewed as a planning tool to guide decision making for actions occurring within the corridor. Actions taken within the corridor should consider their impact on natural process functions contributing to available high-quality habitat. Areas within the corridor that are undeveloped should remain so, be evaluated for invasive plant species presence and need for restoration. Where natural processes are actively creating and supporting aquatic habitat and/or climax riparian communities exist the area should be targeted for preservation and invasive species monitoring. Where development exists within the corridor there is an opportunity to improve resiliency for the Hoh River, and if possible, restoration should be the preferred alternative. If restoration is not achievable, alternatives that minimize impacts and/or mitigate for impairments to natural processes should be pursued to the extent possible. Acquisition and land conservation provide for the greatest opportunity to develop long-term resiliency, that is naturally sustained and provides for people and many other dependent species. The Middle Hoh is unique in that there are currently 194-ac of land in conservation and 374-ac of federal or state ownership within the resiliency corridor that would inhibit future development. This represents 74% of the corridor, a wonderful foundation to continue to build upon with future acquisitions and conservation easements. When private property comes up for sale in these areas it should be purchased. Acquisition will eliminate future flood damages and enable habitat restoration. Landowners that are receptive can consider conservation easements as a means of preserving natural areas while retaining ownership. Funding programs also exist to help homeowners relocate outside flood prone areas. This will also reduce flood damage liabilities to individuals and the public and consistent with current guidance by the Association of State Floodplain Managers (ASFPM) and the Federal Emergency Management Agency (FEMA). Long-Term Desired Conditions (include overarching goals and specific objectives) Looking beyond 20 years the goal is to have completed all of items in the action plan, and focus the effort on continued passive restoration, monitoring and maintaining community engagement to ensure this vision for the Middle Hoh can persist well into the future. The river will appear different that it does today, with more channel threads and instream islands resulting in narrower and deeper channels on average. Most of the energy of the river is partitioned interacting with stable instream wood, instead of being released eroding channel banks, resulting in diminished channel migration rates that have allowed the floodplain riparian forest to mature. It will take more than a century for these forests to produce trees of sufficient size to form stable logjams in the river once recruited, however conifer release and planting are accelerating natural succession of the riparian forest. The ELJs constructed in the river and floodplains mimic the role of these large trees once played in the channel, allowing the system to recover and begin producing new trees to replace the ELJs once they deteriorate. The expected lifespan of the ELJs suggests that additional phases of ELJs construction will be needed to ensure the system remains stable long enough for the riparian forests to mature sufficiently to become self-maintaining. The findings of the Upper Hoh Road Plan successfully identified an alternative route for the road that meets the needs of the community, is outside of the resiliency corridor, CMZ, and is on low-landslide risk terrain (or the risk has been otherwise mitigated). This new route maintains access to the river, providing landowner and recreational access, while problematic chronic washout and side-prone sections of the road have been decommissioned and removed. Traffic heading to and from ONP on the new route has diminished overall traffic close to the river and homes, with individuals looking to access the river locally are directed on spur roads to established sites and local businesses. Recreation opportunities in the Middle Hoh River have been expanded to include a trail network connecting Highway 101 and the Hoh Visitor Center, new interpretive trails along the river educating the public about the