DEPARTMENT OF ECOLOGY, SHORELANDS AND ENVIRONMENTAL ASSISTANCE SPOKANE COUNTY SMP UPDATE: CHANNEL MIGRATION ASSESSMENT

Transcription

1 DEPARTMENT OF ECOLOGY, SHORELANDS AND ENVIRONMENTAL ASSISTANCE SPOKANE COUNTY SMP UPDATE: CHANNEL MIGRATION ASSESSMENT 9/7/2007

2 SPOKANE COUNTY SMP UPDATE: CHANNEL MIGRATION ASSESSMENT Channel migration is a natural process associated with streams 1. Streams may migrate across valleys due to a variety of reasons including channel and bank erosion, meander chute cutoff, avulsion and aggradation. The channel migration zone represents the area within which a given stream may migrate over time. Channel migration is an important ecosystem process of streams supporting a number of ecological functions including wildlife habitat. The channel migration process is also an important risk factor for human settlements in that migrations can result in property damage and change flooding dynamics. Chapter WAC requires that channel migration areas be generally identified during the inventory and characterization phase of Shoreline Master Program updates: WAC (3) (c) (vii): Local government shall, at a minimum, and to the extent such information is relevant and reasonably available, collect the following information: (vii) General location of channel migration zones, and flood plains. Since, channel migration areas vary over space; mapping provides a more efficient means to identify their location. Mapping occurs during the communities Shoreline Master Program updates if it hasn t already been done in their Critical Area ordinance updates. Channel migration zones are to be managed per the SMA to 2 : Reduce potential hazards to human settlements by guiding development near streams. Protect shoreline ecological functions. Not allow development including fill to interfere with natural channel migration processes. Introduction The Washington Department of Ecology, Shorelines and Environmental Assistance Program (SEA) are responsible for managing Shoreline Master 1 The term stream encompasses all sizes of flowing water bodies. 2 Shoreline management code citations referring to channel migration can be found at Patricia Olson, PhD, LHG, SEA, Ecology 2 9/7/2007

3 Program updates and providing technical and policy assistance and guidance. The SEA program is developing web-based technical guidance for identifying when channel migration assessments are needed as part of the update. The guidance also provides a range of approaches and methods to use depending on management objectives and importance of environment and infrastructure values. The Spokane County Channel Migration Assessment provides an example to test the guidance and modify if needed. The web guidance uses a decision flow chart to answer the question: Is a channel migration assessment needed and where? We followed the decision flow chart for answering the above question (Figure 1). Each step is grouped into 1 of 5 tasks: Identify channel confinement and gradient. Evaluate channels for evidence of movement or bank erosion. Identify channel patterns and processes. Determine assessment approach and method. Do assessment. Figure 1: Decision flow chart used to identify streams that migrate is from the Ecology draft channel migration web guidance. Methods and Results We used ArcMap (Version 9.2) to evaluate channel migration from existing data. However, the tasks can be done manually. Limited field visits (3 days) occurred before the assessment. Patricia Olson, PhD, LHG, SEA, Ecology 3 9/7/2007

4 Task 1: Identify channel confinement and gradient. Table 1: Channel confinement and gradient influence channel migration Confinement indicator of space available for channel movement Channel gradient influences channel type, migration, bank erosion Confined: ratio of valley width (vw) to active channel width (bw) <2 Restricts channel movement Moderate confinement: vw/bw 2 and <4 Channel movement and erosion Unconfined: vw/bw 4 Channel can move rapidly >4%--high gradient Bank erosion, slope failure 2-4%--moderate gradient Limited migration & flooding, bank erosion, slope failure <2%--lower gradient Migration, flooding Method For this assessment, the channel characteristics shapefile from the WDFW Salmonscape interactive mapping site was downloaded for Spokane County. The shapefile contained gradient data for the streams but not confinement (Figure 2). Relative confinement was estimated using the 24K hydrography layer (water courses layer of the Washington Hydrography Framework) overlain on the 7½' USGS Quadrangle digital images (1:24,000 scale) 3 (Figure 2 for example). The SMA jurisdiction point layer identified the upstream end of shoreline jurisdiction. In Spokane County, eight streams totaling miles fall under shoreline jurisdiction (Table 2, Figure 2). All SMA streams have gradients 4% but confinement was variable (Figure 2). Table 2: Spokane County Streams Segments under SMP Stream/River Length (miles) Spokane River (including Long Lake portion) 59.0 Little Spokane River 39.2 West Branch Little Spokane River 3.9 Dragoon Creek 12.6 Deadman Creek 7.0 Latah ( Hangman) Creek 46.7 Rock Creek 15.6 Pine Creek Shapefiles publicly available from or Patricia Olson, PhD, LHG, SEA, Ecology 4 9/7/2007

