sign in sign up
<<
Home , Avulsion (river), Canyon, Erosion, Revetment, River channel migration, River morphology

Erosion and Hazard Mapping and Methodologies for use in the Nooksack Migration Zone Mapping

By Paul Pittman LEG, Whatcom County Public Works and Peter , Whatcom County Planning and Development Services

Final Internal Draft September 15th, 2009

5/26/2015 - 0 - and Avulsion Hazard Mapping and Methodologies for use in the Nooksack Zone Mapping

1.0 Purpose The purpose of mapping the erosion and avulsion hazards of the Nooksack River was necessitated by both riverine hazard planning through the Comprehensive Hazard Management Plan and recent updating of the Shoreline Master Plan and Critical Areas Ordinance. The intent of this study is to provide a technical background document and set of maps to help guide decision makers in adopting a channel migration zone and developing comprehensive flood hazard and ecological planning. The combination of the historic, erosion and avulsion hazard mapping can be the foundation for delineating Channel Migration Zones (CMZ) (Rapp et al, 2003).

1.1 Background The Shoreline Master Plan (SMP) and Critical Areas Ordinance (CAO) updates of 2004 and 2005 are the catalyst for mapping a Channel Migration Zone (CMZ). However, the concept of a CMZ has its roots in multiple management plans, acts, and regulations: • 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 and ” (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 encourages development of a “ belt” delineation in Comprehensive Flood Hazard Management Plans. The Whatcom County Lower Nooksack River Comprehensive Management Plan (1999) states that the Channel Migration Limits will be mapped. 1.2 Definitions The “historic migration zone” is a composite of the historic locations of the river as determined from historic information and interpretation, over some length of historic record. The term “erosion hazard” in this analysis refers to the hazards resulting from the natural process of lateral migration of a channel through erosion that occurs from channel expansion, channel meandering, channel course changes, or channel bank and fluvially related slope failures.

5/26/2015 - 1 - The term “avulsion hazard” describes a multiple set of hazards associated with rapid channel course changes or temporary channelization of flow that in addition to having flooding hazards, instantaneously becomes an erosion hazard. The combination of the historic migration zone, erosion hazard area and avulsion hazard zone, in general terms, is “the geographic area where a or river has been and will be susceptible to channel erosion and/or channel occupation” (Rapp et al, 2003). The State of , 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 , 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. The area within which a river channel is likely to move over a period of time is referred to as the channel migration zone (CMZ) or the meander belt.” 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 aerial photos and maps and may require field analysis of specific channel and 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, 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.”

2.0 Hazards mapped on the Nooksack River

2.1 General Methodology 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 Historic Migration Zone (HMZ), Erosion Hazard Area (EHA),

5/26/2015 - 2 - and Avulsion Hazard Zone (AHZ) areas. A summary of DOE Methodology used to delineate Channel Migration Zones (CMZ) follows:

CMZ = (HMZ + EHA + AHZ) – DMA

The HMZ (Historic Migration Zone) is created using historical spatial information (maps, air photos, survey records). Whatcom County has this data going back to 1880’s. The HMZ for the Nooksack River was mapped (Collins and Sheikh, 2004; Appendix A) and modified by Whatcom County staff. The EHA (Erosion Hazard Area) is predicted horizontal channel migration potential through fluvial erosion moving laterally by eroding bank material (Figure 2-1). Lateral erosion is not necessarily limited to the or areas inundated during the 100-year flood event. For this study the fluvial-related geotechnical hazards are incorporated into the EHA, but are shown as an overlay on the maps. The AHZ (Avulsion Hazard Zones) are predicted locations of rapid channel change by capture of a relict channel or topographic low within the floodplain, or of temporary channel-like conditions that can exist and create erosion hazards (for example, a break). 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, , sediment supply and woody debris loading” (Rapp et al, 2003). It predicts possible hazard areas based on existing and historic process observations. This document incorporates the information from the Historic Channel Locations of the Nooksack River (Collins, et al, 2004; Appendix A) to complete the background scientific and historical mapping efforts as a component in mapping a CMZ or as a stand-alone document to assess erosion and avulsion hazards. The hazards mapped in this analysis do not suggest a level of acceptable risk. The mapping should be considered a coarse, assessment. Site-specific analysis should be performed for site level projects or more detailed assessment needs.

2.2 Study Area

The study area includes the Nooksack Valley and all areas potentially impacted by or adjacent to the Nooksack River channel migration (Figure 2-2 - vicinity and study area).

2.3 Base maps and existing information Information, mapping and documented studies were consulted whenever possible in order to build upon existing information. In particular, the Historic Migration Zone mapping effort was conducted by Collins and Sheikh and is presented in Appendix A. Changes to the maps or methods presented in our document may be necessary as new information becomes available.

5/26/2015 - 3 -

2.4 Mapping Efforts Paul Pittman of Whatcom County Public Works River and Flood took the lead on mapping the EHA and AHZs of the Nooksack River. Informal technical review and assistance was provided by: Tim Hyatt and Alan Soicher, LEG, of Nooksack Tribe Natural Resources; Michael Maudlin, LEG of Lummi Nation Natural Resources; John Thompson, LEG and Steve Fox, LG of Whatcom County Public Works; Doug Goldthorp, LEG, LHG of Whatcom County Planning and Development Services; Roger Nichols, LEG of the United States Forest Service; Patrica Olsen, LHG of the Department of Ecology; and Barry Wenger of the Department of Ecology. Oversight, review, and mapping assistance was provided by: Jeff Chalfant, of Whatcom County Planning and Development Services (SMP Update Coordinator); and Peter Gill of Whatcom County Planning and Development Services; Paula Cooper, PE of Whatcom County Public Works; the Whatcom County Zone District Advisory Committee (/Meander Subcommittee). Input, comments and local knowledge from the community was also gathered at public meetings.

2.5 Limitations of this study Flood hazards have historically been focused on rising water levels resulting from flooding. The premise of FEMA maps are based on flood elevations. While this flooding does provide significant risk and causes significant financial damages, additional flood hazards exist, and in some cases, have higher risks and cause even greater public expenditures. These additional flood hazards result from erosion and water velocities. This assessment estimates lateral erosion hazard potential from “normal” alluvial conditions. It should be used in conjunction with other hazard assessments and studies to make comprehensive planning decisions, assess risk, or guide additional studies. Multiple probabilities or levels of risk associated with specific identified hazards were not mapped primarily because of the unpredictable of fluvial systems over a long period of time and the lack of resources needed for an analysis of that depth. This study is considered a snap-shot in time using the current processes operating within the historic context provided by geomorphic , historic maps, photos, and previous studies. It is understood that there are known and unknown future changes that will alter the behavior of the fluvial system. Fluvial response to changes in , hydrology, , sediment supply, woody debris loading, and human modifications are difficult to predict with a high degree of confidence over the life of the study. Specific limitations include: • Mapping scale – temporal and spatial The mapping efforts for this analysis were done on a reach and sub-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

5/26/2015 - 4 - 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 Nooksack River. Other natural hazards in the Nooksack River Valley exist, including but not limited to: flooding, , volcanic eruption, mudslide, , tectonic deformation, flooding (clear-water, debris /flows), and .

5/26/2015 - 5 -

North Fork Nooksack, 1995

Canyon Creek, 1990

FIGURE 2-1: Local examples of lateral erosion impacting development.

5/26/2015 - 6 -

FIGURE 2-2: Nooksack River Vicinity and Study Area

5/26/2015 - 7 -

3.0 Applied Methodologies to delineate Erosion and Avulsion Hazard Areas

3.1 Geologic Setting The Nooksack Basin is located in the western flanks of the North and includes the north and west portions of (Figure 3-1). Metamorphic rocks, Mount Baker and Hannegan volcanics, and Chuckanut Formation sedimentary rocks dominate the bedrock geology of the western North Cascades of Whatcom County (Figure 3-2). The of the region was heavily influenced by the glacial events. Relict glacial deposits still many of the foothills and extensively cover the western portion of Nooksack Basin.

3.1.1 Holocene Geomorphic Setting - Natural Three primary forks of the Nooksack River exit the Cascades and converge near the western edge of the Cascade Foothills near Deming. The North and Middle Forks include the of the west and north flanks of Mount Baker and adjacent peaks. The South Fork drains the foothills of the Twin Sisters and the southwest foothills of Mount Baker. The Holocene (past 10,000 years) Nooksack River floodplain is primarily situated in relict glacial troughs and valleys. The mainstem floodplain splits near Everson and meets the southern at two locations – Vancouver (via the Sumas/Fraser River Valleys) and Lummi/Bellingham (via the Lower Nooksack Valley). The modern Nooksack River flows to Bellingham Bay by following the Lower Nooksack Valley west from Everson, however the Nooksack River occasionally becomes a to the Fraser River via the Sumas Valley during overbank flooding between Lawrence and Everson. Geomorphic evidence supports the theory that the Nooksack was a tributary to the Fraser River for a majority of the Holocene and only recently avulsed into the Lower Nooksack Valley near Everson (Figure 3-3; Maudlin and Pittman, 2003). Similarly, the modern South Fork Nooksack River floodplain splits near Saxon, where the modern floodplain merges with a former glacial valley that now contains the Samish River floodplain. The Samish River has historically conveyed overbank flooding from the South Fork. The Nooksack Basin covers over 800 square miles, with more than half the area in the step Cascade Mountains and Foothills. The Nooksack is a river throughout most of its total length. Geomorphic and historic information suggest that the river was primarily an anastomosing channel in most reaches upstream of RM 21 and in the upper Sumas Valley (south of Sumas), and had an emerging meandering morphology in the downstream reaches where valley gradient decreased and bank cohesion increased, especially in the lower Sumas Valley (north of Sumas) and locally in the lower South Fork Valley. Anastomosing channels migrate both by lateral erosion and avulsion, however in an uncontrolled, natural system it was likely avulsions that were responsible for

5/26/2015 - 8 - the most of the large lateral changes. The geomorphology of the Sumas Valley demonstrates the channel was capable of migrating across the 2 to 3-mile wide valley margin during a portion of the Holocene, and that significant avulsions could instantly move the primary channel from one side of a valley to the other. Much of the pre-disturbed modern channel morphology for the North and Middle Forks is unknown because rapid valley-wide channel migration over the past 100-years has erased most pre-1900 geomorphic channel features; however, channel morphology likely alternated between anastomosing and some degree of braiding depending on local, reach and basin scale changes and conditions. Lateral channel migration in the mountainous valleys of the North Fork, Middle Fork, upper South Fork, and upper Mainstem is constrained, and occasionally constricted by bedrock geology, deep-seated bedrock landslide deposits, and localized deposits of Mt. Baker Lahar. The Mainstem downstream of RM 28 (Lawrence) is in valleys that are flanked predominantly by Holocene and Pleistocene . Limits of lateral channel migration below Lawrence are constrained primarily by time that during the Holocene encroached upon both sides of the Sumas Valley, which is up to three miles wide.

3.1.2 Holocene Geomorphic Conditions - Modern Geomorphologic evidence, early surveys, and anecdotal documentation suggest that the characteristics of the modern Nooksack River channel that we observe today were evolved from the predominant natural conditions within the first few decades of agricultural, , and industrial development that occurred in the late 1800’s and early 1900s. The primary factors that initially modified the channel morphology were removal of debris jams for navigation and clearing of floodplain/riparian forests in the middle and late 1800’s. Beginning in the early 1900’s, hydromodification of the floodplain with and was undertaken to support . The responded to the disturbances by evolving into different channel morphology. The disturbances were numerous and included a decrease in channel roughness (removal of LWD), loss of controls and sediment (removal of LWD), decreased bank cohesion and strength (riparian clearing), decreased flood roughness and storage (floodplain clearing and draining), increased confinement (levees) channel shortening and straightening (for navigation and transport), increased sediment supply (upland clearing, drainage modifications, and increased lateral erosion) and transport. The result of these disturbances was an evolution from a narrower, more stable, often multi-channeled system, to a wider, less stable braided, single thread channel system. For example, the upper mainstem Nooksack River exhibited, based upon geomorphology and survey, an anastomosing/meandering morphology pre-development. By the 1930’s, it had evolved into an entrenched single-thread/braided morphology and many of the side channels and a large portion of the floodplain that were active in the late 1880’s were now isolated on a . In general, the modern Nooksack River channel is shorter, wider, is locally entrenched, has higher velocities, is capable of transporting more sediment, has decreased bank cohesion (from root strength), and lacks much of

5/26/2015 - 9 - the geomorphic complexity that results from floodplain connectivity, intact riparian , and large woody debris jams typical in unaltered fluvial systems of the western Cascade Foothills (Collins and Sheikh, 2002). Modern day hydromodification, riparian/floodplain conditions, and upper watershed land management activities have been the primary driver of the river morphology changes and patterns observed in the past century. The changes to the floodplain and active channel area include: a reduced active floodplain area, an increased active channel area, increased constriction points (reduced channel migration area), altered and depositional areas, increased sediment inputs, increased stream velocities and depths, altered bank cohesion properties (loss of riparian vegetation), reduced channel roughness (loss of large wood debris). Many channel forms are reactions inexorably linked by one action that causes a “domino effect” that impacts not only the localized area where the disturbance occurred, but migrates in multiple dimensions both spatially (upstream, downstream, laterally, vertically) and temporally. An example of a disturbance is LWD removal that was commonly done to manage navigation and flood hazards. Arguably, LWD has many functions, ecological, cultural, and geomorphic. It impacts channel morphology by acting as hydraulic roughness, increasing sediment storage and providing grade control in addition to many other things. Channel morphology in a natural system responds to LWD. When that independent variable is altered, one thing leads to another in a complex domino effect.

3.2 Scope & General Assumptions

3.2.1 Project Area (Figure 3-4 Reaches & Subreaches) The scope of this analysis is the mainstem Nooksack River from the delta to the (RM 37), the North Fork Nooksack River from the confluence to (RM 58), the Middle Fork Nooksack River from its confluence with the north fork to upper Mosquito Bridge (RM 5), and the South Fork Nooksack River from the confluence to the Skookum Creek Hatchery (RM 15). Limited assessments of , alluvial fans, overflow channels (Sumas River, Johnson Creek, and Samish River) were conducted if they fell within the margins or influences of the Nooksack River potential channel migration area.

3.2.2 General Assumptions The study includes mapped lateral erosion potential areas and avulsion prone areas on the Nooksack River for a 100-year time frame influenced by “normal” . The study considers erosion rates and processes observed in the past century as “normal” fluvial processes. The potential exists for catastrophic or rapid changes resulting from less frequent or unpredictable events (e.g. – massive ice or log jams, catastrophic floods, major , volcanic events, , and tectonic deformation, etcetera). It should be understood that unexpected and catastrophic events do occur and conditions as

5/26/2015 - 10 - they have existed during the past 100 years may change over the next 100-year time period. This document should be considered a living document and updated as needed. To map the lateral erosion and avulsion potential, the following assumptions were made: • All man-made structures were ignored and not considered as limiting lateral erosion and avulsion because of the uncertainty of their design capabilities and life expectancy, future upkeep and response to changing conditions. In some conditions, man-made features may increase the risks if they fail. Scenarios for failure of manmade features are extremely unpredictable and were not specifically mapped with the exception of the levee break avulsion hazard describe later in this document. • The hazard areas (Erosion Hazard Area (EHA) and Avulsion Hazard Zone (AHZ)) are defined in this analysis as “possible” erosion or avulsion within a 100-year time frame, as opposed to the “probable”. The specific methodology used is moderately conservative, meaning that more conservative and less conservative methodologies exist. No site-specific probability analyses were performed for this analysis. • The channel morphology and behavior observed in the past one- hundred years will be similar in the next 100 years. The rapid morphological changes that were observed following rapid agricultural and forestry disturbance have predominated over the historic record. Any significant recovery of the pre-disturbed conditions is anticipated to take multiple decades to centuries to dramatically affect the channel on a large scale.

