Guidance for Design Hydrology for Stream Restoration and Channel Stability (2017)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

2C h a p t e r 2 2.1 Overview of the Design Hydrology Process There are two primary phases to the design hydrology for stream restoration and channel stability at stream crossings (hereafter “design hydrology”) process: • Phase 1: Assess the Current Conditions Adjacent to the Stream Crossing and in the Water- shed to Determine Design Effort. This phase includes gathering available data related to hydrology, land use, stream bed material, and current channel conditions, as well as the avail- ability of a stable analog reach, to determine the appropriate level of design hydrology analysis. • Phase 2: Design the Stream Channel Through the Stream Crossing. After the appropriate type of analysis has been determined, a set of analytical and analog tools are used to perform the appropriate hydrological analysis and ultimately design a stable channel. Each phase relies on a mix of decision support and software tools, and requires input from desktop, field, and design phases. Additional details on the scientific basis and development of these tools are provided in Colorado State University’s final report for NCHRP Project 24-40 [downloadable from the NCHRP Research Report 853 summary page on the TRB website (www. trb.org)]. The goal of this process is to identify the appropriate tools and level of analysis given the conditions of each stream crossing. The first step in hydrologic design for stream crossings is to define goals and objectives with respect to channel stability, sediment continuity, flood conveyance, environmental considerations, and many other factors. Once goals and objectives are defined, the two phases can be initiated. 2.2 Overview of Phases 1 and 2 The following subsections present the design hydrology processes for Phases 1 and 2: • Phase 1 (Section 2.3): – Subsection 2.3.1: Bed Material versus Flashiness – Subsection 2.3.2: w* Versus Flashiness – Subsection 2.3.3: Simplified Rapid Geomorphic Assessment – Subsection 2.3.4: Analog Reach Guidance – Subsection 2.3.5: Selection of the Design Hydrology Approach • Phase 2 (Section 2.4): – Subsection 2.4.1: Establish a Sediment Supply Reach – Subsection 2.4.2: Evaluate Whether Additional Field Reconnaissance Is Needed – Subsection 2.4.3: Perform Channel Design Using the Set of Recommended Methods – Subsection 2.4.4: Compare Channel Designs to Analog Design(s) – Subsection 2.4.5: Select a Robust Design The Design Hydrology Process

the Design hydrology process 3 2.3 Phase 1: Assess the Current Conditions Adjacent to the Stream Crossing and in the Watershed to Determine Design Effort Both Phases 1 and 2 can be related to the overall decision table found in Figure 2-1. Phase 1 is the process of relating stream response potential (SRP) and the availability of an analog reach (also referred to as a reference reach) to determine the appropriate level of design analysis. This phase incorporates the following questions: (1) How does the availability of an analog reach change the level of design guidance? (2) What level of hydrologic analysis should be undertaken? (3) Is it appropriate to perform sediment transport analysis, and, if so, what type of analysis is needed? (4) What spatial domain (i.e., how far upstream and/or downstream from the project location) is recommended for conducting the analysis? The fundamental philosophy underlying this approach is that, as SRP increases, it becomes necessary to conduct a deeper analysis over a larger area of the stream and its watershed. At the core of this phase is the evaluation of SRP, which is positively correlated with the amount of design effort (Figure 2-2). The multiple methods to evaluate SRP are discussed in the following subsections. 2.3.1 Bed Material Versus Flashiness The first-cut estimate of the SRP is based upon a visual quantification of the channel bed material adjacent to the stream crossing and the flow regime flashiness (i.e., the frequency and rapidity of short-term changes in streamflow)—as computed in the desktop phase. The general trend related in Table 2-1 is that less erosive bed material and low levels of flashiness typically result in lower SRP than live-bed systems with high flashiness. FDC = flow duration curve; Qeff = effective discharge; Qs50 = discharge associated with 50% of cumulative sediment transport over the sorted flow record; CSR = capacity-supply ratio. If a field rapid geomorphic assessment indicates high or very high susceptibility and response potential in the design reach, then shift to the next higher level of stream response potential and design analysis. Figure 2-1. Decision table providing guidance on the level of design hydrology analysis.