ecosystem, bike lanes along the new Upper Hoh Road to reduce traffic and noise and improved stabilized boat launch sites that persist following high water. Working with entities that manage, own property or enter into conservation easements for ecosystem preservation to consider maintaining existing public access to the river so individuals can return to their favorite places. Intermediate-Term Desired Conditions For the purposes of this discussion, intermediate-term refers to the next 5 - 20-yrs and is a time when active restoration efforts are underway and wrapping up, having set the stage for a self-sustaining ecosystem where passive actions can be used to accommodate unanticipated shifts due to climate change. Active restoration actions include continued focus on restoration of off-channel rearing habitat and partitioning flow to reduce stream energy. Restoration projects have been implemented at the sites ranked as having the highest priority for restoration, and progress is continuing down the priority list toward the bottom. Monitoring is showing more juveniles out-migrating to the sea, indicating restoration actions are having a measured effect that will only continue to improve. Returning adults encounter more deep pools as they migrate upstream from instream ELJs forcing scour, creating resting spots as they move upstream. All these changes are making it easier for the salmonids that use the Middle Hoh to adapt to other changes brought about by the changing climate. Known patches of invasive species have been eradicated and continued monitoring has limited spreading within the Middle Hoh. Work is beginning after years of planning on a new road that will reduce the impact on the environment and is safer for all users and the local community. Most traffic has been routed further from local homes and the river, while maintaining access. Refinements to the alignments continue as new information is gathered and/or conditions change. The chronic road washout locations on the Upper Hoh Road have been stabilized and continue to provide protection from erosion as the new alignment is constructed. A series of trails connecting Highway 101 to the Hoh Visitor Center are under construction providing alternative access to ONP, or an alternative to visiting ONP during high-demand times. Additional interpretive trails along the river are being built that highlight the restoration work underway and the vision for the future. Short-Term Desired Conditions For the purposes of this discussion, short-term refers to the next 5-yrs. Actions that can be achieved in the short-term include those such as riparian conifer inter-planting, invasive species monitoring and treatment, thinning, stabilize known chronic road washout locations, opportunistic acquisition of property within the resiliency corridor, establishing a group of stakeholders that will work to realize the goals of this plan. Floodplain protection from channel migration, establishing islands in the channel that are protected from erosion Over the next five years we envision the establishment of a Middle Hoh working group consisting of a range of stakeholders who work to move the plan forward. A riparian forest management plan for the Middle Hoh was developed that has guided the initiation of silviculture actions aimed at acceleration the conversion to a mature forest, including conifer interplanting, thinning, understory management, invasive species monitoring and control, as well as formation of volunteer groups. Instream restoration projects have been implemented where high-quality floodplain habitat was at most risk of loss through erosion, utilizing ELJs strategically placed in the channel, banks and floodplain to mimic the role large wood once played in the Middle Hoh. These ELJs are used to limit avulsion potential through side and relic channels, reduce channel migration rates through young floodplain forests and split the main stem channel, forming instream islands that can persist over time. As these projects begin to be implemented new islands are forming in the river and vegetation is starting to colonize. Conifers are emerging from the dense alder stands where thinning and planting was completed, and understory invasive species have been eliminated or are under some form of management. In response to a broader national strategy to