3.3 Historic Migration Zone (HMZ) This analysis uses methods and maps from Historic Channel Locations of the Nooksack River, Collins and Sheikh, 2004 (Appendix A)

3.4 Erosion Hazard Area (EHA) The areas of “possible” lateral erosion from the Nooksack River were mapped as the EHA. The mapping reflects fluvial erosion potential based upon interpretations and predictions of geologic processes. Probability of erosion within this area was not analyzed. Best available science methodologies were used for a large, basin scale interpretation of geologic conditions and processes. Site-specific analysis may be needed to assess site-specific conditions. The analysis does not suggest a level of acceptable risk. Characteristics of an area experiencing lateral migration are: vertical or over-steepened eroded banks along the outside of a channel bend; a widening channel with steep banks; and bank sloughing from over-steeping by channel incision. Although lateral channel migration can occur even during low flow, the over-steepened banks typically fail more rapidly during fast moving water or on a

5/26/2015 - 11 - receding flood causing the channel to migrate laterally at faster rates. An area that is has steep banks with under- trees and vegetation that falls into the river can be evidence for lateral migration. Developments that encounter lateral erosion may “fall” into the river as opposed to developments that get wet by rising flood. It is common that developments that do fall into the river are above flood elevations. Revetments, or bank armoring, have traditionally been used to limit lateral migration. A vast majority of the public money spent on river management within Whatcom County over the past decade has been spent on revetments to manage lateral channel migration, not levees to address flood control.

3.4.1 Morphology characteristics and lateral migration patterns Two methods were used to estimate the EHA based upon channel morphology characteristics.

3.4.1.1 Wandering Channel Morphology Wandering channel morphology includes an irregularly sinuous channel that will predominantly exhibit either multiple channels or splits with vegetated (anastomosing) or braided channels with unvegetated bars commonly with seasonal or perennial side channels. Depending on temporal changes or disturbances to the system, a wandering channel may evolve between either of these two channel forms with time or even locally develop other channel forms such as a meander form. 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. In general, the erosion rates are relatively high with wandering channels because of: higher stream energy associated with steeper slopes, channel banks composed of material that is typically coarse and has relatively low cohesion, and channels have bedload sediment transport and rates that are relatively high with channel beds that are typically aggrading. For the wandering channel type, the Erosion Hazard Area (EHA) is a calculated average annual erosion rate for the identified reach as described and calculated by Collins and Sheikh (2004) multiplied by 100 (analysis time-frame, in years). This method represents a “possible” or potential erosion hazard area for the next 100 years. The method makes the assumption that lateral erosion could occur in one direction over the design life of the analysis (100-years). The occurrence of lateral erosion occurring in one-direction over 100 years was not observed on the Nooksack, however areas where significant erosion occurred were managed with revetments and thus, erosion patterns were altered. The probability of erosion occurring in one direction over an extended time interval is unknown. An average annual erosion rate for a reach was used as opposed to maximum erosion observed or site-specific observed erosion because it is felt that maximum erosion over estimates long-term trends, and site-specific erosion may not capture and represent all erosion processes that may occur over a long time span. However, the average annual erosion rate underestimates lateral erosion potential to some degree because the method negated influences that revetments had on erosion. Marrying the 100-year multiplier, which may over-

5/26/2015 - 12 - estimate erosion potential, and the average annual reach erosion rate, which may underestimates erosion rates, was felt to be a prudent and moderate method to predict possible erosion over a 100-year time frame. In many cases the argument is mute since the erosion hazard area along much of the Nooksack River was truncated by geologic factors that impede erosion well within the possible erosion hazard area. In locations where this wasn’t applicable, the erosion rate methodology used in this analysis corresponded well with alternative methodologies calculated by A. Soicher (personal correspondence) and geomorphic indicators (e.g. relict channels). The interpreted erosion hazard area used in this analysis was applied to all areas mapped as Holocene landward of the HMZ. The potential erosion hazard area was modified based upon the interpretation of erosion potential of geologic units ( 3-1), topography or other factors anticipated to limited lateral fluvial erosion. Sub-reach and limited site-specific conditions were also considered in the interpretations and estimates to arrive at a possible 100- year EHA.

3.4.2 Erodability of Banks Table 3-1: Estimated relative erosion potential used for “typical” geologic units. The relative erodablility of the materials used for this analysis is as follows (modified from WSDOT, 2001):

: Very low erosion potential within a 100- Bedrock year time period

: Low erosion potential Bedrock Landslide Deposit within a 100-year time period

Lodgment , Glacio-Marine Drift, Middle Fork Lahar: Moderate erosion potential within a 100-year time period Glacial Outwash: Moderate to high erosion potential within a 100-year time period Holocene Alluvium (including alluvial fans and other unconsolidated, fine-grain Holocene material – landslides, , etc.): Extremely high erosion potential within a 100-year time period General lateral erosion trends were observed. One is a decreased lateral erosion rate moving in the downstream direction on the Mainstem Nooksack. In the forks, lateral erosion was greater in unconfined reaches and immediately downstream of confined reaches. Lateral erosion rates are higher and more erratic in steeper reaches. Lateral erosion rates may periodically increase or decrease, either locally or reach-wide, following changes in sediment influx, riparian vegetation conditions, or obstruction (jams) development.

3.4.3 Erosion Rates This report uses the erosion rates calculated by Collins and Sheikh, 2004 presented in Appendix A.

5/26/2015 - 13 - 3.5 Fluvial induced geotechnical failures (landslides) (Figure 3-5 Photo of fluvially induced landslide) Fluvial erosion induced landslides are a natural and common process of valley widening. Lateral channel migration can undercut and destabilize hillsides and slopes outside the floodplain as the channel erodes the toe of slopes. The failure of a slope or hillside into the channel can block or redirect the channel and exacerbate channel migration processes, especially avulsion. In addition to obvious additional hazards for those on the landslide or in the impact area below the slide, other hazards include landslide dams or landslide induced surge waves that may locally increase flood elevations and velocities. Landslides can happen abruptly, and cause catastrophic channel alteration instantly and have significant economic implications, such as destroying and bridges, rupturing pipelines, and destroying multiple dwellings and facilities, a factor that rapidly accentuates the risk. Other impacts of landslides are increased sediment supply and damage to fish and wildlife habitat. More detailed assessment of fluvial induced landslide potential in the Nooksack River near Deming are described in Appendix B.

3.6 Avulsion Hazard Zone (AHZ) (Figure 3-6 Example photo) Avulsions hazards are characterized by rapid channel location change with high water velocity and flow depths, flooding, erosion, and potential debris impacts. The process of channel flow will cause surface erosion (scour), lateral fluvial erosion, and sediment deposition. All identified “possible” avulsion hazards were mapped based on existing topographic information. No analysis was taken to assess whether the avulsions are an immediate hazard or long-term concern, nor were any detailed analyses performed indicating the relative potential for avulsion at a single site. Based on field observations, some avulsion channels were identified as warranting a detailed risk assessment as soon as possible. All areas of the HMZ that exhibit wandering channel morphology are considered avulsion hazards. Three types of avulsion processes are defined for this study: • Relict/secondary channel capture: Capture of a relict channel or secondary channels within the floodplain during flood events or lateral channel migration into a relict channel. Channels may temporarily convey water only during the flood, or may capture the entire channel flow abandoning the previously occupied channel. Chute and neck cut-off avulsions may also occur as migrate. Hydraulic modeling was used to identify channel forms that were inundated during flood events. 100-year discharges modeled by FEMA were consulted for the mainstem and South Fork, in addition to using more recent hydraulic modeling data. HEC-RAS Modeling performed by Hyatt (2005) was used on the North Fork Nooksack. Modeling was not used on the Middle Fork because the dynamic changes in channel bed elevation observed on a decadal time scale led to conditions that were not appropriate for standard modeling. It should be noted that using the hydraulic modeling available at the time of this study provides only an estimate of inundated areas and the

5/26/2015 - 14 - potential for cross-section changes, measurement error, and modeling errors exists. All historic channel locations (HMZ) are considered to be an avulsion hazard. Relict channels outside the HMZ and within the EHZ were considered to be an avulsion hazard. Channels inundated by a 100-year flood elevation, but outside the EHZ, were shown to be an avulsion hazard if they appeared continuous enough to convey flow.

• Depressional Areas/Swales: Topographical areas subject to rapid or instantaneous high flow velocities with potential debris. Low lying areas adjacent to actively eroding areas were considered avulsions hazards in the Middle Fork, Lower Mainstem, and Delta. The avulsion hazards on the Middle Fork are unique because rapid bed elevation changes may encourage substantial overland flow that may behave like channelized flow or evolve into a channel.

• Levee break () Areas adjacent to man-made levees and road fills that can breach and temporarily create channel-like hazard conditions. A levee break is not necessarily an avulsion by traditional definition, however the nature of the hazard fit best with avulsion hazard characteristics. Field evidence was found for recent levee breaches that suggested channel like conditions and associated hazards do occur where a occurs. These examples were used to estimate possible direct hazard impact areas. For mapping purposes, the generalized estimates of direct hazard impact areas are: 1) Levees elevated four feet or less above the existing floodplain were delineated with a direct hazard area of at least 200 feet. 2) Levees elevated between 4 and 8 feet above the floodplain elevation were delineated with hazard zones between 200 to 600 feet. 3) Levees in excess of 8 feet above the floodplain elevation were delineated with a hazard zone of at least 800 feet wide. Site-specific field investigation must be performed to properly delineate the levee break hazard area.

3.6.1 AHZ Assumptions Several assumptions were made to map potential channel avulsions: the channel bed elevation is somewhat stable over the long-term; however it is understood that short-term, localized of a channel bed may increase avulsion potential at a particular location. Field evidence suggests that long-term channel bed stability has historically existed in certain reaches, and that many reaches are currently incised, likely as a result of the disturbances associated with hydromodification and land clearing. In reaches where it was clearly not possible to ascertain long-term bed stability, or where historic short or long-term bed instability was evident, a more conservative approach to avulsion was taken.

5/26/2015 - 15 - Understanding trends in channel bed elevation, in addition to reliable hydraulic modeling and erosion potential is necessary to make more site-specific analyses. Avulsion risk has likely decreased with time as widespread channel incision has occurred in many reaches and access to side channels and relict channels on the floodplain has been greatly reduced. In addition, the increased channelization and decreased large woody debris jams have also lessened the potential for avulsions. How these factors change over the design period of the study will affect the risk of avulsion occurrence. Topographic lows and relict channels were identified based upon topographic information available at the time of this study. Field checks of topographic conditions were made in locations where immediate and potentially catastrophic impacts to existing infrastructure appeared possible. The observed relict channels and lows are shown on maps. These features may routinely fill with water during floods, and may even convey some flow during flood events, but these temporary conditions do not necessarily mean that they should be identified as an avulsion hazard, although this determination may change with time as the river migrates. With this concept in mind, the relict channels and lows that were identified as Potential Near-term Avulsion Hazard Risk Areas were identified. They were identified because they exhibited conditions that were interpreted to be suitable avulsion pathways within the near-term (within the next decade) and a detailed risk analysis should be conducted as soon as possible. To be considered a near-term avulsion hazard area, one or more of the following conditions needed to exist: • A continuous relict channel or topographic low that was inundated and capable of conveying significant flows for frequent floods • Close proximity to the current channel location (less than the maximum observe erosion for that reach) • Provide a “hydraulic energy benefit” or “short-cut”. Examples are a chute cut-off or a flow path that had an increase in slope and/or a decrease in roughness such that the channel would opt to take a new course and partially or fully abandon its previous course because it is “easier” from an energy standpoint. • Potential for main channel flow obstruction and backwatering, such as bridges or constrictions, log jams or other potential obstructions. The concept is that a downstream obstruction will increase the opportunity for new channels to seek courses around the obstruction.

All areas of the HMZ in wandering channel morphology reaches should be included as Near-term Avulsion Hazard Risk Potential areas.

3.6.2 Potential AHZ consequences An avulsion can drastically alter hazard areas. Should an avulsion occur, a reassessment of the hazard areas mapped for this analysis should occur, and maps should be updated. For this analysis, the Erosion Hazard Areas mapped for potential avulsions underestimates the erosion potential should an avulsion occur and be sustained for any length of time.

5/26/2015 - 16 - The avulsion hazard has more potential health/safety risk than lateral erosion because of the rapid nature of change, the greater extent of change, and the hazard characteristics. Where lateral erosion is somewhat manageable, slow and predictable, avulsions may be difficult to manage because they typically occur rapidly during flood events when responding is complicated by time, access and other field related conditions.

3.7 Alluvial Fan (discussion) Whatcom County Planning and Development Services and others have previously mapped alluvial fan locations. This information and the LiDAR was used to modify the boundaries of alluvial fans where applicable. The alluvial fan locations that are adjacent to or come into contact with the Nooksack River geomorphic floodplain or potential erosion hazard area are shown on the hazard maps. Inherent hazards of alluvial fans include hazards consistent with both EHA and AHZ, with an additional hazard potential including debris impact and sediment deposition. With the exception of showing the interpreted locations of alluvial fans on the hazards maps, site-specific analysis was not conducted.

5/26/2015 - 17 -

FIGURE 3-1: General geomorphic setting of the Nooksack River Basin.

5/26/2015 - 18 -

FIGURE 3-2: Geologic overview of study area (from Lapen, 2000)

5/26/2015 - 19 -

FIGURE 3-3: Late Holocene pre-avulsion Nooksack River course through the Sumas Valley to the Fraser River (Maudlin and Pittman, 2003). The avulsion node(s) is located near present day Everson. Overflow from the Nooksack River enters the Fraser Basin at this location during flooding events that exceed a 5- year interval.

5/26/2015 - 20 -

FIGURE 3-4: Reaches defined for the project area.

5/26/2015 - 21 -

FIGURE 3-5: Example of geotechnical process included within erosion hazard area. The area is locally referred to as the “ bank” and has exhibited close to a century of frequent slope instability resulting from lateral erosion of the by the Nooksack River. Two homes were lost from landslides in 2006, and another is currently at risk. Avulsion hazards are exacerbated from landslide runout which partially or fully blocks the channel (Pittman, 2007).