4 Guidelines for Design hydrology for Stream restoration and Channel Stability 2.3.2 v* Versus Flashiness A fundamental physical relationship that is missing from Table 2-1 is a ratio of flow energy relative to boundary material resistance. A more physically based alternative to Table 2-1 quanti- fies flow energy relative to dominant bed grain size using dimensionless specific stream power (w*), a robust predictor of sediment transport capacity. However, the tradeoff is that field data on grain size will be required for the accurate estimation of dimensionless specific stream power. When it is feasible to collect representative grain size data, the following approach can provide a more rigorous assessment of response potential. Dimensionless specific stream power is defined as: 1 (2-1) 50 3 2g G D  [ ]( )ω = ω ρ − where: w = specific stream power [W/m2] = rgQS/w ; where: r = density of the fluid mixture [kg/m3], g = gravitational acceleration [m/s2], Q = median annual peak flow (Q2) [m3/s], S = channel slope [m/m], and w = channel top width [m]; Lower SRP/Effort: Higher SRP/Effort: Bed Material and/or ω* Versus Flow Flashiness (Tables 4-1 and 4-2) Analog Reach Rapid Geomorphic Assessment as determined by The boxes at the bottom of the arrow are multiple methods to help evaluate the stream response potential. ω* is defined as the dimensionless specific stream power. Figure 2-2. Determination of system risk and associated design effort. Flow Regime Flashinessb Bed Materiala R-B Indexc Ä 0.2 0.2 < R-B Index Ä 0.5 0.5 < R-B Index Boulder / resistant hard pan Low Low Medium Armored cobble / coarse gravel with assorted sizes tightly packed, overlapping, and possibly imbricated; most material > 4 mm (0.16 in.); Fs < 20%, mostly boulders / cobbles / coarse gravel Medium Medium High Transitional: unarmored containing moderately packed to loose assortment with 20% < Fs < 50% Medium High Very High Live bed: very loose assortment with no packing; large amounts of material < 4 mm (0.16 in.); Fs > 50%, mostly sand and finer High Very High Very High a Fs = approximate fraction of sand in bed sediments. b For braiding and/or rapid urbanization (~10% increase in urban land cover per decade), move to next higher category. c R-B Index = Richards-Baker Flashiness Index. Table 2-1. SRP decision table used to define classes corresponding to different design hydrology strategies based on bed material and flow regime flashiness.

the Design hydrology process 5 G = specific gravity of sediment (2.65 is typically assumed); and D50 = median grain diameter of the bed material [m]. Note that for live-bed channels with fine bed materials dominated by sands, silts, and clays, it is recommended to defer to the susceptibility class from Table 2-1 as opposed to comput- ing the class associated with specific stream power (Table 2-2). Discussion and specifications on how grain sizes are sampled are found in Bunte and Abt (2001); note that approaches differ between channel types (e.g., sand versus armored gravel). Also, for guidance on combining sieve and pebble count data, see Bunte and Abt (2001). If Table 2-1 and/or Table 2-2 indicate a low SRP and the system shows no signs of instability based on Hydraulic Engineering Circular No. 20 (HEC-20; Lagasse et al. 2012), it may be possible to proceed to Phase 2. However, if the appropriate SRP for the stream crossing is still unclear, the tools described in Subsections 2.3.3 and 2.3.4 provide additional lines of evidence to support making a decision on the appropriate level of design. 