5 Figure 2: Stream gradient shapefile was downloaded from the WDFW Salmonscape interactive mapping site. The map shows that all streams within SMP jurisdictions have gradients 4%. Relative confinement was estimated from USGS topographic images overlying the DEM. Task 2: Look for evidence of channel movement or bank erosion Indicators of channel movement may include (Figure 3): Multiple channels Channel bars Multiple bar forms Meandering form including oxbows, chute cutoffs Secondary channels -- o Side channels: persistent secondary channels, typically with vegetated islands or other persistent landforms separating it from the main channel o Relic channels -- those channels that may have flow in them only during very high water. They don't have to be connected to main channel but are still in the floodplain. Patricia Olson, PhD, LHG, SEA, Ecology 5 9/7/2007

6 Young disturbance vegetation Wood jams Bank erosion Oxbows Meander scars Chute cutoff Disturbance vegetation Figure 3: Lower Little Spokane River 1982 aerial photo, illustrates channel migration indicators including meandering form, oxbows, young disturbance vegetation, meander scars, and meander chute cut-offs. Method The 10-meter digital elevation model (DEM) obtained from Ecology s intranet GIS database 4 and orthophotos were used in this task. Recent georectified orthophoto layers (2004, 2006) were added to look for indicators of potential channel migration by comparing channels on orthophotos to the 24K hydrography streamlines. The 24K hydrography layer was used as a surrogate for the stream location on the topographic maps (Figure 4). Existing channel and riparian condition reports and Watershed Planning reports provided supplemental information on channel migration and bank erosion along the SMA streams (Spokane County Conservation District 2005, WRIAs publicly available at Patricia Olson, PhD, LHG, SEA, Ecology 6 9/7/2007

7 Figure 4: The 24K stream hydrography (purple line) overlying the USGS 7.5 quadrangle image shows that the 24K streamlines match the blue lines on the topographic map. Flow direction Figure 5: The 24K stream hydrography superimposed on 2004 orthophoto (Source: NAIP Ecology intranet GIS data) shows that migration is evident (yellow arrows). Other evidence of channel movement includes meandering form and oxbows [Dragoon Creek]. Patricia Olson, PhD, LHG, SEA, Ecology 7 9/7/2007

8 The data provides information to continue on one of 3 paths no assessment, channel migration and bank erosion assessment, or only bank erosion assessment (Figure 1). Both bank erosion and channel migration evidence was observed for most stream reaches. For the most part, bedrock control on erosion was limited (Spokane County Conservation District 2005). Common channel movement indicators were meandering form, meander bend cutoffs, channel bars, and secondary channels including oxbow ponds (Figures 3, 5). Task 3: Identify channel patterns and processes Channel patterns provide another indicator of the historic extent of channel migration and dominant erosion processes. Channel pattern is the most obvious expression of a rivers adjustment to changes in discharge and sediment which are influenced by climate and geology. Channel processes, such as lateral and downstream bend migration, chute cutoffs, oxbows and formation of secondary channels all indicate past channel migration (Figures 3, 5). For example, if a meander pattern is a more common pattern then we would expect to see mostly bank erosion on the outside meander bend and bend cutoffs (chute cutoffs). Whereas, if a braided pattern is common, then we would expect to see unvegetated bars that change position overtime. Changes are mostly due to change in sediment load, stream power, or channel gradient. Channel patterns can be divided into single and multi-channel planform. Single channel category includes straight and meandering without oxbows or side channels (Table 3). Straight channels are relatively rare and naturally limited to bedrock controlled reaches and some low gradient reaches. Globally, meandering rivers are the most common channel pattern. Leopold (1994) surveyed alluvial river valleys in the northwestern United States and found that meandering rivers occupied 90% of the total valley length. In multi-channel systems, there are meandering, braided and vegetated island braided patterns with anastomosing a subcategory of vegetated island braided pattern (Table 3). The most important difference between braided and vegetated island braided is degree of stability. Channels and bars are transient features in braided systems and can move on the order of minutes to hours. In comparison, the vegetated islands and natural levees of island braided systems can remain stable for hundreds of years resulting in significant lateral stability. Channel avulsion, sudden channel movement from one channel to another, is an important channel change process in island braided systems. Patricia Olson, PhD, LHG, SEA, Ecology 8 9/7/2007