reduce the resource burden on the National Park Service, a reservation system was established for ONP visitors to the Hoh Rain Forest. This change in policy has increased demand for recreation and engagement with nature in the region, and decreased traffic along the Upper Hoh Road as a result. While overall traffic is down, more recreational users are looking to the Middle Hoh Chronic washout locations along the Upper Hoh Road have been stabilized using wood-based structures, and a plan has been developed to consider realignment where the road is within the Resiliency Corridor and/or CMZ. Improvements to the stability of existing boat launches have been improved or have been moved to limit susceptibility to erosion during high flows. Boat launches that have been lost to erosion over time are under evaluation to determine need and appropriate locations. LOCAL CAPACITY Insert text here PHASE II APPROACH To realize the potential of this Resiliency Plan, the formation of a permanent group of individuals (Middle Hoh Resiliency Initiative) is needed to ensure all groups working in the community are working together in a coordinated effort toward the common vision(s) articulated in the desired conditions. This group should include at a minimum several local landowners, the Hoh Tribe, Jefferson Co., local non-profits and conservation organizations, recreational users and agency representatives to further ensure any potential actions are vetted and considered openly and news can be communicated. The group will be needed to initiate and develop the additional plans needed to realize the desired future conditions, as well as coordinate to help locate funding sources for individual projects. While the Action Plan (Appendix B) identifies specific opportunities and prioritizes their implementation, there is additional, more detailed planning and study needed to move specific elements of the resiliency plan forward in a coordinated and informed manner. Completion of these studies and plans will provide a framework for moving forward larger components of the Action Plan that require additional information and/or coordination. This additional work described as part of Phase II should proceed significant progress on the Action Plan. The anticipated changes to the sediment and flow regime of the Hoh River resulting from climate change is not well constrained and should be a primary consideration for planning purposes. One of the more pressing studies needed is an in-depth evaluation of the anticipated flow and sediment regime changes in the Middle Hoh related to climate change. Warming temperatures coupled with changing weather patterns will alter the rate of surficial processes and sediment production, both of which are anticipated to increase the sediment load entering the Middle Hoh. Flood flows are predicted to increase in magnitude and frequency, and summer base-flows are anticipated to diminish. A better understanding of how these changes will manifest on the landscape over time would greatly benefit the work completed and inform future planning efforts. A comprehensive survey of side channels within the Middle Hoh would further inform and refine prioritization of floodplain habitat protection and enhancement within the reach. A complete inventory of the side channels present and their habitat and hydrologic character, fish usage and riparian conditions would facilitate a more informed decision about the properties to target for preservation and where and what restoration actions are needed. Having functional off-channel habitat is critical to maintain anadromous salmonid populations in the Middle Hoh and needs to be a critical component of the overall restoration strategy. The state of the riparian forest corridor along the Middle Hoh largely dictates the health of the fish using the river. While the Hoh is known internationally for hosting ancient old-growth forests, these are largely confined within the boundaries of ONP. Along the Middle Hoh, most of the riparian corridor has been logged at some time in the past and is in some state of forest succession. A more detailed and complete Riparian Forest Plan could more deeply integrate silviculture actions with invasive species management and restoration actions into a cohesive strategy that can be implemented throughout the reach. A comprehensive