2.3.3 Simplified Rapid Geomorphic Assessment The guidance and figures in this subsection support rapid geomorphic assessments (RGAs) of channel instability and susceptibility at stream crossings. This overtly simple approach is intended to complement more comprehensive and rigorous methods, most notably HEC-20, by orienting engineers to some key considerations during field reconnaissance early in the design hydrology process. If the RGA indicates high or very high susceptibility and response poten- tial in the design reach, then it is recommended that the designer shift to the next higher level of SRP and design analysis. To develop a simplified RGA, the research team reduced a large pool of potential indicators to four: (1) Current stability status—Channel Evolution Model (CEM; Schumm et al. 1984) stage, braiding, alluvial fan (2) Dominant bed material/armoring potential (3) Distance to downstream hardpoint/grade control (4) Bank strength High ratings of stream susceptibility based on these indicators trigger a higher level of design hydrology analysis as defined by the design decision table in Figure 2-1, and underscore the need for a greater stability analysis using more rigorous and comprehensive tools such as HEC-20. A further understanding of the SRP for a stream crossing can be developed through an RGA. The first consideration in the simplified RGA procedure is to identify early “off-ramps” (Figure 2-3)—characteristics that are indicative of fluvial geomorphic extremes. Streams not Flow Power at Q2 Relative to Bed Material ( * ) Flow Regime Flashiness R-B Index Ä 0.2 0.2 < R-B Index Ä 0.5 0.5 < R-B Index * << O[0.1] Low Low Medium * ~ O[0.1] Medium Medium High * ~ 0.3 to O[1] Medium High Very High * ~ O[1] or higher High Very High Very High O = on the order of. ω* = dimensionless specific stream power. ω ω ω ω v Table 2-2. SRP decision table used to define classes corresponding to different design hydrology strategies based on dimensionless specific stream power at the median annual peak flow (Q2) and flow regime flashiness.

6 Guidelines for Design Hydrology for Stream Restoration and Channel Stability exhibiting the early off-ramp conditions would warrant examination of the secondary factors. For example, a stream bed dominated by medium to coarse gravels could be as sensitive as a very high risk category if its banks were weak (alluvium lacking vegetation) and it lacked grade control. However, the same stream bed could be low risk if it had strong banks (bedrock/ boulder) and frequently spaced grade control. Cases in between would be either medium or high risk according to Figure 2-3. Likewise, bed material dominated by small cobbles/very coarse gravels would range from low to high risk depending on bank strength and hardpoint frequency, and beds dominated by large cobbles would range from low to medium risk (Figure 2-4). Channel responses may propagate for significant distances downstream (and sometimes upstream) from a point of influence such as a stormwater outfall or stream crossing. Accord- ingly, it may be necessary to conduct field reconnaissance across a domain spanning multiple channel segments and property owners. The research team recommends that the typical analysis Figure 2-4. RGA risk categories for beds ranging from coarse gravels to large cobbles across a gradient of bank strength and hardpoint (grade control) frequency. Low Medium High Very High Boulder Sand/Fine Gravel CEM Stage III Braiding Figure 2-3. Early off-ramps in the RGA of low risk (boulder-dominated stream beds) and very high risk [sand/fine gravel-dominated stream beds, CEM Stage III (channel incised past critical bank height for geotechnical failure and mass wasting), or active braiding].