5/26/2015 - 22 -

4.0 Reach Discussions

4.1 Delta (RM 0 - 6.4) Mouth to Ferndale (Figure 4-1) 4.1.1 Characteristics (geomorphic summary) - Delta The upstream extent of the geomorphic Nooksack Delta begins at RM 6.4 and continues to the tidal zone below RM 0. The upstream extent of the modern delta with topographic and hydro-modifications is debatable, but would likely be either the Lummi River where historically active distributary channels existed, or downstream of the Marine Drive Bridge where modern distributary channels form. The boundary of the geomorphic delta was interpreted from LiDAR, geologic, and archeological information. Based on recent geomorphic and archeological interpretations, the Nooksack Delta was primarily formed in the mid to late Holocene (Hutchings, 2004). Three primary distributary channel forms are expressed topographically in addition to the numerous tertiary channel forms. Two of these distributary channels have been abandoned and isolated from all but extreme Nooksack flood events. The northern abandoned distributary channel near the town of Ferndale may have been abandoned long ago and is now isolated by levees and fill. However, the Lummi River distributary channel was the primary Nooksack River outlet until it was cutoff in the late 1800’s (Brown, et al, 2005). In the late 1800’s to early 1900’s extensive logjam existed throughout the delta area. One particularly noteworthy logjam was located at approximately RM 5 near the now abandoned Lummi River distributary channel (Brown, et al, 2005). The Lummi River distributary channel would likely still be active without the historic wood removals the subsequent construction of levees. Tidal influence is interpreted to extend upstream to approximately RM 4. Geology of the delta includes overbank , , small gravel, and organics. Based on interpretation of borings done for the Slater Road Bridge, the delta sediments are relatively shallow (less than 4 meters). The delta has changed dramatically in historical times and has prograded seaward well over one mile since the surveys in the late 1800s. Several extensive logjams on the delta were managed for navigational purposes throughout the 1800s and early 1900s (Collins and Sheikh, 2002). The present day active distributary channels are located below RM 2 (near Marine Drive) because the delta floodplain above RM 1.5 has been cutoff by levees and the modern active delta has prograded seaward. LIDAR elevation data shows that the floodplain in the active delta and adjacent to the Hovander/Slater Overflow area has aggraded while the former active areas of the delta now cutoff by levees appear to have subsided. Many of the subsided areas are at or below today’s tidal range, exacerbating flooding hazards and complicating land management. As the delta continues to prograde seaward, aggradation of the active delta is

5/26/2015 - 23 - expected to continue, whereas the detached, inactive delta may continue to subsided creating a compounding elevation difference. The relative difference between active delta elevation and subsided delta elevations will be exacerbated by the predicted rise associated with global warming. Based on this model, an increased potential for avulsion hazards from levee failures and flooding could be expected.

4.1.2 Lateral Erosion Methods - Delta The work by Collins and Sheikh (2004) demonstrated that the lateral erosion observed in the past century has been relatively slow in comparison to the upstream reaches of the Nooksack. The reasons for this include reduced gradient (lower stream velocities) and more cohesive bank materials. Using the historic channel location information was the basis for using the meander migration method. This method was further supported by the current channel morphology and relict channel morphology observed on topographic modeling. The general movement of the channel was historically dominated by avulsions. This process can be observed in the modern day active delta. Deposition of sediment and debris jams often block distributary channels, and floodwater would seek alternate routes. The active delta and distributary channels (south of Marine Driver) were shown entirely within the EHA because the existence of jams and sediment deposition increases the uncertainty of predicting channel change.

4.1.3 Avulsion Hazard Methods - Delta Deltas naturally exhibit distributary channels that frequently avulse. Historically, large volumes of wood debris exacerbated the occurrence of avulsion by obstructing distributary channels and increasing localized backwater flooding. Levees dominate the delta area presently. The presence of levees complicates the prediction of avulsion. Obvious relict channel forms were identified as having an avulsion potential, however catastrophic failure of a levee may be the catalyst for an avulsion at an unpredictable location, and topographic lows in-between levees have the potential carry water with considerable velocity. In most places of the delta, the floodplain elevation adjacent to the channel is considerably lower than the relict channels and natural levees. Predicting specific levee failures and flow paths in such a broad area of hazard potential has a high degree of uncertainty, therefore, avulsion “zones” resulting from levee breaks were mapped adjacent to the levees, in addition to identifying topographic lows that could abruptly convey water temporarily until the low areas become inundated.

5/26/2015 - 24 -

FIGURE 4-1: Lower Mainstem – Delta sub-reach

5/26/2015 - 25 - Lower Mainstem (RM 6.4 - 15) Ferndale to west Lynden (Figure 4-2)

4.2.1 Characteristics (geomorphic summary) – Ferndale to west Lynden The Nooksack Valley in this reach is narrow, averaging only about 2500 feet versus 9000 feet in the immediate upstream reach. The valley is bordered with unconsolidated glacial sediments that show little to no evidence of fluvial erosion from the Nooksack River. Several possibilities might explain this anomalous situation: 1) the avulsion from the Sumas Valley into the relict glacial trough at Everson occurred recently and channel migration has not yet encroached on the valley edges; 2) the valley in this reach was historically a ponded wetland area that may have limited flow velocities because of an inability to convey flow, thus the channel may have been unable to erode the glacial material; 3) deposition has buried evidence of valley widening such as arcuate erosional features. The theory of recent occupation of a relict glacial trough by avulsion is the preferred model used for this analysis. In addition to the lack of lateral erosion of the valley walls, the avulsion model is further supported by the lack of relict channels other geomorphic indicators showing a long- term channel migration history. Furthermore, the Holocene alluvial sediment deposition in the Lower Nooksack Valley is interpreted to be very shallow (personal observations).

4.2.2 Erosion Hazard Methods – Ferndale to west Lynden The channel in this reach was publicly managed for erosion locally beginning in the 1930s when the Work Progress Administration install brush mats. Hydromodification prior to this period may have existed, but were likely small and temporary impediments to channel migration. Historic channel location comparisons showed that the channels have changed very little since the 1938 air photo. To map the EHA, there was not enough geomorphic evidence and there were not enough years of record of unmodified banks to ascertain natural channel patterns and channel migration potential, therefore the more conservative average annual migration rate was used instead of the meander migration method. However, average annual erosion rates were biased to a lesser-than natural rate because the channel management that limited channel migration occurred so early in the photo record. Using a biased erosion rate was felt to be justified because of the lack of geomorphic indicators that demonstrated that lateral erosion was extensive even prior to it be managed. 4.2.3 Avulsion Hazard Methods – Ferndale to west Lynden The topographic data showed little evidence of relict channels outside the HMZ. The floodplain is used primarily for agriculture and

5/26/2015 - 26 - levees exist in portions this reach. The opportunity for a levee break is possible and a levee failure avulsion hazard zone was mapped. The zones were approximated based on limited topographic evidence and should be assessed on a site-specific basis if a more complete hazard or risk assessment is needed. It should also be noted that some of the floodplain areas adjacent to the channel are at elevations lower than the channel bed, and that should significant levee overtopping or a levee break occur, these low areas could convey significant amounts of water with velocity, however the possibility of a full scale avulsion at these locations seems remote.

5/26/2015 - 27 -

FIGURE 4-2: Lower Nooksack Ferndale-Lynden sub-reach

5/26/2015 - 28 -

4.3 Lower Mainstem (RM 15 to 20) Lynden (Figure 4-3)

4.3.1 Characteristics (geomorphic summary) The Lower Nooksack River Valley is a relict glacial trough. The Holocene floodplain in this reach is very broad, over 13,000 feet at its widest, and flanked by unconsolidated glacial deposits. Two significant tributaries enter this reach from the north; Fishtrap Creek and Bertrand Creek. The GLO survey shows that the floodplain in this reach was historically a large wetland complex with a large lake near present day Kamm Creek. Much of the floodplain area is at elevations lower than the natural levee elevations adjacent to the river. Based on the GLO survey and other early survey records, the historic channel pattern was meandering. As with the reach immediately downstream, there is little evidence that the channel migrated into the valley walls except during historic times and relict channel morphology outside of the historic record is lacking, so a comparison of pre-disturbance morphology and migration patterns is hypothetical. The present channel is contained within levees and revetments in a meandering configuration that has been maintained since the 1930s. The channel pattern that would exist without the levees or revetments is uncertain.

4.3.2 Lateral Erosion Methods Prior to large-scale hydromodification and river management, the channel in the early 1900’s appears to have been migrating at accelerated rates in response to disturbances in the late 1800’s and early portion of the 1900’s. The channel migration patterns observed historically and the uncertainty of how the channel would behave if the levees and revetments were removed is the justification for using the wandering channel erosion methods. .

4.3.3 Avulsion Hazard Methods The topographic data showed little evidence of relict channels outside of the HMZ. The HMZ therefore makes up a majority of the channel capture avulsion hazard zone with a few exceptions; most notably Scott and the southern Guide Meridian dry bridge area. As with other sub-reaches within the Lower Nooksack Valley, the floodplain is used primarily for agriculture and extensive levees exist in much of this reach. The opportunity for a levee break is possible and a levee failure avulsion hazard zone was mapped. The zones were approximated based on limited topographic evidence and should be assessed on a site-specific basis if a more complete hazard or risk assessment is needed. It should also be noted that some of the floodplain areas adjacent to the channel are at elevations lower than the channel bed, and that should significant levee overtopping or a levee break occur, these low areas could and do convey significant

5/26/2015 - 29 - amounts of water with velocity. The probability of an avulsion into Scott Ditch was not assessed, however the likelihood would be increased in the presence of significant channel bed aggradation, massive levee failure, catastrophic flood, channel jam (either ice or wood) diverting flood flows, or any combination thereof.

5/26/2015 - 30 -

FIGURE 4-3: Lower Nooksack Lynden sub-reach

5/26/2015 - 31 -

4.4 Upper Mainstem (RM 20 to 23.5) east Lynden to Everson

4.4.1 Characteristics (geomorphic summary) The cutoff between “lower” Mainstem Nooksack River channel migration patterns and the “upper” Mainstem Nooksack channel migration patterns occurs at approximately RM 20. Typically the upper and lower Mainstem dividing point is Everson, however for the purpose of delineating channel migration patterns, the break occurs downstream of Everson. The sub-reach is within the Lower Nooksack River Valley, which is a relict glacial trough. The Holocene floodplain in this reach is not as broad as the immediate downstream reach and averages about 8,000 feet wide. The floodplain in this location flanked by unconsolidated glacial deposits, primarily glacial outwash. Similar to much of the Lower Nooksack Valley, portions of the floodplain area is at elevations lower than the natural levee elevations adjacent to the river. The theory of the major, late Holocene avulsion is supported in this reach by morphology that suggests that a sediment wedge was migrating westward from the point of avulsion (at Everson) creating an extension of anastomosing channel form into a valley that was previously a massive wetland complex. The valley cross-sections show the merging of two different valley morphologies. Based on the GLO survey and other early survey records, the historic channel pattern was also primarily anastomosing. Unlike the reaches immediately downstream, there is relict channel evidence that suggests the channel migrated across the valley and began to encroach upon some of the valley walls. The present channel is contained within levees and revetments in a relatively straight, slightly meandering type configuration that has been maintained since the 1950s. The current channel is very confined from its historic channel form.

4.4.2 Lateral Erosion Methods The historic migration pattern and geomorphic channel form of the river prior to large-scale hydromodification suggest that the lateral erosion potential in this reach is much greater than in the lower reaches. The average annual erosion rates measured by Collins et. al also reflect this. The erosion rates in the early 1900’s may have been accelerated in response to disturbances in the late 1800’s and early portion of the 1900’s. The river was fairly contained with levees and revetments by the 1940s and 1950s. The channel configuration that exists today is not conducive to the past geomorphic channel form or the channel form observed early in the historic record. It is likely that migration patterns would behave like the unconfined reaches upstream of Everson with large lateral erosion rates and a low-order braiding

5/26/2015 - 32 - morphology if the levees and revetments were removed. Based on the historic and geomorphic channel morphology and migration pattern, the wandering channel erosion method seems most applicable. The EHA of this reach is substantially wider than the immediate downstream reach.

4.4.3 Avulsion Hazard Methods The topographic data showed that most of the relict channels were inside the HMZ, however some channel morphology was observed. While the HMZ makes up a majority of the channel capture avulsion hazard zone, there are exceptions; most notably Scott Ditch. The floodplain in this reach is used primarily for agriculture and some levees exist, primarily in the lower portion of this reach. The opportunity for a levee break is possible and a levee failure avulsion hazard zone was mapped. The levee break AHZ were approximated based on limited topographic evidence and should be assessed on a site-specific basis if a more complete hazard or risk assessment is needed. It should also be noted that some of the floodplain areas adjacent to the channel are at elevations lower than the channel bed, and that should significant levee overtopping or a levee break occur, these low areas could and do convey significant amounts of water with velocity. The probability of an avulsion into Scott Ditch was not assessed, however the likelihood would be increased in the presence of significant channel bed aggradation, massive levee failure, catastrophic flood, channel jam (either ice or wood) diverting flood flows, or any combination thereof.

5/26/2015 - 33 -

FIGURE 4-4: Lynden to Everson sub-reach

5/26/2015 - 34 -

4.5 Upper Mainstem Everson to Cedarville (Nugent’s Corner) (RM 23.5 to 31)

4.5.1 Characteristics (geomorphic summary) The Everson to Nugent’s Corner reach is unique in that it is a split in the Holocene geomorphic floodplain, and currently the single valley (Sumas Valley) contains two separate watersheds – the Nooksack and the Fraser, that are separated only by river bank elevations below a 10-year flood stage. Overbank flooding in this reach leaves the Nooksack watershed and enters the Fraser watershed. Between 1990 and 2004, this occurred four times, most seriously in 1990. The geomorphology at this location indicates that the channel recently occupied both sides of the Holocene valley, which is approximately 2-miles wide at this location. The channel morphology based upon relict channels and valley cross sections indicates that this reach was primarily anastomosing, however some evidence for meandering morphology does exist in the Sumas Valley, especially in the east side of the valley, increasing as you move down valley (north). Exactly when the channel avulsed from the Sumas Valley into the Lower Nooksack Valley is not known, but based upon interpretation of archeological sites and morphological interpretation, it occurred in the mid to late Holocene and could have been as late as a few hundred years ago. The valley walls in this reach consist of Pleistocene glacial and Holocene deposits (colluvium and alluvium). Both materials erode easily by fluvial processes, as can be evidenced by numerous arcuate erosional scarps on both sides of the valley and modern observed . The channel is currently situated in the west side of the valley and has occasionally made contact with the Pleistocene glacial deposits in the past century. The right bank of the channel is bound only by Holocene alluvium and erodes very easily, as evidenced by the high annual erosion rates. The eastward lateral migration potential for this reach is very high. This potential is increased by the possibility of avulsions. The potential of an avulsion into the Fraser watershed was not necessarily assessed in this analysis, however because the channel is currently entrenched, primary over-flow paths blocked by infrastructure, and channel revetments used to limit lateral migration, the likelihood for this avulsion occurring under “normal” conditions is probably very low. The avulsion potential may increase with the occurrence of other natural hazards and processes, such as a lahar.

4.5.2 Lateral Erosion Methods The historic migration pattern and geomorphic channel form of the river prior to large-scale hydromodification suggest that the lateral

5/26/2015 - 35 - erosion potential in this reach is substantial. The erosion rates in the early 1900’s may have been accelerated in response to disturbances in the late 1800’s and early portion of the 1900’s, although modern erosion rates are also very high. The floodplain is currently not levied in this reach, however several former active channels have been isolated by agricultural levees. The general lack of levees, and rare occurrence of floodplain flooding, in addition to other geomorphic indicators, suggests that this reach has incised considerably from its pre-disturbed configuration. Whether this incision was initiated by a recent avulsion from the Sumas Valley to the Lower Nooksack Valley is unknown. The right bank floodplain is in essence an abandoned flood terrace for events less than the 100-year flood. Historic placement of revetments began in the 1940s and 1950s. In local places within this reach, the channel configuration is very confined. It is likely that migration patterns would behave like the unconfined reaches with a larger migration area and a low-order braiding to evolving anastomosing morphology if the and revetments were removed. Based on the historic and geomorphic channel morphology and migration pattern, the wandering channel erosion method was used.