plan for the Upper Hoh Road is needed to layout a shared vision for what the road should be, that best serves the people that use it and the surrounding environment. The plan should explore opportunities for relocation, abandonment, improvements or some combination(s) thereof to the road that could improve the user experience and reduce environmental impacts and risks. Identifying an improved alternative route out of the northern side of the valley should be included in the plan for the Upper Hoh Road to improve community resiliency. Recreation Plan to explore opportunities for larger recreation features (trails), ensure visitor use does not detract from the environment (garbage?), and the number, location and types of day-use areas along the river (including boat launches). Bring together local community, guides, day-users in community, NPS?, county and landowners to explore alternatives that can improve the experience for everyone while maintaining the wildness of the river (limited paved parking areas) REFERENCES Abbe, T. and D.R. Montgomery. 1996. Large woody debris jams, channel hydraulics, and habitat formation in large rivers. Regulated Rivers: Research and Management, 12, 201-221. Abbe, T.B. and D.R. Montgomery, 2003, Patterns and Processes of Wood Accumulation in the Queets River Basin, Washington, Geomorphology 51:81-107. Abbe, T., J. Bountry, L. Piety, G. Ward, M. McBride, and P. Kennard. 2003. Forest Influence on Floodplain Development and Channel Migration Zones. Geological Society of America Abstracts. Abbe, T. and A. Brooks. 2011. Geomorphic, Engineering and ecological considerations when using wood in river restoration. In Simon, A., Bennett, S., and Castro (Eds.), Stream Restoration in Dynamic Fluvial Systems: Scientific Approaches, Analyses, and Tools. Geophysical Monograph Series 194. AGU. pp.419-451. Abbe, T., B. Belby, M. Fox and D. Shields. 2016. Geomorphology and Hydrology Considerations. Chapter 4, National Large Wood Manual: Assessment, Planning, Design, and Maintenance of Large Wood in Fluvial Ecosystems: Restoring Process, Function, and Structure. pp.234-326. Bureau of Reclamation and U.S. Army Corps of Engineers. 2016. National Large Wood Manual: Assessment, Planning, Design and Maintenance of Large Wood in Fluvial Ecosystems: Restoring Process, Function, and Structure. 628 pages + Appendix. Babinski, Z. 2005 The relationship between suspended and bed load transport in river channels. Sediment Budgets 1. (Proceedings of symposium S1 held during the Seventh IAHS Scientific Assembly at Foz do Iguacu, Brazil, April 2005) IAHS Publ. 291. Beechie, T. J., M. Liermann, E. M. Beamer, and R. Henderson. 2005. A classification of habitat types in a large river and their use by juvenile salmonids. Transactions of the American Fisheries Society 134: 717–729. Beechie, T. J., and T. H. Sibley. 1997. Relationships between channel characteristics, woody debris, and fish habitat northwestern Washington streams. Transactions of the American Fisheries Society 126:217-229. Beechie, T. J., O. Stefankiv, B. Timpane-Padgham, J. Hall, G. R. Pess, M. Rowse, M. Liermann, K. Fresh, and M. Ford. 2017. Monitoring salmon habitat status and trends in Puget Sound. NOAA Technical Memorandum NMFS-NWFSC-137. https://doi.org/10.7289/V5/TM-NWFSC-137. Bisson, P. A., J. L. Nielsen, R. A. Palmason, and L. E. Grove. 1982. A system of naming habitat types in small streams, with examples of habitat utilization by salmonids during low stream flow. Pages 62-73 in N. B. Armantrout, editor. Acquisition and salmonid. Environmental Biology of Fishes 8:203-216. CHaMP (Columbia Habitat Monitoring Program). 2016. Scientific protocol for salmonid habitat surveys within the Columbia Habitat Monitoring Program. Prepared by CHaMP for the Bonneville Power Administration. Collins, B.D., D.R. Montgomery, K.L. Fetherston and T. Abbe. 2012. The Floodplain large-wood cycle hypothesis: a mechanism for the physical and biotic structuring of temperate forested alluvial valleys in the North Pacific Coastal Ecoregion. Geomorphology 139/140, 460-470. Cram, J., N. Kendall, A. Marshall, T. Buehrens, T. Seamons, B. Leland, K. Ryding, and E. Neatherlin. 2018. Steelhead at risk report: Assessment of Washington’s steelhead populations. Fausch, K. D., and T. G. Northcote. 