the Design hydrology process 7 domain for conducting the RGA should be at least 20 channel widths upstream and downstream in accordance with the recent Caltrans (2015) guidance. Begin by defining the points or zones along the channel reach(es) where changes in discharge or channel type are likely to occur (e.g., potential locations of outfalls or tributary inputs). Docu- ment any observed outfalls for final desktop synthesis and define the minimum upstream and downstream extents of analysis as follows: • Upstream—for a distance equal to 20 channel widths or to grade control in good condition— whichever comes first. Within that reach, identify (1) hardpoints that could check headward migration and (2) evidence that headcutting is active or could propagate unchecked upstream. • Downstream—until reaching the closest of the following: – At least one reach downstream of the first grade control point (but preferably the second downstream grade control location) – Tidal backwater/lentic waterbody – Equal-order tributary (Strahler 1952)1 – A two-fold increase in drainage area2 This (practicality-driven) guidance should not supersede the consideration of local conditions and sound engineering judgment. Within the analysis domain, there may be several reaches that should be assessed indepen- dently based on either length or change in physical characteristics. In more urban settings, seg- ments may be logically divided by other stream crossings, which may offer grade control, create discontinuities in the conveyance of water or sediment, etc. In more rural settings, changes in valley/channel type, natural hardpoints, and tributary confluences may be more appropriate for delineating assessment reaches. In general, the following criteria should trigger delineation of a new reach and hence a separate susceptibility assessment: • 200 m or ca. 20 bankfull widths—it is difficult to integrate observations in the field over longer distances • Distinct or abrupt change in grade or slope due to either natural or artificial features • Distinct or abrupt change in dominant bed material or sediment conveyance • Distinct or abrupt change in valley setting or confinement • Distinct or abrupt change in channel type, bed form, or planform 2.3.4 Analog Reach Guidance The analogy method has sometimes been used recklessly in design: Streams from different watersheds and even different physiographic regions with disparate hydrologic and sediment supply characteristics were used to define channel geometry in dissimilar settings. The decision support tool in Tables 2-3 and 2-4 helps users to identify upstream analogs that are very similar in terms of key criteria—such as the valley setting, boundary conditions, and inflowing loads of water and sediment—and to define supply reaches for sediment continuity analysis. A good analog reach can serve to inform the level of analysis that is appropriate for a project’s design hydrology; however, finding a suitable analog can be challenging. The four criteria in 1In the absence of proximate downstream grade control or backwater, the confluence of an “equal-order tributary” should correspond to substantial increases in flow and channel capacity that should, in theory, correspond to significant flow attenu- ation; however, there is no scientific basis to assume that downstream channels of higher stream order are less susceptible than their upstream counterparts. 2An increase in drainage area greater than or equal to 100% would roughly correspond to the addition of an equal-order tributary.

8 Guidelines for Design hydrology for Stream restoration and Channel Stability Table 2-3 are required to be met. If the analyst is unable to find an appropriate analog reach with a similar drainage area, similar channel type, and similar hydrology and that is stable, then exclusive reliance on an analytical design method is recommended. The eight criteria in Table 2-4 are important; hence, it is recommended that at least six of these criteria are satisfied to ensure the analog reach is an appropriate analog. If the analog reach is not on the same river, then the first question in Table 2-4 does not apply and that criterion is not met. The research team defines “stable” after Biedenharn et al. (1997): “In summary, a stable river, from a geomorphic perspective, is one that has adjusted its width, depth, and slope such that there is no significant aggradation or degradation of the stream bed or significant planform changes (meandering to braided, etc.) within the engineering time frame (generally less than about 50 years).” 2.3.5 Selection of the Design Hydrology Approach Phase 1 culminates in selection of the design hydrology approach. Use (1) the SRP for the stream crossing, (2) the existence (or absence) of a good analog reach, and (3) the outcome Topic Question Criterion Met If: Context References Flow Regime Do the sites have similar drainage area (within 20%)? Yes Rivers and streams are scaled in size by their drainage areas. • Environmental Resource Assessment & Management System (eRAMS; erams.com) online tool for delineation (simplified version in development) • StreamStats online tool [U.S. Geological Survey (USGS) 2012] • Arc Hydro Tools ArcGIS Toolkit Flow Regime Is the hydro- climatic system the same (i.e., how and when does the precipitation come: snow, winter rain, convective rain, monsoon)? Yes This is often less troublesome if the analog reach is close to the restoration reach, but can become an issue in mountainous areas with strong orographic effects (i.e., wet and dry side of the mountains). • Cheng et al. (2012) • Poff (1996) • Reidy Liermann et al. (2012) • Sawicz et al. (2014) Channel Type Are the channel types the same? Yes Target channel type represents: (1) prevailing historical channel type that was previously stable (diagnose why departure occurred) in that location under current land use; OR (2) channel type is stable under same current land use, flow, and sediment supply in analog reach. • See Table 3-2 • Church (2006) • Lagasse et al. (2012) • Montgomery and Buffington (1997) • Rosgen (1994) Stability Is the analog reach largely stable? Yes CEM Stage I or V per Schumm et al. (1984) banks stable, no evidence of trends in aggradation/degradation, planform change, etc. over engineering time scales. • Lagasse et al. (2012) • Schumm et al. (1984) • Hawley et al. (2012) Table 2-3. Required criteria for analog reach selection. The analog reach must meet 100% (4/4) of the criteria.