4.5.3 Avulsion Hazard Methods The topographic data showed that significant relict channels existed outside the HMZ, some relict channel morphology was observed within the adjacent Sumas/Fraser basin. The geomorphic floodplain in this reach is used primarily for agriculture and few levees exist, likely because the reach is fairly incised and seldom accesses the geomorphic floodplain. To further complicate mapping the AHZ in this reach is the understanding the potential for an avulsion from the Nooksack valley into the Sumas valley. If mapping were being done on unaltered topography, the AHZ would include the Sumas Valley. However, significant bank armoring, incised channel conditions, and infrastructure limit the avulsion hazard potential. Because an avulsion in this location would cause significant damage, it is likely that considerable effort would be undertaken to prevent such a condition from occurring. However, some conditions may or will exist in which engineered response may be impractical or impossible. Such scenarios may be or other volcanic events, major landslides, catastrophic flooding (greater than 500-year event or outburst flooding), channel jamming or blockages, or significant tectonic disturbances. The recurrence interval of these events is very low, however each has occurred at least once in the past 10,000 years. Analysis for this study was based on limited topographic evidence and should be assessed on a site-specific basis if a more complete hazard or risk assessment is needed.

5/26/2015 - 36 -

FIGURE 4-5: Upper Mainstem Nooksack River Everson sub-reach

5/26/2015 - 37 -

4.6 Upper Mainstem Nugent’s Corner to Deming (confluence) (RM 31 to 37)

4.6.1 Characteristics (geomorphic summary) The Nugent’s Corner to Deming reach is the lowest-most geologically confined reach. Historic mapping and geomorphic interpretation demonstrate that the Nooksack River occupied both side of the valley. Overbank flooding in this reach typically would re-enter the Nooksack watershed however catastrophic flooding and the migration of the channel to the north side of the valley would increase the potential for water to enter the Fraser watershed via the Sumas Valley. The Holocene valley is less than 1-mile wide in this reach. The channel morphology, based upon historic mapping, relict channels and valley cross sections, indicates that this reach was primarily an anastomosing channel form. The valley walls in this reach consist of bedrock, Pleistocene glacial deposits, and Holocene deposits (colluvium and alluvium). The Pleistocene and Holocene deposits erode easily by fluvial processes, as can be evidenced by numerous arcuate erosional scarps on both sides of the valley and modern observed bank erosion. The channel is currently situated in the south side of the valley and occasionally makes contact with the Pleistocene glacial deposits and bedrock geology. The right bank of the channel is bound only by Holocene alluvium and erodes very easily, as evidenced by the high annual erosion rates. The northward lateral migration potential for this reach is very high. This potential is increased by the possibility of avulsions. Numerous lahars have traveled down the Nooksack in the Holocene (Kovanen et.al., 2001). Deposits of a lahar originating in the Middle Fork Nooksack drainage traveled at least as far as Nugent’s Corner and may have deposited material several feet in thickness in the channel bed (Pringle et. al., 2001). An event of this magnitude or larger would have serious implications for flood hazards for this valley confined reach. Major infrastructure within the reach includes the Mount Baker High School, Nooksack Casino, Mount Baker , and high-pressure gas pipelines.

4.6.2 Lateral Erosion Methods The Nugent’s Corner to Deming Reach is very similar to the Everson to Nugent’s Corner Reach in its lateral migration patterns. Portions of the floodplain are currently levied in this reach, and several former active channels have been isolated by agricultural levees. The right bank floodplain is inundated in events less than the 100-year flood. Historic placement of revetments began in the 1940s and 1950s. In many places within this reach, the channel configuration that exists today is very confined. It is likely that migration patterns would

5/26/2015 - 38 - behave like the unconfined reaches immediately upstream of Everson with a larger migration lateral and a low-order braiding to evolving anastomosing morphology if the and revetments were removed. Based on the historic and geomorphic channel morphology and migration pattern, the wandering channel erosion method was used. A significant fluvially destabilized landslide potential exists within this reach (RMs 32 to 34). The river has significantly eroded the toe of a slope composed of unconsolidated Pleistocene glacial material. Assessment of erosion rates, slope retreat rates, slope failure mechanisms, repose angles, and future channel configuration scenarios was conducted for as part of a river management plan concept initiated by the Whatcom County Flood Control Zone District and Whatcom County Public Works (Appendix B). A risk analysis was not formally conducted in this assessment, however one is highly recommended because of the hazard potential. Based on the assessment that was conducted, future slope failures are believed to be immanent. Not known is the potential size and likelihood of more catastrophic failures. Even small landslides can impact channel migration patterns at this location. A large slope failure could cause extreme lateral migration, and especially channel avulsions as described below.

4.6.3 Avulsion Hazard Methods The geomorphic floodplain in this area is confined by glacial deposits, Holocene deposits or bedrock valley walls. The HMZ spanned valley wide in early mapping of this reach. Topographic data showed that relict channels existed outside the HMZ also exist. The geomorphic floodplain in the narrow valley of this reach is used for agriculture, residential, and dense rural development in addition to significant transportation infrastructure and levees do exist. Avulsion hazards outside the HMZ in this reach are low except for at the downstream portion of this reach near Nugent’s Corner. Where avulsion hazards do exist, the risk is exacerbated by landslide potential and constrictions (bridges and revetments), resulting in a reach that has potentially high avulsion risks in a localized area. Indeed, historic temporary avulsions did occur in this area. The risk of a channel avulsion into the Sumas Valley in this reach is very low because of the distance to the drainage divide and infrastructure, especially the elevated railroad grade, that would be difficult for a new channel to erode through. However, an avulsion at the lower end of the reach could cause significant damage to the Mount Baker Highway because avulsion paths cross this roadway. As with the Everson to Nugent’s Corner Reach, because an avulsion in this location would cause significant damage, it is likely that considerable effort would be undertaken to prevent such a condition from occurring. However, some conditions may or will exist in which engineered response may

5/26/2015 - 39 - be impractical or impossible. As stated previously, such scenarios include lahars or other volcanic events, major landslides, catastrophic flooding (greater than 500-year event or outburst flooding), channel jamming or blockages, or significant tectonic disturbances. The recurrence interval of these events is very low, however each has occurred at least once in the past 10,000 years. Analysis for this study was based on limited topographic evidence and should be assessed on a site-specific basis if a more complete hazard or risk assessment is needed.

5/26/2015 - 40 -

FIGURE 4-6: Upper Mainstem Nooksack River Deming sub-reach

5/26/2015 - 41 -

4.7 North Fork (Glacier Creek to Confluence: RM 37-58)

4.7.1 Characteristics (geomorphic summary) – Figure 4-7 The North Fork Nooksack River reach is located within the steep, relatively narrow remnant Pleistocene glacial valleys of the western Cascade Mountains. Three sub-reaches with different geomorphic conditions have been identified: • the Glacier Reach, upstream of Kendall (RM 49) to the upstream limit of the analysis at Glacier (RM 58) (Figure 4-7A); • the Kendall Reach upstream of the Middle Fork Confluence (RM 40.5) to Kendall (RM 49) (Figure 4-7B); and • the Welcome Reach, upstream of South Fork Confluence (RM 37) to the Middle Fork Confluence (RM 40.5) (Figure 4-7C). The Glacier Reach is located within a narrow valley that was occupied by recent alpine glaciation in the latest Pleistocene/early Holocene (Kovanen, 2001). The valley in this reach is narrow, incised, and flanked by bedrock and -slope processes that limit, and occasionally confine, channel migration. Several Holocene landslides, including the Church Mountain Mega-Landslide, in this reach limit and constrict channel migration (Van Siclen, 1994). The floodplain is limited primarily to the historic and geomorphic channel migration zone. The erosive potential of the channel as observed in the historic record provides evidence that the river is capable of migrating across the entire valley bottom within a very short time. Further support of valley wide channel migration is that only localized remnant pockets of glacial deposits are found at the margins of the geomorphic channel migration area, indicating that Holocene migration has eroded away much of the Pleistocene deposits, except in the lowest most portions of the reach. In contrast, the modern North Fork channel and floodplain in the Kendall Reach is slightly incised into unconsolidated remnant glacial outwash deposits that are not considered significant limitations to lateral channel migration. The river in the Kendall reach seldom encounters the bedrock valley walls. Although modern erosion rates would suggest that the North Fork River in the Kendall has the potential to have eroded across its entire valley and through the glacial deposits within a relatively short period of time, in general this has not occurred. In part, this may be because the valley width in this reach is considerably wider than the Glacier Reach the North Fork here enters the Paradise Valley, a remnant glacial trough associated with Fraser Valley glaciation. However, the observed erosion rates do not correlate well with the relatively narrow observed geomorphic migration

5/26/2015 - 42 - area flanked by unconsolidated glacial deposits, especially between RM 43 and 41. It is not fully understood why lateral channel migration of the North Fork was relatively limited over the past 10,000 years. The active floodplain of the North Fork is relatively small, and today primarily occupies an area coincident with the historic migration area. The historic data suggests that the migration area has been expanding and that the Kendall reach is very capable of occupying a larger percentage of the modern Nooksack River floodplain/valley bottom. Three primary hypotheses are proposed to explain the inconsistencies with the interpreted pre-historic migration patterns and observed historic patterns: 1) overall Holocene trend of the river in the Kendall Reach either incision or net channel bed elevation stability and that the channel was primarily a transport reach between RM 45 and 41, or 2) pre-disturbance erosion rates were relatively low compared to the rates observed in the past century because of sediment and bank stability, or 3) hydrologic changes, either as a result of land use alterations or climate variability/changes, have increased the magnitude of stream events. The first hypothesis assumes that excessive lateral channel migration is primarily sediment load dependant, meaning that if the channel is efficient at transporting all or more than the input, then lateral migration is reduced. If this is true, it means that in the long- term, the North Fork has been an efficient transport reach capable of transporting at least all the sediment that is delivered to it over time. It should be noted that even in pre-disturbed conditions, the sediment loading of the North Fork was likely quite high because of glaciers, landslides, steep slopes of the mountain sides and tributary valleys; however modern sediment rates have likely only increased from pre- disturbed conditions. The increased sediment delivery may be causing the lateral erosion rates we have observed this past century. The second hypothesis lends credence to the abilities of wood debris and root cohesion provided by natural /riparian area forests and forested islands to slow lateral erosion rates. The observed lateral erosion rates we have observed this century may have resulted from the clearing and modification of these areas, and the corresponding loss bank strength and channel roughness. The third hypothesis is that hydrologic changes have increased the frequency and magnitude of larger discharge events. Plotting of maximum discharges of the North Fork Nooksack River at the USGS gage near Glacier demonstrate that a large majority of the ten largest floods of record have occurred in the past two decades (Tim Hyatt, personal communication). At this point, it is uncertain as to what the observed information demonstrates for the future or long- term evolution of channel geomorphology, meaning are the recent hydrograph changes part of a long-term change or are the changes part of the variability of shorter-term hydrologic/climatic cycles, and

5/26/2015 - 43 - what would the channel response be for either condition. Longer-term climate forecasts do predict higher snow level elevations (Dan Moore, UBC [email protected]). The influences that increased snow level conditions may have on a hydrograph are unknown. Speculatively, the result could be a “flashier” hydrograph with a steeper hydrographic curve resulting from a loss water of storage in snow at lower elevations. What the impacts of a flashier hydrograph on channel morphology are also unknown. It is possible that components of multiple or all these hypotheses are true, that the North Fork as a whole was an effective transport reach but that disturbances in the past century have altered the system and erosion rates and lateral migration potential are greater now than in the past. An active Holocene was identified near Kendall in 2005 (Haugarud, 2005). The impacts of tectonic deformation and channel migration have not been fully explored. It is possible that some of the landslides in the North Fork reach are associated with movement along this fault. It is also possible that some of the terraces identified in this study are not Pleistocene in age as presumed, but rather Holocene terraces associated with uplift along the fault. Further work or analysis to answer these questions may require review of the interpretations and assumptions made in this analysis. The North Fork Nooksack River in the Welcome Reach differs from the upstream North Fork reaches in that it dramatically increases in: discharge, historic lateral migration rates, and geomorphic migration corridor widths. Most remnant glacial valley deposits and older terraces have been eroded and the terraces that do exist are interpreted to be Holocene based on the relict channels found on them. The increased discharge and sediment added to the channel by the Middle Fork directly influence channel migration in this reach. In addition, the downstream limit of the reach confines and limits channel migration with a bedrock on the north side of the valley and a Holocene landslide on the south side of the valley. In addition, infrastructure (State Highway 9 and Burlington Northern Railway) further constricts the channel. The Middle Fork lahar filled this valley approximately 6,000 years ago, and are occasionally uncovered by lateral erosion. The channel has migrated from valley margin to valley margin for much of this reach over the historic record. A majority of the geomorphic evidence that would help illustrate the pre-disturbed channel form has been obliterated by modern lateral erosion. Interpretation of the few remaining areas that do contain pre- disturbance geomorphic channel form, suggest that an anastomosing or braided channel form existed, at least locally. Anastomosing channel forms typically occurs in stable or slowly aggrading reaches, whereas braided channel forms can occur in areas of rapid deposition and poor bank strength. The channel forms of the past may have

5/26/2015 - 44 - varied dramatically depending upon changes in sediment input rates, amongst other factors, and may have alternated, both temporarily and spatially, between anastomosing, braided, or low amplitude meandering over time. Presently however, low to high order braiding dominates the modern channel form. The channel migration observed during the historic record demonstrates that the North Fork is very dynamic. Further evidence supporting dynamic conditions in the valley is found outside the historic record. The geomorphic history of the North Fork valley suggests that numerous geologic events (landslides, tectonic deformation, and lahars) outside of the “normal” events observed in the past century have occurred in the recent geologic past and are likely to occur again. The nature of the steep mountainous and the proximity to an active volcano presents numerous altering hazards. These events include, but are not limited to lahars, debris flows/floods, and glacial outburst floods, mega-landslides, and other volcanic and tectonic events. Because of the narrow valley width amongst other factors, these types of events can catastrophically impact the valley.

4.7.2 Lateral Erosion Methods The North Fork Nooksack River Valley is relatively narrow, and is locally incised into Pleistocene Glacial sediments, Holocene landslides or alluvial fans, or bedrock. The active migration patterns of the river as observed in the past 100 years would indicate that it has the potential to migrate, either by lateral erosion or by avulsion, across the entire active valley floodplain within a short period of time. The North Fork Reach is similar to the Everson to Nugent’s Corner Reach in its lateral migration patterns, except that the North Fork is dominantly braided and has larger observed maximum single event erosion potential. In places within this reach, the channel configuration that exists today is confined both by natural obstructions and infrastructure (bridges). These constrictions affect lateral migration patterns upstream and downstream of the constriction. Most revetments located on the North Fork are at the fringe of the active floodplain. Many of the revetments constructed over the pas half century have been lost to lateral migration or are routinely damaged because of the high energy of the river. Based on the historic and geomorphic channel morphology and observed migration patterns, the wandering channel erosion method was used. It is not known whether the channel is still actively incising or if a cycle of channel aggradation is occurring. Because of the conservative nature of this analysis, an aggrading channel is presumed, which would suggest that the active floodplain may be increasing in size. Fluvially-destabilized slopes with landslide potential may exist within the North Fork Reach, especially in the reach upstream of

5/26/2015 - 45 - Kendall, however none were identified based on the course resolution of this assessment. A detailed geotechnical risk analysis of potential failures adjacent to the channel may be warranted. Not known is the potential size and likelihood of catastrophic deep-seated landslide failures.