1992. Large woody debris and salmonid habitat in a small coastal British Columbia stream. Canadian Journal of Fisheries and Aquatic Sciences 49(4):682-693. Frissell, C. A., S. B. Adams, and N. H. Hitt. 2000. Identifying priority areas for salmon conservation in Puget Sound basin. Flathead Lake Biological Station Open File Report, The University of Montana, Polson, MT. 155 p. Halofsky, J.E., D.L. Peterson, K. O’Halloran, and C. Hawkins Hoffman. 2011. Adapting to climate change at Olympic National Forest and Olympic National Park. USDA Forest Service General Technical Report PNW-GTR-844. Pacific Northwest Research Station, Portland. Hoh Tribe. 2016. State of Our Watersheds Report. Hoh River Basin. Available at: https://geo.nwifc.org/SOW/SOW2016_Report/Hoh.pdf Huntington, C., W. Nehlsen, and J. Bowers. 1994. Healthy stocks of anadromous salmonids in the Pacific Northwest and California. Oregon Trout, Portland, Oregon. Isaak, D., S. Wenger, E. Peterson, J. Ver Hoef, D. Nagel, C. Luce, S. Hostetler, J. Dunham, B. Roper, S. Wollrab, G. Chandler, D. Horan, S. Parkes-Payne. 2017. The NorWeST summer stream temperature model and scenarios for the western U.S.: A crowd-sourced database and new geospatial tools foster a user community and predict broad climate warming of rivers and streams. Water Resources Research, 53: 9181-9205. https://doi.org/10.1002/2017WR020969 Kramer, N., and E. Wohl. 2017. Rules of the road: A qualitative and quantitative synthesis of large wood transport through drainage networks. Geomorphology. 279:74-97. Leopold, L. B. and M. G. Wolman. 1957. River channel patterns: braided, meandering and straight. U.S. Geological Survey Professional Papers 262B:39-85, Washington, D.C. Lieb, A., T. Perry. 2005. Seasonal Variability for Select Streams within WRIA 20, Olympic Peninsula, Washington. Bureau of Reclamation Technical Services Center. Lyons, E.W. 2003. Mass wasting in the Upper Hoh watershed, Olympic National Park, Washington: Interpreted from 1939 and 2001 Aerial Photography. U.S. Bureau of Reclamation. McHenry, M.L. 2001. Fisheries habitat module. Middle Hoh River Watershed Analysis, Washington State Department of Natural Resources, Forks, WA. McHenry, M.L., J. Lichatowich, and R. Hagaman. 1996. Status of Pacific Salmon and their habitats on the Olympic Peninsula watersheds. Washington Department of Ecology, Olympia. McMillan, J. R., and J. C. Starr. 2008. Identification and prioritization of salmon tributaries for conservation in the Hoh River Basin, Washington State. Wild Salmon Center, Portland, Oregon. Montgomery, D.R. and T.B. Abbe. 2006. Influence of logjam-formed hard points on the formation of valley-bottom landforms in an old-growth forest valley, Queets River, Washington, USA. Quaternary Research 65: 147-155. Montgomery, D. R., J. M. Buffington, R. D. Smith, K. M. Schmidt, and G. Pess. 1995. Pool spacing in forest channels. Water Resources Research 31(4): 1097-1105. Nelson, L. 1986. Fluvial Sediment in the Hoh River. In Reconnaissance of the Water Resources of the Hoh Indian Reservation and the Hoh River Basin, Washington. USGS Water-Resources Investigations Report 85-4018. North Pacific Coast Lead Entity. 2020. North Pacific Coast (WRIA 20) salmon restoration strategy. NV5 GeoSpatial. 2021. Hoh River, Washington orthophotography, technical data report. Technical Report to Jefferson County, Port Townsend, Washington. Parks, D. 1999. Mass wasting module, in Hoh River Watershed Analysis, Middle Hoh and Rainforest Watershed Analysis Units. Level 2. Washington Department of Natural Resources. Appendix A. Peterson, N. P., and L. M. Reid. 1984. Coho riverine habitat: Its evolution, distribution, and use. Pages 215-225 in Proceedings of the Olympic Wild Fish Conference, Fisheries Technology Program, Peninsula College, Port Angeles, Washington. Perry, L.G., L.V. Reynolds, T.J. Beechie, M.J. Collins and P.B. Shafroth. 2015. Incorporating climate change projections into riparian restoration planning and design. Ecohydrology. Piety, L. A., J. A. Bountry, T. J. Randle, and E. W. Lyon. 2004. Summary report for geomorphic assessment of Hoh River in Washington State. River Miles 17 to 40 between Oxbow Canyon and Mount Tom Creek. Report to the U.S. Department of the Interior Bureau of Reclamation. Quinn, T. P. 2018. The behavior and ecology of Pacific salmon and trout. University of Washington Press, Seattle. Smith, C. J. 2003. Salmon habitat limiting factors in Washington State. Washington State Conservation Commission, Olympia, Washington. Thackray, G. 2001. Extensive Early and Middle Wisconsin Glaciation on the Western Olympic Peninsula, Washington, and the Variability of Pacific Moisture Delivery to the Northwestern United States. Quaternary Research 55(3): 257-270 WDFW (Washington Department of Fish and Wildlife) and NWIFC (Northwest Indian Fisheries Commission). 