Flow/ Sediment Regime Alterations Are there any noteworthy tributaries, dams, or intervening flow augmentations or extractions? No Tributaries, dams, and flow augmentations or extractions can initiate changes in the flow and sediment regimes. Learn more about the National Hydrography Dataset Viewer (http://nhd.usgs.gov/index.html) Valley Type Energy Is the valley stream power (defined as the Sv * Q20.5, where Sv = valley slope and Q2 = 2-year return interval discharge) similar (within 20%)? Yes Stream power is the stream’s ability to do work including the entrainment and transport of sediment. Desktop estimates for the required parameters can be performed using: • Valley slope from mapping tool that includes elevation (e.g., Google® Earth™) and • Q2 (for most locations in the United States) è StreamStats (USGS 2012) • Bledsoe and Watson (2001) • van den Berg (1995) Valley Type Lateral Constraints Are the lateral constraints (i.e., the influence or connectivity of the valley walls) similar? (Is the ratio of floodplain width to channel width within 30% between the analog and project reaches?) Yes The narrower and steeper the valley walls, the more connection (and influence) they will have on a river’s planform, sediment inputs, and ability to self-adjust. • Desktop estimates can come from Google Earth imagery • Nanson and Croke (1992) • Whiting and Bradley (1993) Flow Regime Same hydrologic flashiness (within 30%)? Yes Flashiness (i.e., the frequency and rapidity of short-term changes in streamflow) • Estimates for gaged sites can be found using the Flow Analysis toolkit in eRAMS (variable currently in downloadable data summary) • Baker et al. (2004) Topic Question Criterion Met If: Context References Location If on the same river, is the analog reach upstream of the project reach? (If analog reach is not on the same river, that criterion is not met.) Yes Restoration is often in response to disequilibrium (or instability), thus if instability exists at the restoration site, it is likely that this instability may persist downstream. Learn more about the National Hydrography Dataset Viewer (http://nhd.usgs.gov/index.html) Land Use Are the extent and nature of land use (e.g., curve number) similar between the two watersheds (within 20%)? Yes Watershed land use influences both flow regime and sediment supply in a river. • Can use SWAT-DEG (Soil and Water Assessment Tool— channel DEGradation) in eRAMS (currently in beta version) to estimate composite curve number for a watershed • NRCS (1986) Geologic Setting Analogous physiographical region / geologic setting with respect to topography / valley slopes, soil types, and vegetation cover? Yes These watershed characteristics influence the magnitude and timing of runoff. • Booth et al. (2010) • Reid and Dunne (1996) • Vigil et al. (2000) Bed Surface Sediment Characteristics Are the bed surface grain size distributions similar (do not differ by more than ± one half phi class for Yes The bed surface grain size is linked to sediment supply [e.g., Dietrich et al. (1989)]. • Bunte and Abt (2001) influences the stability of a river channel. D50 and D84) Table 2-4. Important criteria for analog reach selection. The analog reach must meet 75% (6/8) of the criteria.