9 Table 3: The following four channel patterns represent only the end members of a morphological continuum but discrete categories simplify this task. Channel pattern Straight single channeled. Meandering--usually single-channeled, but may have oxbows and side channels Description Single channel, sinuosity < 1.5, alternating bars Single channel, sinuosity >1.5, usually dominated by suspended sediment load. Channel cutoffs or avulsions can produce side channels. Multi-channeled, unvegetated: Braided Multi-channeled, vegetated: Island-braided, anastomosing, anabranching Multiple flow paths separated by transient bars, relatively stable bars, or islands, typically characterized by high or increased bed load supply Multiple flow paths separated by forested islands usually in gravel bed rivers in the PNW and overlaps with braided patterns. Method: The map produced in task 2 provides the information to evaluate channel types and processes. The Spokane River was not assessed because the majority is within the City of Spokane and is artificially or bedrock controlled thus exempting the incorporated portions from channel migration delineation [WAC (3) (b):]. The Spokane River downstream of City of Spokane is in a park and not developable. Also, river discharge is controlled by dams. All the Spokane County SMP streams exhibited some migration or bank erosion. Single channeled meander pattern is the dominant planform with some multichannel reaches. Multi-channel meandered planform with oxbows, chute cutoffs, and secondary and relict channels predominantly occurs along the Little Spokane River, especially in reaches from River Mile (RM) 1-9.5, 13-15, and 17. Anastomosed channel pattern is not evident within Spokane County. With a meandered pattern we would expect that the dominant processes are pool scour and erosion along the outside meander bend which leads to lateral and downstream channel movement. This appears to be the case (see Figure 5 for example and SCCD 2005 reports). However, braided channels do occur, mostly in the Little Spokane River but also in Latah, Deadman and Dragoon Creeks. Braiding within these streams most likely occurs due to a low-energy (low channel slope) setting and the associated reduction in available stream power in combination with an increase in sediment due to bank erosion. So in addition to increased sediment load from bank erosion, the reaches have reduced energy. The braided islands may also be bank slumps dissected by discharge. Consequently we would expect that the braiding will occur in areas of bank failure and below bank failure where channel gradient or stream power is low. Patricia Olson, PhD, LHG, SEA, Ecology 9 9/7/2007

10 Task 4: Determine assessment approach and method The draft CMZ web guidance includes matrices that provide guidance on the minimal assessment level based on environment and infrastructure values, management objectives, and predominant channel patterns (Figures 6-8). The lowest minimal standard may be appropriate for watershed characterization on the basin level in Spokane County (Figure 6). However, increasing development in potentially hazardous areas and high riparian corridor environmental values suggest a moderate level of effort (Figure 7). We incorporated moderate level elements because the dominant channel pattern is meandered with multichannel reaches, environmental values are high, and shoreline planning is a management objective (Figure 8). Moderate level elements used include approximating migration rates based on comparing stream traces between the topographic maps and more recent orthophotos. Figure 6: The numbers identify the linked summary of approaches and methods. M=meandered pattern, B=braided pattern, A=anastomosed pattern, W= wandering pattern. Figure 7: This chart provides criteria besides resources to determine the minimal level of effort on channel migration assessment for shoreline environmental and infrastructure values and potential channel movement occurring in terms of time Figure 8: Matrix cross-references channel migration assessments by scale (basin, reach, and site), management objectives, and level of effort Patricia Olson, PhD, LHG, SEA, Ecology 10 9/7/2007

11 Task 5: Conduct the assessment The approaches and methods used in this study follow the standards of practice matrix in the draft web guidance (partially shown in Table 4). Table 4: Describes approaches and method used to map the migration and bank erosion hazards for all the SMP streams except the Spokane River. The blue text summarizes data and analysis used in this assessment. Basin/watershed scale Approaches used 1. Low accuracy method Minimum Level of Effort Most applicable to basin or watershed Scale: all stream types Synthesize available data to understand control or relevant basin-scale processes such as geology, soils, topography/gradient, hydrology, land use, vegetation Information and data obtained from SMP inventory, characterization, WRIA reports, and Spokane County Conservation District (SCCD) shoreline inventories. GIS based with limited field verification 24k hydrography, USGS 7.5 quadrangles, SHIAPP channel gradient, 10-meter DEM, orthophotos, miscellaneous aerial photographs, SCCD shoreline inventory shapefiles, DNR geology, NRCS soil data. 3-day field observation 2. Moderate accuracy method Identify the purpose of the CMZ delineation To generally map the CMZ for the shoreline update Use available tools such GIS, DEM, LiDAR and hydrologic data to minimize costs Determine the characteristics and processes operating within the basin, including sources of sediment, climate and land use changes by researching floodplain & geologic and soils maps, and historic stream flow and climate data Much of this information was included in the WRIA and SCCD reports Identify possible hazard areas using DEM, LiDAR or aerial photographs, topographic maps Compared channel alignment between 2-3 time series using USGS 7.5 digital images, orthophotos Minimum Level of Effort Above plus the following CMZ Hazard zone=average rate of bank erosion per year x 100 years Mapped the 50 and 100 year buffers Suggested maximum buffer for meander and braided patterns: The probable migration area over 100 years plus the erosion hazard width or FEMA 100-year floodplain, whichever is greater Erosion hazard area Intersection between the 100-year CMZ and geology with high erosion potential Intersection between the 50-year CMZ and soil with Kffact 0.36 Patricia Olson, PhD, LHG, SEA, Ecology 11 9/7/2007