4.7.2.1 Comparisons of potential erosion rates The Washington State Department of Transportation “North Fork Nooksack River Corridor Analysis,” prepared by GeoEngineers in 2001, shows an anticipated 50-year erosion hazard zone. For the most part, we produced similar results. Appendix C provides justification for areas in which discrepancies between the two erosion hazard zones exist.

4.7.3 Avulsion Hazard Methodologies Predicting avulsion hazards on the North Fork was done by mapping the relict channel locations on the terraces and floodplain and then using hydraulic modeling to determine potential inundation of channels in a frequently recurring flood (10-year event). This method assumes that the channel bed elevation is relatively stable. Several observed relict channels and obvious topographic paths that could support an avulsion outside of the HMZ were identified. Because of the rapid channel migration patterns, avulsions within the HMZ are very likely and have occurred numerous times in the historic record. Although long-term overall channel bed stability was assumed, abrupt and dramatic bed elevations changes could locally elevate the channel to the surface of terraces, which could convey flow in topographically low areas. All mapped avulsion hazard zones were within the erosion hazard area.

5/26/2015 - 46 -

FIGURE 4-7: North Fork Nooksack Reach

5/26/2015 - 47 -

FIGURE 4-7A: Kendall-Glacier Sub-reach

5/26/2015 - 48 -

FIGURE 4-7B: Welcome – Kendall Sub-reach

5/26/2015 - 49 -

FIGURE 4-7C: Confluence – Welcome Sub-reach

5/26/2015 - 50 -

4.8 Middle Fork (Mosquito Lake Road to Confluence: RM 0 to 5) (Figure 4-8)

4.8.1 Characteristics (geomorphic summary) The geomorphology of the Middle Fork most closely resembles the geomorphology of the North Fork Nooksack River upstream of Kendal. The valley gradient is steep and confined, and tributaries originating in steep, high altitude basins feed the river. Sediment is also abundant in the Middle Fork drainage, both because of the terrain, and because of the presence of glaciers. Additionally, at least one, likely more, Holocene era lahars have flowed down the Middle Fork Nooksack, the most significant lahar event was approximately 6,000 years ago. The Middle Fork valley is very dynamic and not well understood. Obvious Holocene era terraces with preserved channel forms are lacking, however, the lahar deposit still flanks the margin of the active Middle Fork floodplain and channel. As with the interpretations made with the North Fork, it seems that long-term net bed elevation changes appear stable, however we observe rapid lateral and vertical changes in the historic record. Recent vertical bed elevation changes of several meters have been observed occurring on a decadal scale. The channel form, as with the North Fork, is predominantly braided, although some stable islands do exist, suggesting that anastomosing channel form is possible. Little historic accounts of the channel and any jams were found. Because of the steep gradient and confined channel, large channel spanning jams may have been limited. The average valley width to HMZ ratio is extremely low when compared to the rest of the Nooksack River, with exception to of portions of the North Fork. Lateral erosion is frequently controlled by geologic features, either bedrock or lahar deposits. In the historic record, the channel was situated primarily in the western portion of the valley. Recent migration has been occurring in the eastward direction and has expanded HMZ in several locations. Erosion rates in the alluvial material are very high. Erosional rates of the lahar could not be measured, but they are estimated to be moderately low because of demonstrated erosion resistance. In many locations, the lahar confines and constricts channel migration. The dynamic geomorphic history of the Middle Fork valley suggests that geologic events outside of the “normal” events observed in the past century have the potential to occur again. The nature of the steep mountainous terrain and the proximity to an active volcano presents numerous landscape altering hazards. These events include, but are not limited to lahars, debris flows/floods, dam and glacial outburst floods, mega-landslides, and other volcanic and tectonic events. Because of the narrow valley width amongst other factors, these types of events can catastrophically impact the valley.

5/26/2015 - 51 -

4.8.2 Lateral Erosion Methods The Middle Fork Nooksack River Valley is relatively narrow, and is locally incised into either a Holocene lahar deposit, Holocene landslides or alluvial fans, or bedrock. The active migration patterns of the river as observed in the past 100 years would indicate that it has the potential to migrate, either by lateral erosion or by avulsion, across the entire active valley floodplain within a short period of time. The Middle Fork Reach is similar to the North Fork in its lateral migration patterns and similarly has relatively large observed maximum single event erosion potential. In places within this reach, the channel configuration that exists today is confined both by natural obstructions and revetments. These constrictions affect lateral migration patterns upstream and downstream of the constriction. Most revetments located on the Middle Fork are at the fringe of the active floodplain. Based on the historic and geomorphic channel morphology and observed migration patterns, the wandering channel erosion method was used. As with the North Fork, it is not known whether the channel is still actively incising or if a cycle of channel aggradation is occurring. Because of the conservative nature of this analysis, an aggrading channel is presumed, which would suggest that the active floodplain might be increasing in size. Field and recent channel migration patterns support the model of an increasing migration area. Fluvially destabilized slopes with landslide potential may exist or could occur; especially on the west side of the valley, however no specific slides were identified in this assessment. A detailed geotechnical risk analysis of potential failures adjacent to the channel may be warranted if any slides are identified.

4.8.3 Avulsion Hazard Methods Predicting avulsion hazards on the Middle Fork provide a unique challenge when compared to the rest of the Nooksack River. Lacking are easily observed relict channels and obvious topographic paths that would support an avulsion outside of the HMZ. However, abrupt and dramatic bed elevations changes could potentially elevate the channel locally to the surface of the lahar deposit, which could convey flow in topographically low areas. Several areas that were topographically low were identified as potential avulsion hazard areas because increasing bed elevations could exacerbate this hazard. Because of the dynamic bed elevation changes and rapid channel migration, avulsions within the HMZ are very likely and have occurred numerous times in the historic record. All mapped avulsion hazard zones were within the erosion hazard area.

5/26/2015 - 52 -

FIGURE 4-8: Middle Fork Nooksack

5/26/2015 - 53 -

4.9 South Fork (Skookum Creek to Confluence: RM 0 to 15) 4.9.2 Characteristics (geomorphic summary) The South Fork Nooksack River reach is located within remnant Pleistocene glacial valleys. The valley walls in the South Fork Reach consist of bedrock, Pleistocene glacial deposits, and Holocene deposits (colluvium and alluvium). The Pleistocene and Holocene deposits erode easily by fluvial processes, as can be evidenced by numerous arcuate erosional scarps on both sides of the valley and modern observed bank erosion. The channel is currently situated in the north side of the valley above Acme and on the west side of the valley below Acme, and occasionally makes contact with the Pleistocene glacial deposits, Holocene deposits and bedrock geology. The right bank of the channel below Hutchinson Creek near Acme is bound only by Holocene alluvium that can erode easily, as evidenced by the high annual erosion rates in areas with no revetments. The eastward lateral migration potential for this portion of the river is very high. This potential is increased by the possibility of avulsions. The pre-historic South Fork geomorphology varied substantially and has been significantly altered in the past century. Three geomorphic sub- reach were identified based upon topographic interpretation: • Skookum Creek to Acme (RM 8 to 15) (Figure 4-9A), • Acme to Van Zandt (RM 2 to 8) (Figure 4-9B), and • Van Zandt to the confluence (RM 0 to 2) (Figure 4-9C). Bedrock and relict glacial deposits flank the South Fork valley in the Skookum Creek to Acme reach. The South Fork in historic, or in recent geological past has had contact with both north and south valley walls and is presently flowing along the northern valley edge. The channel above Skookum Creek is confined by a bedrock constriction (Dye’s ) and a sharp gradient change near the Saxon Bridge occurs where the channel transitions from a transport reach to a depositional reach. The historic channel form in this reach was likely dominantly a single thread, with a low amplitude channel meandering form upstream of the Saxon Bridge. Because of the valley confinement and steep gradient in this area, fluvial energy is fairly high and likely dominates channel-forming processes, meaning that sediment and large woody debris were likely stored only temporarily before they were transported downstream. Downstream of the Saxon Bridge the valley widens and the gradient reduces. Geomorphic and historic information show that the channel form evolves into an alternating anastomosing/braided channel form. This channel form is common where there are higher rates of sediment deposition and longer-term storage of large woody debris. Wide historic channel migration and channel instability in this area demonstrate the rapid lateral and vertical migration associated with prolific sediment deposition and avulsions common with anastomosing channel form.

5/26/2015 - 54 - Pre-historically, wood jams and intact riparian forests likely added to channel stability. Historic alteration of wood jams in this reach has caused dramatic channel changes. Rapid channel bed elevation changes are also evident in this reach, based both upon field observations and anecdotal information. The lower end of the Skookum to Acme sub-reach transitions from a relatively narrow valley into a larger remnant glacial valley. The South Fork is primarily under fit within this larger glacial valley that once contained a large river that flowed from north to south and out the present day Samish River valley. Just upstream of Acme, the Samish Valley and the South Fork valley floodplains merge. Historically, the South Fork flooded over this very subtle drainage divide and floodwaters entered the Samish River (personal comm. Bob Knudson). Historic mapping and geomorphic interpretation suggests that the South Fork Nooksack River has occupied both side of the valley in the reaches between Saxon and approximately Acme, and Van Zandt to the confluence. The topographic information shows a complex network of numerous relict channels demonstrating that the pre-historic channel form just upstream of Acme was strongly anastomosing. The presence of so many tertiary anastomosing channel branches may be the result of extensive logjams. The present channel form has been altered into a single thread channel that, based upon field observations and modeled flood inundation, has incised considerably such that it has abandoned the relict channels substantially enough that even considerable floods typically don’t inundated these relict channels let alone the floodplain. The South Fork in the Acme to Van Zandt sub-reach is under fit within the remnant Pleistocene glacial valley as demonstrated by both the average valley width to HMZ ratio and the valley width to geomorphic (relict channel) migration zone ratio. The Holocene valley is approximately 8,000-feet wide in this sub-reach. The channel is situated on the west side of the valley. Relict channels are numerous adjacent to the present channel location but lacking on the east side of the valley except at the downstream end of the sub-reach near Van Zandt. The relict channel forms were mapped and used to identify the geomorphic migration zone, which is an area that was occupied by channels over some longer period of time as evidence by the geomorphic record. The geomorphic migration zone was used to compare to the EHA and AHZ to evaluate the concept of a “maximum” migration corridor width needed to support natural channel migration processes over an extended period of time. The Black Slough sub- basin located on the east side of the valley in this sub-reach was mapped as a vast wetland/lake complex in the late 1880s. Presently, the Black Slough channel and ditch network have replaced this feature. In cross-sections of the valley, the Black Slough sub-basin is lower than the South Fork. Overbank flooding in this reach between Acme and Van Zandt in pre-historic conditions would likely have flooded into

5/26/2015 - 55 - the Black Slough topographic low, and if fact there is some evidence that the two basins were connected by some channel form downstream of Acme (GLO survey, ~1885). Presently, small recurrence interval flooding from the South Fork now only enters the Black Slough sub- basin at the downstream most end of the sub-reach, however larger or catastrophic flooding could likely distribute large quantities of water into this sub-basin. Lateral channel migration or failure of levees would increase the potential for floodwater to travel northeast (the downward gradient in the Acme to Van Zandt sub-reach) and from the South Fork channel into Black Slough sub-basin. The pre-historic floodwater storage capacity in this area was likely very large. The channel morphology, based upon historic mapping, relict channels and valley cross sections, indicates that this reach contained both anastomosing and meandering channel form, and was evolved into a dominantly meandering channel form in the downstream portion of the sub-reach. The South Fork channel morphology in the lowest sub-reach, Van Zandt to the confluence, is influence by the constriction of the channel and floodplain by the late Holocene landslide located at the confluence. The gradient is low in this area and the floodplain between Van Zandt and the upstream end of the landslide is inundated more frequently by lower recurrence interval flooding relative to upstream reaches. The relict and historic channel form exhibits both meandering and anastomosing characteristics. Alluvial-formed erosional features cut into the Holocene alluvial fan deposits along the margin of the valley show large radius arcuate form likely created by long-term lateral meander migration of a single dominant channel; however early historic mapping demonstrates multiple channels did exist. Overall the South Fork channel has been modified from an anastomosing, meandering system into single thread morphology. The transition to single-thread morphology results in a “larger river” based on channel width, decreased , increased flow capacity in addition to many other changes. The instream modifications of /levee construction and removal of instream roughness (wood jams) made the South Fork more efficient and transporting water and sediment. The result is that the river incised into its floodplain throughout most of the reach. The interpretations made from air-photograph, revetment mapping, and geomorphic evidence support the concept that single thread system has been maintained by revetments, levees, reach-wide channel incision, loss of bank stability via root-strength cohesion decrease, and loss of stable wood jams. These changes alter the fluvial response from the pre-historic geomorphic observations and the early historic observations. Since the channel is maintained within its banks for most flood events (less than 10-year), the depth of flows and the velocities are increased and thus more lateral erosion potential exists. However, because of the disconnect with the floodplain and relict channels and the decrease in

5/26/2015 - 56 - stable wood debris jams, avulsion potential has been dramatically decreased from the pre-modified system.

4.9.3 Lateral Erosion Methods The South Fork Reach is most similar to the Everson to Nugent’s Corner Reach in its lateral migration patterns. In many places within this reach, the channel configuration that exists today is very confined, both because of revetments and because of recent channel incision. The assumption was made that migration patterns in controlled reaches would behave like the unconfined reaches immediately upstream or downstream of the confined reach. The present channel form is a low amplitude meandering river with some low-order braiding that has the potential to evolve into an anastomosing morphology if the revetments were removed and the lateral channel migration area increased. Based on the historic and geomorphic channel morphology and observed migration patterns, the wandering channel erosion method was used. It is not known whether the channel is still actively incising or if the cycle of channel aggradation has once again resumed.

A potential fluvially destabilized slope with landslide potential exists downstream of Acme at RM X (near McCarty Creek). The river has eroded the toe of a slope composed of unconsolidated sediment deposits. Assessment of erosion rates, slope retreat rates, slope failure mechanisms, repose angles, and future channel configuration scenarios was not conducted for this analysis nor was a risk analysis formally conducted, however one may be warranted because of the hazard potential. Not known is the potential size and likelihood of catastrophic failures. Even small landslides can impact channel migration patterns at this location. A large slope failure could cause extreme lateral migration, and especially channel avulsions as described below.

4.9.4 Avulsion Hazard Methods The South Fork reach was pre-historically and in early times an avulsion prone system because of the channel form, low gradient, and abundance of wood jams. The avulsion potential has been reduced by modern channel changes. Further reducing avulsion potential is infrastructure (roads, railroads, and levees). The most likely avulsion potential is meander cut-offs, several of which appear immanent in the near term. Other avulsion hazards do exist, which could be exacerbated by levee or road failures or lateral migration. The assumption stated in Avulsion Hazard Methods, were used to assess near-term avulsion hazard zones. Herrera (2005) conducted a more site specific avulsion hazard analysis for the Acme area. Bed elevation changes can alter the avulsion hazard potential over time. In addition,

5/26/2015 - 57 - catastrophic levee failures, especially upstream of a constriction, can instantly cause an avulsion hazard to exist where one was not previously identified by based upon geomorphic conditions. For the South Fork, all relict channels identified in the existing topographic information were mapped. Then and assessment was made to identify what channels were: inundated by floods; within close proximity to existing channels; could convey quantities of water and flow; had a gradient and course length suitable to encourage flow; upstream of constriction points; and located in a depositional reach that could experience channel bed aggradation. These channels were identified as having near-term avulsion hazard potential.