10 Guidelines for Design hydrology for Stream restoration and Channel Stability from the RGA analysis to select the most appropriate combination of analyses (Figure 2-1) and transition to Phase 2: Design the Stream Channel Through the Stream Crossing. Note that if the RGA indicates high or very high susceptibility and response potential in the design reach, then it is recommended that the designer shift to the next higher level of SRP and analysis. 2.4 Phase 2: Design the Stream Channel Through the Stream Crossing Phase 2 begins when the design analysis has been specified and includes a broad range of tools each targeted to a specific type of design. It is important for the designer to keep in mind that there are a range of channel design tools that each have their own inherent strengths, weaknesses, and assumptions. Thus, as the overall project risk increases, it is recommended that the number of evidence lines used in the final design increase as well (Figure 2-5). 2.4.1 Establish a Sediment Supply Reach If flow duration curve (FDC)/half-load discharge (Qs50) or FDC/capacity-supply ratio (CSR) sediment analysis is recommended based on Figure 2-5 and RGA, establish a sediment sup- ply reach in the zone upstream of the crossing. This step includes collecting bed-material data and channel geometry [profile and cross section(s)] within the sediment supply reach. If the upstream reach is currently unstable, attempt to collect data within a subreach that appears clos- est to stability (if possible). Try to select a reach that (1) appears to be transporting its sediment without appreciable downcutting or aggradation and (2) is far enough upstream from hard- points (culverts, exposed bedrock, etc.) to avoid a depositional reach that could have overly fine bed material (Figure 2-6). If the sediment supply reach is currently stable, it may serve as the best analog reach for the design (i.e., there is no need to find another analog). If the sediment supply reach cannot also serve as the analog reach, establish an analog reach to meet as many suitability criteria as possible (Tables 2-3 and 2-4), and survey the analog if existing analog data are insufficient. Peak Discharge Analog Reach Daily Flow Analysis Peak Discharge Analog Reach Peak Discharge Capacity-Supply Ratio (CSR) Sediment Transport Rates Analog Reach Check Sediment Balance at Qs50 Daily Flow Analysis Sub-daily Flow Analysis Lower SRP/Effort: Higher SRP/Effort: Blue = hydrologic analysis Brown = sediment transport and field analysis Figure 2-5. Lines of evidence required for design increase with stream response potential.

the Design hydrology process 11 When selecting the sediment supply reach, the designer should use the reach or subreach that is most representative for sediment continuity. Avoid scour areas immediately downstream of headcuts or hardpoints, as well as aggradational areas immediately upstream of hardpoints. Subreaches that appear to be transporting their bedload without incision or aggradation (even temporarily) are more representative than segments that are more clearly downcutting or aggrading. temporarily balanced Qs? original channel bed scour aggradation hardpoint Figure 2-6. Selecting the sediment supply reach. 2.4.2 Evaluate Whether Additional Field Reconnaissance Is Needed Additional field reconnaissance may be needed to provide requisite data for performing the recommended level of analysis [see Subsection 3.3 for the input requirements for the CSR Sta- ble Channel Design Tool (CSR Tool)]. Perform additional bed material and geometric surveys as necessary. The field reconnaissance steps described previously are not intended to supplant engineering judgment and existing guidance. Instead, this approach is designed to be a rela- tively simple and user-friendly complement to the more comprehensive procedures described in HEC-20 (Lagasse et al. 2012). 2.4.3 Perform Channel Design Using the Set of Recommended Methods For FDC/Qs50 design, apply the decision tree in Figure 2-7 and Environmental Resource Assessment & Management System (eRAMS; erams.com) guidance to compute Qs50. For full FDC sediment analysis design based on CSR, apply the CSR Stable Channel Design Tool (Subsection 3.3 and Chapter 4). Details on how to approach each node in the decision tree, and examples of how Qs50 is cal- culated using the decision tree and the tools, are provided in Subsection 3.2. Note that Q1.5 is typically a reasonable surrogate for Qs50 in coarse-/armored-bed channels. If CSR design is recom- mended, generate FDC using eRAMS or other means and apply the CSR Stable Channel Design Tool using appropriate inputs and selected sediment transport equation per guidance (in Sub- section 3.3 and Chapter 4). The decision tree presents a series of questions regarding land use change, potential non-stationarity of the flow record, and the availability of stream gage data and sediment transport measurements to guide the user toward the best approach for calculating Qs50. 