12 GIS Based Method The 24k hydrography layer was used as the historic stream layer. This layer was assumed to be developed from the USGS 7.5 quadrangles. We tested the assumption by overlaying the stream layer on the USGS quadrangle images and comparing the two for each stream. The assumption appeared to be the case (for example Figure 4). The orthophoto layers were used to determine channel changes from the 24K hydrography layer to recent conditions. The USGS topographic maps were developed from aerial photographs of varying dates. The date information for each quadrangle was obtained from the USGS. Dragoon and Deadman Creeks are relatively small (channel width <50 feet). This posed little problem for those reaches with sparse canopy cover. But where riparian conditions were properly functioning the channels were more difficult to see. For these reaches, the DEM and other air photos and maps were crossreferenced to check stream locations. Another issue was that the streams were generally smaller than their valleys where the valleys were formed by the Missoula glacial floods. While Task 3 showed evidence of channel movement and bank erosion, the stream power of present day streams, with the exception of the Little Spokane River, needed to be estimated. Stream power measures the rate of work done by flowing water and indicates the water s ability to erode the bank and bed sediments. Stream power estimates. Before migration rates were measured on Pine, Latah, Rock, Deadman, and Dragoon Creeks, stream power was estimated for each stream reach using: = gqs; where is water density, g is gravity, Q is the 2-year flood discharge, S is channel gradient and is stream power per unit channel length. For each reach, the upstream and downstream elevations were measured from the DEM. The difference in elevations divided by reach length provided the channel gradient. Where mean annual peak flow data were not available to directly calculate the 2- year flood for a reach, the 2-year flood was estimated from the USGS regional flood frequency equations (USGS Streamstats, Although the average bankfull discharge in Washington approximates the year flood frequency (Castro and Jackson 2002), the bankfull discharge varies substantially for eastern Washington streams. The average bankfull discharge recurrence interval for streams in the same flood hydrologic region as Spokane County streams is approximately the 1.75 year flood. Since there are no predictive equations for this flood recurrence interval, we assume that the 2- Patricia Olson, PhD, LHG, SEA, Ecology 12 9/7/2007

13 year flood is bankfull. Our definition for bankfull discharge is the discharge where sediment transport, channel movement and other geomorphic work is done. Empirical data suggests there is a relationship between the stream power index (Bankfull discharge*channel slope) and the median bed material size (e.g., Richards 1982). The empirical relation defines the threshold where stream power may initiate bed material movement. A relationship between the stream power index and sinuosity also exists (Richards 1982). Sinuosity, the ratio of channel length to valley length, is sometimes used to categorize single-thread channel patterns. For example, a meandering river generally has a sinuosity >1.5. This later relationship provides an indicator of a stream s tendency to meander. The stream power estimates suggest that the majority of SMP stream reaches have adequate power to move some of their bed load (Table 5). Sediment size distribution was not measured but descriptions in the SCCD shoreline reports (SCCD 2005) provide a basis for estimating sediment size ranges. Even if there is not enough power to move larger sediment such as cobbles, there may be adequate power to erode banks. The SCCD study reports that the predominant bank composition is non-cohesive soils and sediments that are finer than bed sediment size. Our field reconnaissance supports this. Many factors influence bank erosion, but generally banks of non-cohesive fine sediment will fail before more cohesive silt or clay banks. Migration rate estimates. After each streamline was traced from orthophotos, the linear channel change was measured between the years for each segment. The annual migration or erosion rate was estimated by dividing the average migration rate 5 for each reach by the time in years between the photo and map time series. The annual rate was converted to 50 and 100-year potential erosion distances by multiplying the annual rate by the appropriate number of years. The 50-year buffer provides an indicator of a higher hazard channel migration area. The 100-year buffer is suggested from Ecology guidance on channel migration areas (WAC (3) (b) :) (also refer to Rapp and Abbe 2003). The buffer tool in ArcGIS 9.2 toolbox was used to create 50-year and 100-year buffers around each segment. Polygons for the entire shoreline jurisdiction stream length were created from the buffers. The polygon lines were adjusted where the buffers extended past the valley wall and where buffers intersected existing secondary channels. SMA stream CMZ maps are in Appendix A. 5 We used the average rate because this was a general analysis. In more detailed channel migration analyses, the maximum migration rate is used. Patricia Olson, PhD, LHG, SEA, Ecology 13 9/7/2007