2014. Statewide Washington Integrated Fish Distribution (SWIFD) Schema and data dictionary. Pages 5 in a. t. N. I. F. C. Washington Department of Fish and Wildlife, editor. Olympia, Washington. Welber, M., W. Bertoldi, and M. Tubino. 2013. Wood dispersal in braided streams: Results from physical modeling. Water Resources Research 49:7388-7400. Winkowski, J. 2020. The Olympic Peninsula Thermalscape Model: A Spatial Stream Network Model for August Stream Temperature on the Outer Olympic Peninsula, WA (WRIA 20 and 21). Washington Department of Fish and Wildlife Coast Ecology and Life Cycle Monitoring Unit. APPENDIX A HABITAT SURVEY DATA COLLECTION Table 1: Data collected at the habitat unit level for habitat surveys for mainstem habitats. UNITS DESCRIPTION DROPDOWNS UNITS  fk_Survey Unique Survey Number - -  pk_Units Unique Unit Number - -  DateCreated Date Surveyed - DD/MM/YYYY  DateModified Date Modified - DD/MM/YYYY  GPS_Top_Lat GPS coordinate taken from the Bad Elf GNSS Surveyor - DD  GPS_Top_Long GPS coordinate taken from the Bad Elf GNSS Surveyor - DD  GPS_Top_EPE GPS coordinate taken from the Bad Elf GNSS Surveyor - -  GPS_Bottom_Lat GPS coordinate taken from the Bad Elf GNSS Surveyor - DD  GPS_Bottom_Long GPS coordinate taken from the Bad Elf GNSS Surveyor - DD  GPS_Bottom_EPE GPS accuracy taken from Bad Elf GNSS Surveyor - -  ChannelType MS Channel type Main channel, braid, side channel -  UnitNumber Non-unique sequential unit number for survey - -  UnitType_Dominant Unit type as defined by Bouwes et al. 2011, Units must be at least one wetted width in length and occupy 50% of the wetted width. Pool, riffle, glide, ponded area, backwater -  PoolType Pool type as defined by hydrology (Bisson et al. 1982) Plunge pool, scour pool, trench pool, dammed pool, backwater pool -  PoolFormingFeature Feature causing scour and pool formation LW Piece, LW jam, boulder, bedrock, beaver dam, confluence, channel bend -  UnitType_SubDominant A subdominant unit must be at least one wetted width in length Pool, riffle, glide, ponded area, end point, backwater -  SubDomUnitPercent Percentage of channel occupied by the subdominant unit - %  UnitLength Total wetted length of unit - m  UnitWidth25 Wetted width at 25% of unit length - m  UnitWidth50 Wetted width at 50% of unit length (taken for pond units) - m   UnitWidth75 Wetted width at 75% of unit length - m  Comments - - -  NoJamsPresent - Y/N -  NoLWDPiecesPresent - Y/N -   Table 2: Data collected for edges unit level for habitat surveys for mainstem habitat types. EDGES DESCRIPTION DROPDOWNS UNITS  fk_Units Unique Unit Number - -  pk_Banks Unique bank identifier - -  DateCreated Date Surveyed - DD/MM/YYYY  DateModified Date Modified - DD/MM/YYYY  Bank Left or right bank Left, right -  EdgeType Edge habitat as defined by Hayman et al. (1996) Bar edge, natural bank edge, other, hydro-modified bank edge, riprap bank edge, levee bank edge, placed LW, deflectors, pilings -  EdgeLength Percent of channel unit length occupied by each edge type for left and right bank. 1-100 (increments of 10) %  SlowWater Presence of slow water next to bank. Y/N -  SlowWater_EdgeWidth Width of slow water next to edge. - m   Table 3: Data collected for large wood for each unit for habitat surveys for Mainstem habitats. Large wood was only recorded if a qualifying piece of the jam was wetted at the time of the survey. LWD DESCRIPTION DROPDOWNS UNITS  fk_Units Unique Unit number - -  pk_LWD Unique LW piece identifier - -  DateCreated Date Surveyed - DD/MM/YYYY  DateModified Date Modified - DD/MM/YYYY  LWDJamNum Count of LW jam located within the wetted channel or within 10m of distance and 1m of vertical elevation from the wetted edge - -  JamLength Visually estimated length of the jam. - m  JamWidth Visually estimated width of the jam. - m  JamUnitCoverPercent Percent of the jam that is wetted. - -   APPENDIX B MIDDLE HOH RIVER ACTION PLAN