2.4.4 Compare Channel Designs to Analog Design(s) Compare channel designs, via the Qs50 and/or CSR Tool under current design hydrology, to analog design(s), if available. The FDC input to the CSR Tool can reflect current or projected

12 Guidelines for Design hydrology for Stream restoration and Channel Stability Is the site gaged? Yes No No No Yes Yes Is the gage record stationary? Is there a more recent subset of stationary data with sufficent length? Does the site have sediment transport measurements? Create sediment rating curve, extract beta Do you have channel geometry, slope measurements? Yes No Construct sediment rating curve using a sediment transport equation (e.g., Brownlie (1981), Bagnold (1980), Wilcock- Kenworthy (2002)) Use RB, beta, and acceptable error to determine whether daily or subdaily flow records are needed in the estimation of Qs50 beta Use appropriate flow data (daily or subdaily) with sediment rating curve to calculate Qs50 Choose index flow Transfer flow record from gaged station to ungaged station Is there an acceptable reference gage? Yes, use index flow method Use hydrologic model to produce streamflow time series from precipitation records Flow Record / FDC No Calculate Richards-Baker flashiness index Will watershed land use change over future time period of interest?No Yes Level of analysis (effort, time, money, accuracy): High Collect sediment transport measurements to determine beta Medium Get beta from estimated sediment transport capacity on erams.com Low Use regression equation (Syvitski et al. 2000) to determine beta Yes No Qs50 St re a m flo w Re c o rd s Se di m e n t T ra n s po rt D at a Step 1 (see Subsection 3.2.1) Step 2 (see Subsection 3.2.2) Step 3 (see Subsection 3.2.3) Step 4 (see Subsection 3.2.4) Step 5 (see Subsection 3.2.5) Step 6 (see Subsection 3.2.6) Step 7 (see Subsection 3.2.7) Figure 2-7. Decision tree supporting Qs50 calculations.

the Design hydrology process 13 future flow regimes that result from land use change. If hydrology is non-stationary, scenario analysis that examines sediment and water continuity of both near bankfull and overbank flood flows under both current and potential future hydrologic conditions [e.g., Soil and Water Assessment Tool—channel DEGradation (SWAT-DEG), Storm Water Management Model] is recommended. Additionally, if the sediment supply reach is currently unstable or there is uncertainty in its representativeness for future conditions, consider computing channel designs using several sediment supply reaches. 2.4.5 Select a Robust Design Use weight of evidence to select a robust design and incorporate other design considerations per Soar and Thorne (2001), NRCS (2007), Shields et al. (2008), Hotchkiss and Frei (2007), and U.S. Forest Service (2008). Acknowledge other objectives, such as flood conveyance, managing debris, aquatic organism habitat and passage, etc., and use sound engineering judgment to make final recommendation. 2.5 Limits to the Application of This Process The hydrology, form, and conditions of streams vary widely. As such the research team attempted to make the guidance as general as possible but, in the end, had to define the limits of applicability for the toolset. Table 2-5 summarizes the types of streams to which the guidance and tools apply and those to which they do not apply. Applies: Does Not Apply: • Alluvial channels • Dune/ripple, pool-riffle, plane bedforms • Single-thread channels • Channel slope ≤ ~ 0.03 • D50 of sand and larger • Near-perennial flow • Non-alluvial channels • Multi-thread, braided, fan channels • Channel slope > ~ 0.03 • Ephemeral, dryland rivers • Abrupt transitions • Channels that lack capacity to transport inflowing sediment load at valley slope • Severely backwatered/tidal situations • CSR Tool does not apply to boulders Table 2-5. Streams and situations to which the guidance and tools apply and situations where they are not directly applicable.