5/26/2015 - 58 -

FIGURE 4-9: South Fork Nooksack Reach

5/26/2015 - 59 -

FIGURE 4-9A: Acme - Skookum Creek Sub-reach

5/26/2015 - 60 -

FIGURE 4-9B: Van Zandt – Acme Sub-reach

5/26/2015 - 61 -

FIGURE 4-9C: Confluence – Van Zandt Sub-reach

5/26/2015 - 62 - Closure Paul Pittman of Whatcom County Public Works prepared this report for Whatcom County Planning and Development Services and the Whatcom County Flood Control Zone District for use in developing Channel Migration Zones. The material in it reflects the judgment of Whatcom County staff in light of the information available at the time of report preparation. Any use which a third party makes of this report, or any reliance on decisions to be based on it are the responsibility of such third parties. Whatcom County or its staff accept no responsibility for damages, if any, suffered by any third party as a result of decisions made or actions based on this report.

The report and all associated drawings are submitted for the confidential information of Whatcom County for planning purposes. If this report is issued in an electronic format, an original paper copy is on file at Whatcom County Public Works and that copy is the primary reference with precedence over any electronic copy of the document, or any extracts from our documents published by others.

5/26/2015 - 63 - The assessment and this report were conducted by:

______Paul David Pittman, LEG Whatcom County Public Works

Statement of Limitations

This document and the associated maps have been prepared by Paul David Pittman of Whatcom County Public Works for the exclusive use and benefit of Whatcom County. No other party is entitled to rely on any of the conclusions, data, delineations, and opinions or any other information contained in this document and the associated maps.

The identified in this document and the associated maps represent Whatcom County’s best professional judgment based on the information available at the time of its completion and as appropriate for the project scope of work. Services performed in developing the content of this document have been conducted in a manner consistent with that level and skill ordinarily exercised by members of the geologic profession currently practicing under similar conditions. No warranty, expressed or implied, is made.

5/26/2015 - 64 -

References (in progress) Rapp, C, and Abbe, T. 2003. A Framework for Delineating Channel Migration Zones. Department of Ecology (DOE). Ecology Publication #03-06-027.

Collins, B and Sheikh, A. 2004. Historic Channel Locations of the Nooksack River. Whatcom County Public Works.

Collins, B and Sheikh, A. 2002, Historic Riverene Dynamics and Habitats of the Nooksack River. Nooksack Indian Tribe

Lappen, T. 2000.

Haugarud, Ralph, 2005 Kendall …

Herrera, 2005 South Fork Avulsion Hazard Analysis - memorandum

Hutchings, Richard, 2005 – master’s thesis

Indrebo, Mark. 1998. Stream Channel Classification and Channel Changes Along the North Fork Nooksack River, Washington. Master’s Thesis – Western Washington University

Maudlin and Pittman, 2003 – Avulsion work….

Pittman, 2007 – Upper Reach 4 assessment

WSDOT Corridor Analysis…2001

LiDAR

Review and reference as applicable Alan Soicher’s study and other SF mapping efforts,

GLO survey,

5/26/2015 - 65 -

APPENDICES A, B, and C

Appendix A: Historic Locations of the Nooksack River (Collins & Sheikh, 2004)

Appendix B: NOOKSACK RIVER UPPER REACH 4 (DEMING TO CEDARVILLE BRIDGE) GEOMORPHIC ASSESSMENT (Pittman, 2007)

Appendix C: Justification of discrepancies between the Whatcom County mapped EHA and the WSDOT North Fork Nooksack River Corridor Analysis mapped EHZ

5/26/2015 - 66 - 5/26/2015 67

Appendix A: Historic Locations of the Nooksack River (Collins & Sheikh, 2004)

5/26/2015 68 Appendix B: NOOKSACK RIVER UPPER REACH 4 (DEMING TO CEDARVILLE BRIDGE) GEOMORPHIC ASSESSMENT (Pittman, 2007)

5/26/2015 69

NOOKSACK RIVER UPPER REACH 4 (DEMING TO CEDARVILLE BRIDGE) GEOMORPHIC ASSESSMENT

Photo by Steve Seymour, 2007

October 30, 2007

Whatcom County Public Works River and Flood Division 322 North Commercial Street Bellingham, WA 98225

5/26/2015 70

Overall Project Objective: This document is one of up to five planned documents designed to provide comprehensive information to develop “meander limit” delineation consistent with the Lower Nooksack River Comprehensive Flood Hazard Management Plan (Whatcom County, 1999). The other planned documents include: Engineering Analysis of Existing Conditions – Upper Reach 4 (Whatcom County, draft-2007); Hydraulic Modeling Results for Upper Reach 4 (Whatcom County, in progress-2007); Economic Analysis for Upper Reach 4 (Whatcom County, draft-2007); and a summary of habitat data/analyses. No specific recommendations or conclusions are drawn from these background informational documents; however, it is hoped that these documents, as best available science, will provide the foundation for informed decisions with broad public benefits.

Geomorphic Analysis Purpose: The purpose of this detailed geomorphic analysis is to map erosion and avulsion hazards of the Nooksack River in Upper Reach 4 between the South Fork confluence and the Cedarville Bridge (Figure 1). This document accompanies and provides supporting documentation for the DRAFT Nooksack River Erosion and Avulsion Hazard Zone Mapping. The intent of this specific study is to provide scientific information and an associated set of maps to support a hazard and risk analysis.

Historically, analyses of flood hazards have focused on rising water levels resulting from over-bank flooding and inundation of the floodplain. In fact, the underlying premise of FEMA maps is based on flood elevation. While this flooding does involve significant risk and may result in substantial financial loss, additional flood hazards exist; in some cases, these are more hazardous with higher risks and may result in even greater public expenditures. The additional flood hazards considered in this document are lateral erosion, avulsions (rapid channel shifts), high water velocities, and alluvial fan flood hazards. This assessment considers the hazard potential from lateral erosion and near-term hazard probability from “normal” alluvial conditions. The Nooksack Valley is a dynamic environment and other hazards, such as lahars, landslides, and catastrophic flooding, do exist; therefore, this report should be used in conjunction with other hazard assessments and studies to guide comprehensive planning decisions, assess relative hazard and risk, and support additional studies.

Holocene Geomorphic Setting - Natural Three forks of the Nooksack River exit the Cascades and converge near the western edge of the Cascade Foothills near Deming. The Holocene (past 10,000 years) Nooksack River floodplain is situated in relict glacial valleys shaped by both ice and water. The valley walls in the subject reach consist of bedrock, Pleistocene glacial deposits, and Holocene deposits (colluvium and alluvium) (Figure 2). The Pleistocene and Holocene deposits erode easily by fluvial

5/26/2015 Page 71

processes, as can be evidenced by numerous arcuate erosional scarps on both sides of the valley, relict channels, and modern observed bank erosion (Figure 3). The channel is currently situated in the south side of the valley and occasionally makes contact with the Pleistocene glacial deposits and bedrock geology. It should be noted that landslide hazards also increase as the active river channel erodes into and over-steepens the glacial deposits. Current landslide activity at the “Clay Banks” and landslide scarps along the south bank terrace in glacial material provide evidence of this. Under natural conditions, the right bank of the channel is bound only by Holocene alluvium. Downstream of the study reach, the right over-bank of the Nooksack River floodplain between Lawrence and Everson is actually within the Fraser River basin. When the Nooksack River overtops its right bank north of Lawrence, it flows toward the Fraser River via the Sumas Valley (Figure 4). This unusual artifact results from the recent condition when the Nooksack flowed down the Sumas Valley into the Fraser River; this occurred for most of the Holocene (Pittman, et al, 2003). Over-bank flooding in Upper Reach 4 typically would re- enter the Nooksack watershed via Smith Creek; however catastrophic flooding and the migration of the channel to the north side of the valley could increase the potential for water to enter the Fraser watershed via the Sumas Valley. A massive influx of sediment into Reach 4 from upstream could also cause increased overflow to the Fraser. A lahar, or volcanic , originating from Mount Baker could provide such a slug of sediment. Numerous lahars have traveled down the Nooksack from Mount Baker in the Holocene (Kovanen et.al., 2001). Deposits of a lahar originating in the Middle Fork Nooksack drainage approximately 6,000 years ago traveled at least as far as Cedarville and may have deposited material several feet thick in the channel bed (Pringle et. al., 2001). Another lahar event of this magnitude or larger would have serious implications for flood hazards for both the Nooksack and Sumas Valley. The Nooksack is a gravel bed river throughout most of its total length, including the subject reach. Geomorphic and historic information suggest that the river was likely an anastomosing channel in this reach (Figure 5). Anastomosing river patterns consist of two or more channels that are evident at both low and high flows (Figure 6) and often occur where river gradients are at the break between the more commonly recognized braided and meandering channel forms. The islands separating the channels are typically well vegetated. Anastomosing channels migrate both by lateral erosion and avulsion; however in an uncontrolled, natural system it is likely that avulsions are responsible for most of the large lateral changes. The geomorphology of the valley in Upper Reach 4 demonstrates that the channel was capable of migrating across the entire valley margin and that significant avulsions could quickly move the primary channel from one side of a valley to the other. Work by Collins indicates that anastomosing channel patterns were common in the Puget Lowland in rivers similar to the Nooksack (Collins and Sheikh, 2004).

5/26/2015 Page 72

Casual observation of modern conditions would lead some to believe that Reach 4 is unusually steep and trending towards a braided channel. Yet examination of the river gradient illustrates that it is more likely to have trended towards an anastomosing channel form in the undisturbed condition. The elevation at the upstream end of the reach is 204 feet and 142 feet (NAVD 1988) at the Cedarville Bridge, giving the Upper Reach 4 a slope of approximately 0.2%. The slope is similar to the slope of the South Fork Saxon to Acme Reach (0.3%) and the Cedarville to Everson Overflow Reach (0.2%), whereas the North Fork and Middle Fork Nooksack have slopes much greater (Table 1);

Reach Beginning El. Ending El. Distance Slope Deming to Cedarville (subject reach, Mainstem Nooksack) 204 feet 142 feet 25,340 feet 0.2 % Cedarville to Everson Overflow (downstream Reach 4, Mainstem Nooksack) 142 feet 94 feet 23,740 feet 0.2 % Saxon to Acme (South Fork) 350 feet 286 feet 21,875 feet 0.3 % North-Middle Fork confluence to North-South confluence (Truck Road – North Fork) 280 feet 210 feet 16,250 feet 0.4 % Mosquito Lake Bridge (MF) to North-Middle confluence (Middle Fork) 522 feet 282 feet 23,390 feet 1.0 %

Table 1: Slopes of comparative reaches of the Nooksack River

Holocene Geomorphic Conditions - Modern Geomorphologic evidence, early surveys, and anecdotal documentation suggest that the characteristics of the modern Nooksack River channel that we observe today have evolved from the predominant natural conditions within the first few decades of agricultural, forestry, and industrial development during the late 1800’s and early 1900s. The primary factors that initially modified the channel morphology in the middle and late 1800’s were removal of debris jams for navigation and clearing of floodplain/riparian forests. Beginning in the early 1900’s, hydromodification of the floodplain with levees and revetments was undertaken to support agriculture. The river morphology responded to the disturbances by evolving into a different channel morphology. The disturbances were numerous and included: • A decrease in channel roughness and loss of grade controls and areas of stored sediment through removal of large woody debris (LWD) • Decreased bank cohesion and root strength resulting from riparian clearing • Decreased flood plain roughness and water and sediment storage from floodplain clearing and draining • Increased confinement resulting from levee construction • Channel shortening and straightening for navigation and transport

5/26/2015 Page 73

• Increased sediment supply from upland clearing and drainage modifications, which leads to increased lateral erosion

The result of these disturbances was an evolution from a broader active floodplain/valley area with a complex multi-channeled system of individually narrow threads (wandering and anastomosing channel forms), to a single, wide thread channel system (single thread sinuous, braided, wandering channel forms). Geomorphic and survey information demonstrate that the mainstem Nooksack River between Everson and Deming exhibited channel forms of anastomosing, wandering, and meandering morphology pre-development. However, by the 1930’s, the same reach had evolved into a single-thread sinuous/braided morphology and many of the side channels and a large portion of the floodplain that were active in the late 1880’s became isolated on a terrace (Figure 7). In general, the modern Nooksack River channel is shorter and wider, is locally entrenched, has higher velocities, is capable of transporting more sediment, has decreased bank cohesion (from root strength), and lacks much of the geomorphic channel pattern complexity that results from floodplain connectivity, intact riparian vegetation, and large woody debris jams typical in unaltered fluvial systems of the western Cascade Foothills (Collins and Sheikh, 2002). While not the topic of this assessment, it should be noted that the loss of channel pattern complexity also reflects a loss of salmon habitat diversity and complexity affecting two ESA-listed salmonid species and multiple other salmon populations.

Presently, bank armoring is the primary driver of the river morphology changes and patterns observed in Upper Reach 4 and likely has been over the past half century (Figure 8). The presence and locations of the armored banks cause changes to the floodplain, active channel area and channel morphology evolution. The effects of the bank armoring including: • A reduced active floodplain area • Preference for, or forcing of, a single thread channel • An increased active channel with a larger proportion of either very young or un-vegetated area • Increased channel constriction resulting in reduced channel migration area and increased hydraulic energy • Altered sediment transport, depositional and storage areas • Increased sediment inputs (e.g. clay banks) • Increased stream velocities and depths • Altered bank cohesion properties through loss of riparian vegetation • Reduced channel roughness from loss of large wood debris

Channel forms and migration are systems of reactions inexorably linked, such that one action that surpasses a critical state causes a “domino effect” to any

5/26/2015 Page 74

number of elements that affect not only the localized area where the disturbance occurred, but creates impacts that migrate in multiple dimensions both spatially (upstream, downstream, laterally, vertically) and temporally. The modern channel form of the Nooksack River in the subject reach currently exhibits characteristics of a low ordered braiding, single thread sinuous channel that appears to evolve towards a “wandering” channel morphology in- between periods of disturbances. A wandering channel morphology contains both elements of an anastomosing channel and a braided channel. Channel forms in low-gradient, gravel bed systems, such as the subject reach, are often characterized by having an irregularly sinuous channel or multiple channels that can have anastomosing, (splits with stable vegetated islands), wandering, single- thread sinuous, or braided (unvegetated bars commonly with seasonal or perennial side channels) channel forms; and that these forms can evolve from one form to another temporally in response to disturbances or exceedances of critical kinematic thresholds (Church, 1992. Stolum, 1996. Hooke, 2003).

The wandering and braided in low-gradient gravel bed rivers are less predictable over time than an anastomosing or meandering channel form. In general, braided channels have relatively high erosion rates associated with: higher stream energy (steeper slopes), channel banks with relatively low cohesion, and significant bedload sediment transport and deposition. High aggradation rates, either locally or widespread, can exacerbate lateral migration, and thus bank erosion. For comparison, the unconfined reach downstream of Cedarville has a similar gradient, discharge, and sediment load as Upper Reach 4. However, in recent decades the lower reach has exhibited a wandering morphology with the development of anabranching channels (side channels) that have been “stable” through numerous flood events, whereas the confined subject reach appears unable to develop or organize itself into a more stable channel condition. Similarly, erosion rates in the downstream reach have been relatively slow in the past decade when compared to the Upper Reach 4, which may reflect the lower energy associated with the wandering channel morphology versus the braided morphology.