14 Since erosion potential of soils and underlying geology can influence channel migration and bank erosion, shapefiles of geologic formations and soil characteristics that intersected with the SMP streams were created. Geology qualitative erosion potential (low-high) was assigned to each lithology class in the geology layer. The data were obtained from the geology 6 and soils [NRCS, layers for Spokane County. Geology with high erosion potential and soil erodibility (K Factor, kffact) shapefiles attributes were intersected with the CMZ buffers using the intersecting tool of ArcToolkit. The geology data were intersected with the 100-year CMZ polygon because the 100-year CMZ may be controlled more by the valley wall geology. The soil kffact was intersected with the 50-year CMZ polygon (Figure 10). The areas identified may have higher potential for bank erosion or slope failure. Figure 9: The 50-year and 100- year CMZ buffers were combined with high geology erosion potential and high soil erosion potential (kffact 0.36) to identify potentially high bank erosion areas. 6 public access: < 7 A kffact 0.36 is on the upper end of values for silt loams which erode more than clay soils and have greater runoff values than sand. Patricia Olson, PhD, LHG, SEA, Ecology 14 9/7/2007

15 Table 5: A stream power analysis indicates, with the exception of a few reaches, that most streams have adequate power to mobilize bed and bank sediments. The sinuosity indicator also suggests that the stream pattern tends towards meandering. The stream reaches are defined in SCCD reports (2005). Reach number Q_2yr gradient Stream m 3 s -1 power (watts) Stream power index Estimated D50 in mm from power index Estimated sinuosity indicator from power index Qualitative sediment sizes in SCCD inventory (2005) Bed sediment size that could move at 2-yr flood Latah Creek hc none reported Very coarse (VC) gravel hc none reported Medium (M) gravel hc none reported VC gravel hc none reported VC gravel hc none reported VC gravel hc none reported Cobble hc none reported VC gravel hc none reported Cobble hc none reported Coarse (c ) gravel hc14a none reported VC gravel hc14b none reported VC gravel hc none reported VC gravel Rock Creek rc silts C gravel rc silts M gravel rc silts C gravel rc silts, sand C gravel rc silts, sand, cobbles M gravel Dragoon Creek dgc Sand (S), gravel (G) Fine (f) gravel dgc S, Gs Very fine (VF) gravel dgc S, G F gravel dgc S, G, cobble(c) VF gravel dgc S, G, C M gravel dgc S, G, C, C sand bedrock (BR) dgc S, G, C, BR M gravel dgc S, G, C, BR VF gravel dgc S, G, C, BR C gravel Pine Creek PC Silt, S, G C gravel Deadman Creek dmc S, G M gravel dmc S, G M gravel 6 dmc S, G M gravel dmc S, G, C, boulder C gravel Patricia Olson, PhD, LHG, SEA, Ecology 15 9/7/2007

16 Errors Data error can lead to under or over estimated migration rates. The lower level assessment method used assumes the maps will have lower accuracy than a more detailed assessment. We assume that the assessment only provides a general map of the channel migration areas. More detailed assessments, conducted by qualified, and preferably licensed, geomorphologists, hydrologists, geologists (including engineering and hydrogeology specialties) or hydraulic engineers should be required for regulatory purposes and future development. Since this assessment has a lower accuracy, the errors are generally discussed by not specifically quantified for this assessment. Geo-rectification error Aerial photograph and map geo-rectification causes some error. The 10-meter DEM and the USGS 7.5 digital quadrangles have a maximum error of 7 meters ( 22 feet). The exact error for the DEM and images used in this assessment is not known. Orthophoto rectification tends to have less error but it depends on the DEM quality. The error associated with the NAIP orthophotos is 3 meters ( 9.6 feet) with a 2.7 meter root mean square error. However, the draft 2006 orthophotos have poorer horizontal resolution than previous years. Corrected orthophotos will be available later this fall. For the Little Spokane there was approximately 30 ft difference between 2004 and Because of this error, we used the 2004 orthophotos as the most recent time series. Digitizing channel traces and width of lines Error associated with free-hand drawing and width of lines was reduced by correcting line errors and using the smallest scale orthophoto resolution would allow (approximately 1:2000). At the higher resolution, the error is less than 2 feet. Small stream size Small streams (channel width < 50 feet) can be hard to see where there is vegetation canopy obscuring the channel. To reduce mistaking stream location in areas with canopy, the DEM and other air photos and maps and on ground photographs were cross-referenced to check stream locations. However, the potential error is higher in these reaches than for reaches not obscured by vegetation or larger streams. Factors influencing SMA stream channel migration The geomorphology created by the Missoula floods exerts a major influence on the stream corridors by forming scoured valleys. Despite moderately confined nature of these valleys, channel migration does occur within the channel valley because the streams are much smaller than the ones that caused the scablands. Moreover, the glacial flood deposits tend to be erosion prone, increasing migration and bank erosion potential. Patricia Olson, PhD, LHG, SEA, Ecology 16 9/7/2007