Hazards Identified Five channel migration hazards were identified for this analysis: lateral erosion hazards; fluvial-induced landslides; avulsion hazards; high water velocities, and alluvial fan hazards. The methods for assessing each hazard are described below and all five hazards are reflected in the Summary of Channel Migration Hazards Map for Upper Reach 4 – Deming to Cedarville Bridge (foldout map – Plate 1).

Lateral Erosion Hazards Hazards associated with lateral erosion are primarily to structures which can be undermined and fall into the river, this may include residential/agricultural

5/26/2015 Page 75

structures, public infrastructure including flood structures. Erosion rates can occur “quickly” in that several meters of bank can erode in the course of a flood event (several hours to days), but humans can minimize their exposure to the hazard by getting out of the way. The hazard is significant to humans if they happen to be too close to the edge of an eroding bank. Table 2 shows the erosion rates observed in upper Reach 4 over the past century:

Average time-weighted bank erosion rate for forested 9.4 m/yr floodplain/high bank (measurements based on interpreted (31 ft/yr) conditions observed in orthophotographs for the period 1933- Median: 8.2 m/yr Standard Deviation: 2006) 19 m/yr Maximum time-weighted erosion rate for forested floodplain/high bank (measurement observed in the interval 21.3 m/yr between 1938 and 1950 from orthophotographs) (70 ft/yr) Time-weighted average annual lateral migration rates 1885 – 2002 (measured channel offset along transect lines averaged 8.4 m/yr by reach; Collins & Sheikh, 2004) (28 ft/yr) Table 2: Measured erosion and lateral migration rates for Upper Reach 4. Note: Data for single-event or short-term maximum erosion rates were not available and higher maximum rates may exist. In addition, erosion rates for historic river bottom or unvegetated bars were not assessed and may be higher.

Historic placement of revetments to address lateral erosion began in the 1940s and 1950s and revetments to control lateral erosion continue to be constructed and maintained. As a consequence, the channel configuration that exists today is very confined in many places within this reach. A majority of the channel migration within Upper Reach 4 is constrained by either bank revetments or by geologic features, such as bedrock, that limit lateral erosion. The reach is sufficiently confined such that it limits downstream migration of meander bends and locks them in-place. Historically however, this reach was very similar to the Cedarville to Everson Reach in its lateral migration patterns and rates. Significant portions of the floodplain are currently leveed in this reach and several former active channels and river bottom have been isolated by agricultural levees. A result of the levee configuration is in-channel flow depths are increased over the natural condition. The consequence of increasing flow depths is that the hydraulic forces are greater and increased scour and higher velocities occur, thus bed and bank erosion rates can greatly increase. It is likely that if the and revetments were removed, migration patterns would behave similar to the unconfined reach downstream of Cedarville with a large degree of migration contained within the Historic Migration Zone and a low-order braiding evolving towards a wandering or anastomosing morphology. The benefits of a wandering or anastomosing morphology is that a reduced overall erosion rates outside of the meander corridor could be anticipated, flow velocities are reduced (increased wetted perimeter, increased channel/floodplain roughness), and channel depth is decreased (reduced scour).

5/26/2015 Page 76

In order for the wandering or anastomosing channel morphology to fully develop, the river would need access to a wider migration corridor (a minimum of one meander amplitude wide) (Figure 9).

Fluvially-Induced Landslides Landslides induced by fluvial erosion are a natural and common process of valley widening. Lateral channel migration can undercut and destabilize hillsides and slopes outside the floodplain as the channel erodes the toe of slopes (see Figure 3). The failure of a slope or hillside into the channel can block or redirect the channel and exacerbate channel migration processes, and potentially result in an avulsion (see example in Appendix A, pg 30). In addition to obvious additional hazards for home, roads or other infrastructure on the landslide or in the impact area below the slide, indirect hazards include landslide dams or landslide-induced surge waves that may locally increase flood elevations and velocities. Unlike lateral erosion, landslides can happen abruptly and without warning, and can cause catastrophic channel alteration instantly. Landslides can have significant economic implications, such as destroying roads and bridges, rupturing pipelines, and destroying multiple dwellings and facilities; due to the potential consequences the risk is accentuated. Other impacts of landslides are increased sediment supply and damage to fish and wildlife habitat.

Currently, a significant landslide resulting from fluvial processes destabilizing a hillside exists within this reach. The historic record shows that the Pleistocene deposits composing the “clay banks” have repeatedly failed after periods during which the river erodes into the deposit. Recently, the river eroded the toe of a portion of this deposit and a series of significant landslides occurred. This pattern repeats itself in areas in which contact with the channel occurs as evidenced by a series of reforested landslide scars at the edge of the glacial terrace. Historical aerial photographs from the 1930’s show active landslides at several of these now reforested scars.

Recent 2006 slide event at the Clay Bank shown in photos below:

5/26/2015 Page 77

February 15, 2007 by GeoTest April 22, 2006 Whatcom County

An assessment of erosion rates, slope retreat rates, slope failure mechanisms, repose angles, and debris run-out was conducted and is summarized in Table 3: Maximum Toe Retreat Observed 1933-2004 116 m (380 feet) (time-weighted average) (1.6m/yr; 5.35 ft/yr) Averaged Toe Retreat Rate 1906-2006 upstream point 1.4 m/yr (4.5 ft/yr) Averaged Toe Retreat Rate 1906-2006 mid point 0.8 m/yr (2.5 ft/yr) Averaged Toe Retreat Rate 1906-2006 downstream point 1.5 m/yr (5 ft/yr) Maximum Top of Bank Retreat Observed– single event, instantaneous (2006) 61 m (200 ft) Estimated Potential Top of Bank Retreat After Toe Failure (2006) 120-150 m (400-500 ft) Unstable Slope Angle (Deming Sand) 35 degrees Areas currently exceeding unstable slope angles (Deming Sand) 2 Repose Angle (Glaciomarine Drift – unweathered) ~12 degrees Debris Runout Distances Observed in Three 2006 Events (Note: A pre-historic slide created a debris runout that may have 152 m (500 ft) exceeded 300 m; 1000 feet – Figure 10). Table 3: Summary of slope failure data and observations. Work maps and supporting documentation are shown in Appendix A.

Based on the historic observations and current channel configuration, future slope failures are believed to be imminent. Two areas currently exhibit conditions that demonstrate slope instability exists and failure at either of these two areas could be anticipated. The potential size and likelihood of future catastrophic failures (larger than observed 500-ft debris run-out) is not known, although geologic evidence suggests that larger failures have occurred (Figure

5/26/2015 Page 78

10). Even small landslides can impact channel migration patterns at this location, and more importantly they can greatly increase flood levels and avulsion potential. A large slope failure could cause extreme lateral migration or channel avulsions as described in Avulsion Hazards below. A 2005 landslide (Hazel Landslide) into the North Fork Stillaguamish River channel east of Everett created conditions similar to the conditions discussed in this analysis (Appendix A). A large, previously unmapped landslide immediately north of Deming on Sumas Mountain was observed on the 2006 LiDAR and mapped by the author (Appendix B). The slide geomorphology demonstrates potential instability and recent movements in the late Pleistocene and possibly Holocene based on the offset of glacially derived surface fabric. The toe of the landslide appears to have been eroded away, likely by the Nooksack River. A remnant portion of debris run-out may exist on the south side of the Nooksack River on Stewart Mountain, indicating that this slide may have temporarily blocked the Nooksack River. The current stability of the landslide has not been assessed; however in light of the potential consequences (significant threat to life and public safety, damming of the Nooksack River, rupture of the Williams Gas Pipeline, severing of the Mount Baker Highway), some additional assessment may be warranted. It should be noted that there has been significant seismic activity near the town of Deming in recent decades and a recent discover of surface ruptures from an active fault have been found in close proximity to Deming. Seismic activity can be the trigger slope instability and failure.

Avulsion Hazards Avulsion hazards are characterized by rapid change in channel location with high water velocities and flow depths, flooding, erosion, and potential debris impacts. The process of channel flow will cause surface erosion (scour), lateral fluvial erosion, and sediment deposition. Identified “possible” avulsion hazards were mapped based on existing topographic information, hydraulic modeling, and levee and road locations. Based on field observations, some avulsion channels were identified as warranting a detailed risk assessment. All areas within the historic migration zone are included within the avulsion hazard area. Three types of avulsion processes are defined for this study: • Relict/secondary channel capture: Capture of a relict channel or secondary channels within the floodplain during flood events or lateral channel migration into a relict channel. Channels may temporarily convey water only during floods; these channels may capture the entire channel flow causing the previously-occupied channel to be abandoned. Chute and neck cut-off avulsions may also occur as meanders migrate. • Depressional Areas/Swales: Topographically low areas subject to rapid or instantaneous high flow velocities with potential debris.

5/26/2015 Page 79

• Levee break (crevasse splay): Areas adjacent to natural or man- made levees and road fills that can breach and temporarily create channel-like hazard conditions. A levee break is not necessarily an avulsion by the traditional definition, however the nature of the hazard can lead to avulsions or avulsion-like hazards.

Hydraulic modeling was used to identify channel forms that were inundated during flood events. A 100-year discharge was routed though the hydraulic model and the resulting flood elevations were then used to assess avulsion potential. In addition, alternative scenarios were assessed for their impacts on avulsion potential; this is discussed in more detail in the hydraulic report that will accompany this document. Two assumptions were made to map potential channel avulsions: 1. The channel bed elevation is somewhat stable over the long-term 2. In the short-term, localized aggradation of a channel bed may increase avulsion potential at a particular location. A recent sediment study determined that in the interval between 1996 and 2004, no net channel-bed aggradation occurred within this reach (KWL, draft 2006). In addition, field evidence suggests that long-term (1906-2006) channel bed stability with net incision has historically existed in the subject reach. Understanding trends in channel bed elevation, in addition to reliable hydraulic modeling and an understanding of erosion potential is necessary to make more site-specific analyses and to assess condition changes with time. Avulsion risk has likely decreased when compared to natural conditions as channel incision has occurred in much of the subject reach and access to side channels and relict channels on the floodplain has been greatly reduced. In addition, the increased channelization and decreased large woody debris jams have also lessened the potential for avulsions. How these factors change over the design period of the study will affect the risk of avulsion occurrence. Topographic lows and relict channels were identified based upon topographic mapping from LiDAR and historic topographic maps (Figure 11). Field checks of topographic conditions were made in locations where immediate and potentially catastrophic impacts to existing infrastructure appeared possible. These features may routinely fill with water during floods, and may even convey flow during flood events. With this concept in mind, the relict channels and low areas are considered near-term avulsion hazards. They were identified because they exhibited conditions that were interpreted to be suitable avulsion pathways within the near-term (within the next decade). To be considered a near-term avulsion hazard area, one or more of the following conditions needed to exist: • A continuous relict channel or topographic low that was inundated and capable of conveying significant flows for frequent (greater than 2- year interval) floods • Close proximity to the current channel location (less than the maximum observed erosion rate for that reach)

5/26/2015 Page 80

• Area should provide a “hydraulic energy benefit” or “short-cut”. Examples are a chute cut-off or a flow path that had an increase in slope and/or a decrease in roughness such that the channel could opt to take a new course and partially or fully abandon its previous course because it is “easier” from an energy standpoint. • Potential for main channel flow obstruction and backwatering, such as bridges or constrictions, log jams or other potential obstructions. The concept is that a downstream obstruction will raise upstream water levels and increase the opportunity for new channels to seek courses around the obstruction.

The likelihood of experiencing avulsion hazards in the subject reach are exacerbated by the existing landslides and potential for additional landslides originating on the opposite (southwest) bank. Hydraulic modeling examining alternative scenarios was used to assess potential scenarios and outcomes (Whatcom County and Chang draft – in progress 2007). Indeed, historic avulsions did occur in this reach, most notably during the 1990 flood.

The potential of a channel avulsion into the Sumas Valley in this reach is considered very low because Smith Creek routes most of the overflow back into the Nooksack and infrastructure, especially the elevated railroad grade, creates a levee that would be difficult for a new channel to pass through. However, a major overflow, even if it did not result in an actual avulsion, could cause significant damage to the Mount Baker Highway as multiple historic flow paths cross this roadway. It should be noted that some conditions may or will exist in which engineered response to address an avulsion may be impractical or impossible to implement during a large flood. As stated previously, scenarios such as landslides, lahars or other volcanic events, catastrophic flooding (greater than 100-year event or outburst flooding), or channel blockages, greatly increase the potential for avulsions. With the exception of landslides, the recurrence interval of many of these events is very low; however, each has occurred at least once in the past 6,000 years.

Alluvial Fan (discussion) Whatcom County Planning and Development Services previously mapped alluvial fan locations (PDS, 1992). Two primary alluvial fans are present in Upper Reach 4, McCaulay Creek and Smith Creek. The mapped alluvial fan locations mapped by PDS are shown with modifications by the author in Figure 12. Inherent hazards on alluvial fans are similar to both erosion hazards and avulsion hazards, with an increased potential for debris impacts and sediment deposition. A preliminary geomorphic assessment of the Smith Creek Alluvial Fan indicates that the hazard potential there appears great. Observations include: • Two large, active landslides are near the apex of the fan

5/26/2015 Page 81

• A new fan is in the process of rebuilding the fan as past erosion from the Nooksack River truncated the historic fan, • The fan gradient is over-steepened by the eroded fan • The Smith Creek basin is very large and large discharges and debris flows are possible • Logging activity in the upper basin has occurred altering basin hydrology and possibly slope stability At risks are the Mount Baker Highway, Burlington Northern Railroad, Williams Gas Pipeline, and numerous residences that could be impacted by a . A detailed site-specific analysis would provide more hazard and risk information for flood hazard management at this location.

5/26/2015 Page 82

The assessment and this report were conducted by:

Statement of Limitations

This document and the associated maps have been prepared by Paul David Pittman of Whatcom County Public Works for the exclusive use and benefit of Whatcom County. No other party is entitled to rely on any of the conclusions, data, delineations, and opinions or any other information contained in this document and the associated maps.

The geologic hazards identified in this document and the associated maps represent Whatcom County’s best professional judgment based on the information available at the time of its completion and as appropriate for the project scope of work. Services performed in developing the content of this document have been conducted in a manner consistent with that level and skill ordinarily exercised by members of the geologic profession currently practicing under similar conditions. No warranty, expressed or implied, is made.

5/26/2015 Page 83

REFERENCES

Church, M. 1992. Channel morphology and typology. In The Rivers Handbook. Edited by P. Calow. Blackwell Scientific Publishers, Oxford, UK. pp. 126-143.

Collins, B. D., A. J. Sheikh. 2002. Historical riverine dynamics and habitats of the Nooksack River. Interim Report to the Nooksack Indian Tribe. 93 p.

Collins, B. D., A. J. Sheikh. 2004. Historical Channel Locations of the Nooksack River. Report to Whatcom County Public Works. 62 p.

Hooke, J. 2003. Trans Inst Br Geogr. NS 28 (2003). pp. 238-253.

Kovanen, D. J., D. J. Easterbrook. 2001. "Late Pleistocene, Post-Vashon, Alpine Glaciation of the Nooksack Drainage, North Cascades, Washington." GSA Bulletin 113 (2001): pp 274-288.

Ellis, E., H. Weatherly. Draft - 2006. Report to Whatcom County Public Works by Kerr Wood Leidal and Associates (KWL). p 64.