17 For streams flowing through this environment, stream power is sufficient to move sediment during low magnitude high frequency floods (Table 5). Stream power relationships indicate that the streams have migration potential. The moderately confined reaches have room to move and comparison between map and orthophoto time-series provides evidence that the channels do move (e.g., Figure 5). Evidence of migration includes meander chute cutoffs, oxbows, secondary channels, and lateral and downstream migration (Figures 3, 5). In upper Latah and Rock Creek and Pine Creek, loess deposits are the primary influence on the channel movement. The loess soils tend to have high erosion potential leading to increased bank erosion especially where vegetation has been removed (e.g., SCCD 2005). The present day loess deposits are areas where sheet and rill erosion tends to account for almost 90 percent of the soil loss from cropland (USDA, 1978). In the loess dominated area, channel movement is mostly a response to channel incision and subsequent bank failure. The channel evolution model illustrates this response (Figure 10, Table 6). Summary-Stream descriptions Pine Creek The Pine Creek reach consists of 3.8 river miles within Spokane County. The SCCD properly functioning stream and riparian assessment indicated that the creek appears to be laterally and vertically stable pooled reach with low to moderate erosion. The results of the channel migration assessment generally concur with this observation. Some channel migration is evident upstream of the confluence with North Pine Creek (Appendix A). A relict channel appears in the orthophotos which may have been the original North Pine Creek alignment. Single channel and meandering describes the channel pattern. The PFC assessment indicates that the dominant bed materials are sands and gravels with bedrock in some areas. The stream banks are composed mostly of bedrock and unconsolidated, heterogeneous, non-cohesive materials. The stream power assessment indicates that there is sufficient power to mobilize coarse gravel during an approximate 2-year flood event. The sediment supply is moderate in this reach. Map (Appendix A) shows high soil erosion potential along the entire reach. The primary soil is loess, with high silt content, which generally has high erosion potential especially where banks have been disturbed. Patricia Olson, PhD, LHG, SEA, Ecology 17 9/7/2007

18 Figure 10: Schumm et al (1984) developed a conceptual channel evolution model for degraded alluvial channels. The model describes the systematic response of a degraded channel over time. Table 6: Descriptions for the channel evolution model schematic in Figure 10. Type I Sediment transport capacity > sediment supply. For example, an increase in discharge increases sediment capacity. Bank height (h) < critical bank height (hc). U-shaped channel cross-section. Type II Immediately downstream of knickpoint. A knickpoint is an abrupt change in thalweg elevation, and may be visualized as a small rapid or waterfall. Sediment transport capacity > sediment supply. h< hc Bed slope < than Type I because water depth has increased Type III Sediment transport capacity varies with respect to sediment supply h>hc with bank erosion often due to slab failure Bank loss rates at a maximum variable sediment deposition may start forming bars Channel depth generally > Type II depth Channel widening due to bank failure Type IV Sediment supply>transport capacity resulting in bed aggradation h approaches hc Less bank failure than Type III & failure may change from slab to circular-arc failures Channel widened Berm, natural levees at the edge of the effective discharge channel may form Type V Sediment supply and transport capacity in balance dynamic equilibrium h<hc vegetation is colonizing bank with decreased bank angle from accumulation of failed bank materials at toe slope aggraded channel channel sinuosity increases newly formed floodplain Patricia Olson, PhD, LHG, SEA, Ecology 18 9/7/2007

19 Figure 11: Channel migration is evident along Pine Creek upstream of its confluence with North Fork Pine Creek. The yellow streamline is from the 24K hydrography layer and the blue streamline is traced from the 2004 orthophoto. A relic channel is visible in the upper left corner of the photo. It may have been the original alignment of the North Fork. The blue line indicates direction of flow. Latah Creek (Hangman Creek) The SCCD report (2005) identified Latah Creek as the most critical system in Spokane County. The primary channel process is bank erosion along longitudinal vertical bank with areas of slumping caused by toe erosion with areas of lateral migration (Figure 11). The stream is not laterally and vertically stable except in areas of bedrock control and artificial constraints (e.g. road corridors). Stream bank slumping and erosion has resulted in significant braiding in some reaches. The SCCD report states that significant erosion and inadequate or absent riparian plant communities were characteristic, especially above the canyon (RM 34.2) and near the confluence with Rock Creek where the stream is naturally meandering. The report also indicates that the majority of Latah river corridor has land uses that are generally incompatible with high erosion prone soils. The potential bank erosion map (Appendix A) shows occurrence of erosion prone soils and geology. The combination enhances channel instability as evidenced by widespread streambank erosion (SCCD 2005) in most reaches. Vertical eroding banks and toe erosion resulting in bank slumping and channel widening are Patricia Olson, PhD, LHG, SEA, Ecology 19 9/7/2007

20 common (Photo 1). Our field reconnaissance concurred with this observation as do the recent orthophotos (Figure 12). Figure 12: A reach in upper Latah Creek with bank erosion and braiding (within orange oval) is identified on the 2006 orthophoto. The yellow line is the streamline from the 24K hydrography layer. The photo illustrates the change in channel alignment due to bank recession and downstream migration of meander bend. The blue arrow indicates the flow direction. Photo 1: Photograph taken June 2006 shows bank erosion and braiding along Latah Creek. Rock Creek The SCCD report indicates that much of Rock Creek channel and riparian conditions are in fairly good condition. Exceptions are above Rockford and near the confluence with Latah Creek. Upstream of Rockford, the stream flows Patricia Olson, PhD, LHG, SEA, Ecology 20 9/7/2007

21 through the erosive Palouse Formation and loess soils. Here the stream is moving laterally and widening due to past incision down to bedrock (e.g., type III channel in the channel evolution model, Figure 10, Table 6). Even in the properly functioning reaches within the Rock Creek basalt canyon, bank erosion and lateral movement occurs but at a slower average rate between photo and map time series. The mean migration rate =3.8 feet for the PFC reaches whereas the FAR reaches mean migration rate = 4.6 feet) (Figure 13). The stream power assessment concurs with this observation indicating there is sufficient power to move sediment larger than that reported by SCCD (2005) (Table 5). The relationship between stream power and sinuosity (Table 5) also suggests the stream has the potential to migrate. Figure 13: Natural channel migration is evident along a relatively undisturbed reach of Rock Creek. The yellow streamline is from the 24K hydrography layer and the blue streamline is traced from the 2004 orthophoto. The blue arrow indicates flow direction. Deadman The existing SMP only regulates the first seven miles of Deadman Creek but the new SMA point has moved upstream to RM 9.7. We determined the CMZ buffers up to that new point. The outburst flood deposits from Glacial Lake Missoula influence the reaches downstream of the new SMA point. While the flood deposits have high erosion potential, bank erosion is not common where riparian vegetation is still intact. Dominant bed materials are sand and gravel. Banks are unconsolidated, heterogeneous, non-cohesive materials. The stream power assessment suggests that the stream can transport a large fraction of its bedload and erode banks (Table 5). While the SCCD report (2005) found the stream to be laterally and vertically stable, stability does not preclude channel migration especially in streams prone to meandering such as Deadman Patricia Olson, PhD, LHG, SEA, Ecology 21 9/7/2007

22 Creek. Overtime, the stream has migrated mostly in a lateral direction and reoccupied side channels but downstream migration is also evident (Figure 14). Figure 14: Natural lateral and downstream channel migration and channel avulsion into a side channel is evident along Deadman Creek. The yellow streamline is from the 24K hydrography layer and the blue streamline is traced from the 2004 orthophoto. The blue arrow indicates flow direction. Dragoon Creek Dragoon Creek flows through Pleistocene glaciolacustrine and outburst glacial flood deposits. Above RM 8.0 the dominant bed materials range from sand, gravel, and cobble where the stream meanders through glacial outwash. The bank materials are unconsolidated, heterogeneous, non-cohesive materials that are finer than the bed material. The SCCD report (2005) suggests that the soil has moderate erosion potential. The GIS generated geology erosion potential indicates that the geologic materials have high erosion potential (Appendix A) which encourages bank erosion, channel meandering and subsequent lateral and downstream migration, chute cutoffs and formation of oxbows (Figure 5, Photo 2). Increased development, including riparian clearing, rural residential homes, railroads, parks, golf courses, highways, and roads, apparently have increased bank instability and migration (SCCD 2005). Downstream of RM 8.0, Miocene Grand Ronde Basalt and landslides influence the stream where bed material changes from predominantly sand and gravel to a Patricia Olson, PhD, LHG, SEA, Ecology 22 9/7/2007