Lapen, T. 2000. Geologic Map of the Bellingham 1:100,000 Quadrangle, Washington. Open file report 2000-5, plate 1. Washington Division of Geology and Resources, Washington State Department of Natural Resources.

Planning and Development Services - Staff. 1992. Alluvial Fan Hazard Areas. Whatcom County Environmental Resources Report Series.

Pittman, P. D., M. R. Maudlin. 2003. Geological Society of America Abstracts with Programs, Vol. 35, No. 6, September 2003, p. 334.

Pringle, P. T.; S. M. Kevin. 2001. Postglacial influence of on the landscape and of the Puget Lowland, Washington—A review of geologic literature and recent discoveries, with emphasis on the landscape disturbances associated with lahars, lahar runouts, and associated flooding. In Puget Research 2001, Proceedings: Washington State Puget Sound Water Quality Action Team, 23 p.

Stolum, H. 1996. River meandering as a self-organizing process. Science 271. pp. 1710-1713.

Whatcom County Public Works, S. Chang. In progress - 2007. Interim data from hydraulic modeling – upper Reach 4, Nooksack River. Internal correspondence and documents.

5/26/2015 Page 84

5/26/2015 Page 85

5/26/2015 Page 86

5/26/2015 Page 87

5/26/2015 Page 88

5/26/2015 Page 89

Examples of anasotomosing and wandering channel forms

5/26/2015 Page 90

5/26/2015 Page 91

5/26/2015 Page 92

5/26/2015 Page 93

5/26/2015 Page 94

5/26/2015 Page 95

5/26/2015 Page 96

Appendix A Clay Bank Slope Stability Assessment Slope Retreat Rates

Data Sources 1909 Topographic Map (USGS) 1933 Aerial photography (US Military) 2006 Aerial photography (Whatcom County) 2006 LiDAR (USGS)

Data and work map used to assess toe retreat between 1909 and 2006.

5/26/2015 Page 97

Appendix A Clay Bank Slope Stability Assessment Slope Angle Analysis

2006 slope failures

Historic slope failures

Slope angles from 2006 LiDAR. Red colored slopes exceed 35 degrees. Slopes that show red at the base of the slope are more susceptible to stability issues and failure. The two areas circled in blue exhibit slope conditions that are at higher risk of larger failure because of over- steepened conditions at the base of the hillside. Top of bank retreat from the recent 2006 failures could continue until a more stable slope at the top of bank exists.

5/26/2015 Page 98

Appendix A Clay Bank Slope Stability Assessment Field Visit Photos (2006-2007)

Photo A1: 2006 – Three “lobes” from three separate events Photo A2: Oversteepened slopes east of 2006 failures

Photo A2: Oversteepened slopes west of 2006 failures Photo A3: Beginning signs of slope stability issues at historically stable area (flanked by recent failures)

Appendix A Clay Bank Slope Stability Assessment Comparable river erosion-induced slope failures

5/26/2015 Page 99

Hazel Landslide, North Fork Stillaguamish River. In 2005, a bank over- steepened by lateral channel migration (river erosion) failed and blocked the North Fork Stillaguamish River. A subsequent avulsion around the toe of the landslide occurred.

5/26/2015 Page 100

Appendix B Deming Landslide

Approximate boundary of the “Deming Landslide” mapped by the author. The slide was previously unmapped as one large slide, and instead was shown as three much smaller slides. The basis for mapping the smaller slides as one large slide is the geomorphic interpretation of the 2006 LiDAR. The geomorphology demonstrates large head-scarps with evidence of slumping that offsets the glacial deposits. A remnant portion of slide debris run-out including a ponded area may exist on the south side of the river and could give clues as to the timing of the slide event. The Nooksack River may have removed slide debris from previous slope failures. Appendix B Deming Landslide Seismicity near Deming 5/26/2015 Page 101

Pacific Northwest Seismic Network Deming Seismic “Swarm” Red dots denote epicenters >4.0 M Specific data shown in box below

NEAREST HISTORICAL SORTED BY DISTANCE FROM REFERENCE EVENT

Year MoDy Hr:Mn Depth Mag Distance Azimuth Event Location (UT) (km) (km) (Deg) 1990 0621 04:49 1.5 3.6 3.4 268.8 4.9 km ENE of Deming, WA 1990 0414 05:33 12.6 5.0 4.0 279.7 4.7 km ENE of Deming, WA 1990 0731 23:49 1.3 3.5 4.8 259.7 3.4 km ENE of Deming, WA 1990 0403 02:18 1.7 4.0 5.0 265.7 3.3 km ENE of Deming, WA 1990 0414 06:02 3.1 3.6 5.3 259.0 2.9 km E of Deming, WA 1990 0402 11:13 0.8 4.3 6.0 262.5 2.3 km ENE of Deming, WA 1990 0414 05:40 3.6 4.0 6.3 252.4 2.1 km E of Deming, WA 1990 0523 01:02 1.9 3.6 6.5 258.6 1.7 km E of Deming, WA 1931 0418 03:55 0.0 5.0 16.9 203.8 14.0 km S of Deming, WA 1964 0714 15:50 0.0 4.6 29.6 283.4 15.3 km N of Bellingham, WA

______NEAREST HISTORICAL EARTHQUAKES SORTED CHRONOLOGICALLY

Year MoDy Hr:Mn Depth Mag Distance Azimuth Event Location (UT) (km) (km) (Deg) 1931 0418 03:55 0.0 5.0 16.9 203.8 14.0 km S of Deming, WA 1964 0714 15:50 0.0 4.6 29.6 283.4 15.3 km N of Bellingham, WA 1990 0402 11:13 0.8 4.3 6.0 262.5 2.3 km ENE of Deming, WA 1990 0403 02:18 1.7 4.0 5.0 265.7 3.3 km ENE of Deming, WA 1990 0414 05:33 12.6 5.0 4.0 279.7 4.7 km ENE of Deming, WA 1990 0414 05:40 3.6 4.0 6.3 252.4 2.1 km E of Deming, WA 1990 0414 06:02 3.1 3.6 5.3 259.0 2.9 km E of Deming, WA 1990 0523 01:02 1.9 3.6 6.5 258.6 1.7 km E of Deming, WA 1990 0621 04:49 1.5 3.6 3.4 268.8 4.9 km ENE of Deming, WA 1990 0731 23:49 1.3 3.5 4.8 259.7 3.4 km ENE of Deming, WA

5/26/2015 Page 102

Appendix C: Justification of discrepancies between the WC EHA and the WSDOT EHZ (reference reach breaks to Figure…)

• Reach 1: Left Bank between Reach 2 and Middle Fork confluence (Note A) The left bank in this reach consists of either glacial outwash or older Holocene alluvium. The deposit was mapped as glacial outwash by the DNR (Lapen, 2000), but I did not verify this in the field. It is my opinion, based on the elevation of the deposit and the terrace features, that it is glacial outwash. The bank is between 5 and 12 meters above the current floodplain elevation and marked with up to two distinct terraces. No evidence of recent channel occupation was observed on the upper terrace. The abruptness of the upper terrace may indicate the presence of more recent fluvial erosion, but at a time when the channel bed elevation was considerably higher. Drastic changes in channel bed elevation may occur from landslides that block the channel. The erodability of the material is potentially high. The right bank of the floodplain across river from this area is marked with numerous relict channels which indicates that the river has used much of the modern valley bottom at this location in the recent past, and historic channel locations show that this reach is unstable and subject to rapid channel location changes. This reach is also impacted by several alluvial fans (Kenney Creek and Canyon Creek) and potential channel changes of the Middle Fork Nooksack.

I was unable to estimate an erosion rate of the bank in this reach with the topographic data I had. A rate of 17 feet per year was estimated for a similar deposit in reach 3. Because of the dynamic nature of the channel in this reach, it is likely that the channel would not be located at the toe of the bank in this location for 100 continuous years. The work from Collins indicated that in the past 100 years, the channel was against this bank an average of 30% of the time. Using a rate of 17 feet per year over 30 years, I came up with an erosion hazard zone width of about 500 feet from the HMZ. The height of the bank may cause a slower rate of erosion to some extent, although the amount, if any, is unknown. If the interpretation of the bank material is glacial outwash is correct, it demonstrates that the bank has been there for 10,000 years of unchecked fluvial erosion. Rutsatz Road follows the top of the terrace and bank on the outwash deposit. Currently, lines much of the bank where the river has historically encroached upon the road. I elected to use a minimum EHZ width of 500 feet beyond HMZ, or 250 feet beyond the upper terrace when the river was away from the terrace edge or when greater than 500 feet from the HMZ.

In the WSDOT North Fork Corridor Analysis, their mapped Erosion Hazard Zone is placed at the edge of the bedrock slope and glacial valley

5/26/2015 Page 103

bottom (approximately 1000 feet outside of the HMZ). • Reach 2: Left Bank (Note B) I conducted a field visit of this area and determined that the landslide material in this reach would likely have a very low rate of erosion. An EHZ of between 50 and 100 feet was used at this location.

• Reach 3: Right Bank (Note C) The right bank between Bennett’s knoll and downstream of the Kendal Hatchery is glacial outwash. The bank height varies from about 2 to 4 meters above the present NF floodplain. Using historic topography and aerial photography, I tried to estimate the erosion rate of the outwash. I observed only minor erosion of the outwash during the period in which I had adequate topographic information. In the period between 1998 and 2001, an average erosion rate of 17 feet per year was observed, but was sustained for only three years time before the channel changed course and abandoned flow against the bank. Because of the wide active channel cross-section in this area, large floods do not increase substantially in height, such that only a few feet of flow erodes the toe of the bank in this area. Because of the broad channel migration area and the frequent channel changes, the channel has not been observed to occupy the margins of the HMZ for any great length of time. The channel migration area decreases as you move downstream from Bennett’s knoll, and thus the increased potential for the channel to occupy the margins of the HMZ. With increase occupation time, a greater potential for erosion exists.

I assumed that the channel may in the future occupy the right bank margin of the HMZ 25% of the time at the upper end of the reach (Bennett’s knoll to Kendall Creek), and 40% of the time in the lower end of the reach (Kendall Creek to landslide). This assumption was based somewhat on historic channel migration patterns in which the channel was against the right bank between 10 to 30% of the past 100 years at the upper end of this reach, and 30 to 50% of the time in the lower end. Some bias may exist in the lower reach because of existing riprap and some influence from the landslide debris ().

For the upper reach, I took the 17 feet per year average erosion rate with 25 to 30 years of predicted occupation against the right bank and arrived at an EHZ width of approximately 500 feet. In the lower reach, I took the 17 feet per year average erosion rate with a predicted 35 to 40 years of predicted occupation against the right bank and arrived at an EHZ width of approximately 650 feet. The WSDOT report has a 50-year predicted EHZ between 1500 and 2000 feet wide in this reach. Our estimated erosion rates for the glacial outwash are similar, but I assumed that the channel would not occupy the right bank continuously over the design interval based on its historic migration patterns.

5/26/2015 Page 104

• Reach 3: Right Bank (Note C) The right bank between Bennett’s knoll and downstream of the Kendal Hatchery is glacial outwash. The bank height varies from about 2 to 4 meters above the present NF floodplain. Using historic topography and aerial photography, I tried to estimate the erosion rate of the outwash. I observed only minor erosion of the outwash during the period in which I had adequate topographic information. In the period between 1998 and 2001, an average erosion rate of 17 feet per year was observed, but was sustained for only three years time before the channel changed course and abandoned flow against the bank. Because of the wide active channel cross-section in this area, large floods do not increase substantially in height, such that only a few feet of flow erodes the toe of the bank in this area. Because of the broad channel migration area and the frequent channel changes, the channel has not been observed to occupy the margins of the HMZ for any great length of time. The channel migration area decreases as you move downstream from Bennett’s knoll, and thus the increased potential for the channel to occupy the margins of the HMZ. With increase occupation time, a greater potential for erosion exists.

I assumed that the channel may in the future occupy the right bank margin of the HMZ 25% of the time at the upper end of the reach (Bennett’s knoll to Kendall Creek), and 40% of the time in the lower end of the reach (Kendall Creek to landslide). This assumption was based somewhat on historic channel migration patterns in which the channel was against the right bank between 10 to 30% of the past 100 years at the upper end of this reach, and 30 to 50% of the time in the lower end. Some bias may exist in the lower reach because of existing riprap and some influence from the landslide debris (boulders).

For the upper reach, I took the 17 feet per year average erosion rate with 25 to 30 years of predicted occupation against the right bank and arrived at an EHZ width of approximately 500 feet. In the lower reach, I took the 17 feet per year average erosion rate with a predicted 35 to 40 years of predicted occupation against the right bank and arrived at an EHZ width of approximately 650 feet. The WSDOT report has a 50-year predicted EHZ between 1500 and 2000 feet wide in this reach. Our estimated erosion rates for the glacial outwash are similar, but I assumed that the channel would not occupy the right bank continuously over the design interval based on its historic migration patterns.

• Reach 3: Left Bank (Note D) This area was excluded from the WSDOT EHZ. I included it in the 100- year EHZ because of geology (mapped as Holocene Alluvium), relict channel forms and low elevation.

5/26/2015 Page 105

• Reach 3: Left Bank (Note E) This area was included in the WSDOT EHZ. I excluded it because of geology, and topography. The area is mapped as partially as glacial outwash, and with detailed topographic data, I interpreted the deposit extending further north following the topographic feature. • Reach 6: Right Bank (Note F) This area was included in the WSDOT EHZ. I excluded it because of geology. I observed that the landslide mapped to the south of the channel extended across the channel and onto the right bank. The portion of the landslide deposit I observed included large bedrock boulders and blocks and I felt would limit lateral migration to approximately one-hundred feet from the edge of the geomorphic channel area.

5/26/2015 Page 106

Appendix ? Key points discussed with Technical Group- summary: • Department of Ecology manual basis for methodologies • Anthropomorphic features and structures influences not considered for mapping possible erosion or avulsion potential • Hydrologic conditions will not change dramatically • Anthropomorphic alterations and behaviors will not dramatically alter the conditions • Erosion assumption – possible (not probable) erosion within 100 year time frame • Erosion rate methodologies o Average annual x 100 (predominantly wandering channel morphology or where historic records not suitable to draw long-term trends) – North, Middle, South Forks, and Mainstem Nooksack River upstream of delta . Inherent assumptions – unidirectional over 100 years . Reach breaks and Collins methods of measure o Meander Migration Rate (predominantly meandering channel morphology) – Localized sub-reaches on Mainstem Nooksack River near Ferndale and delta reach . Inherent assumptions – meander migration will occur predictably in future • Relative erosion rates based on generalized “geology type” • Geology from Lapen, 2000 (DNR), and modified based on field observation/verification • Avulsion Hazard o Channel capture criteria – relict channel form capable of sustaining flow if inundated during a 100-year flood event or relict channel form within the EHZ o High velocity flow/debris inundation area criteria - Middle Fork and levied channels exception – depressional areas spontaneously occupied during rapid bed level change or levee breach • AHZ – map both possible Avulsion Hazard types, consider adding both and using a symbol to differentiate • AHZ – assumption: channel bed elevation changes over next 100- years not significant (Middle Fork exception) • Identify geologic hazards not considered in mapping exercise: flooding, lahar (and other volcanic), landsliding, tectonic deformation, alluvial fan flooding (clear-water, debris floods/flows), tsunamis… what did I forget?

5/26/2015 Page 107