23 mix of sand, gravel, cobble, boulder and bedrock. The stream power assessment indicates that the stream can only move the sand and gravel fraction during the 2-year flood event (Table 5). However there are also unstable bluffs that are part of a landslide complex. The soil on these bluffs have high erosion potential (Map, Appendix A) and become more susceptible to erosion and failure when vegetation is removed. Grazing also appears to affect their stability. The steepness of the bluffs increases erosion potential. Photo 2: Bank erosion and meander bend formation within the middle reaches of Dragoon Creek. Little Spokane River The Little Spokane River has the most diversity in channel patterns and processes because of the landscape diversity. The Little Spokane, above Dragoon Creek is predominantly influenced by the Missoula glacial flood deposits. These deposits have a high erosion potential. Dominant bed materials are sand, gravel and cobble where the stream flows through the glacial flood deposits. The banks are unconsolidated, heterogeneous, non-cohesive materials, dominated by silt, so are finer than the bed material. Silt has a high Kffact and erosion potential (Appendix A). While glacial flood deposits enhance stream meandering potential, the high soil erosion potential makes bank erosion the dominant channel change process above RM 15. The SCCD report (2005) indicates that bank erosion increases where livestock have unrestricted access to the stream. Bank erosion also accelerates where vegetation has been removed, mostly during rural residential Patricia Olson, PhD, LHG, SEA, Ecology 23 9/7/2007

24 development. Increased bank erosion leads to braided sections due to increased sediment load. Development along the stream includes shoreline modifications such as bulkheads, landscaping to the water s edge and sometimes illegal fill (Photo 3). These conditions lead to alterations in high water capacity and higher migration rates especially where vegetation is removed and upstream and downstream of bank hardening. The end result is larger CMZ buffers. Photo 3: Illegal fill within the Little Spokane floodplain and channel migration zone alters flow dynamics and affect upstream migration rates. Photo was taken June Upstream of RM 15, lateral and downstream movement also occurs. Channel pattern is slightly meandering interspersed with straight reaches particularly in areas of bedrock control (Figure 15). Beaver activity likely influences the formation of secondary channels in the upper reaches. Downstream of RM 15, the character of the Little Spokane River changes to predominantly meandering stream with a wide floodplain. Lateral and downstream migration and meander chute cutoffs are the dominate channel change processes (Figure 3). Indicators include oxbow ponds, meander scars and disturbance vegetation such as the Black cottonwood (Populus trichocarpa) galleries and high diversity in scrub-shrub communities (Figure 3). Black cottonwood relies on flooding and channel movement to maintain communities. Channel migration also promotes vegetation diversity. Patricia Olson, PhD, LHG, SEA, Ecology 24 9/7/2007

25 Land management practices such as unlimited stream access to livestock appear to increase bank erosion producing vertical cutbanks (SCCD 2005). Residential houses along the banks also increase lateral instability. Despite these disturbances, this portion of the stream is properly functioning. Figure 15: The dominant channel pattern for the middle reaches of the Little Spokane River is meandering interspersed with straight channels. The yellow streamline is from the 24K hydrography layer and the blue streamline is traced from the 2004 orthophoto. Bank erosion is the dominant process for channel change between the 2 time series. The blue arrow indicates flow direction. Patricia Olson, PhD, LHG, SEA, Ecology 25 9/7/2007

26 References Castro, J.M., and P.L. Jackson, 2001, Bankfull discharge recurrence intervals and regional hydraulic geometry relationships: patterns in the Pacific Northwest, USA. Journal of the American Water Resources Association 37(5): Leopold, L.B., 1993, a View of the River: Berkeley, California, University of California, 298 p. Rapp, Cygnia and Timothy Abbe, 2003, A Framework for Delineating Channel Migration Zones, Ecology Publication , Richards, K 1982, Rivers: Form and Process in Alluvial Channels, Methuen & Co, NY, NY USA, pp Schumm, S. A., M. D. Harvey, and C. C. Watson, 1984, Incised Channels: Morphology, Dynamics and Control, Water Resources Publications, Littleton, Colorado. Spokane County Conservation District (SCCD) 2005 Spokane County proper functioning condition stream inventory & assessment, funded by Centennial Clean Water Fund Grant # G , Washington State Department of Ecology, 340 pp U.S. Department of Agriculture (USDA), Forest Service, and the Soil Conservation Service, Palouse Co-operative River Basin Study, U.S. Government Printing Office, , 182 pp. WRIA 56, Hangman (Latah) Creek Watershed Planning Unit, 2005, The Hangman (Latah) Creek Water Resources Management Plan, May 19, 2005, Funded under, Watershed Planning Grant # G , Washington State Department of Ecology, pp 153. Patricia Olson, PhD, LHG, SEA, Ecology 26 9/7/2007