A roughness-corrected index of relative bed stability for regional stream surveys

Quantitative regional assessments of streambed sedimentation and its likely causes are hampered because field investigations typically lack the requisite sample size, measurements, or precision for sound geomorphic and statistical interpretation. We adapted an index of relative bed stability (RBS) for data calculated from a national stream survey field protocol to enable general evaluation of bed stability and anthropogenic sedimentation in synoptic ecological surveys. RBS is the ratio of bed surface geometric mean particle diameter (D-gm) divided by estimated critical diameter (D-cbf) at bankfull flow, based on a modified Shield's criterion for incipient motion. Application of RBS to adequately depict bed stability in complex natural streams, however, has been limited because typical calculations of RBS do not explicitly account for reductions in bed shear stress that result from channel form roughness. We modified the index (RBS*) to incorporate the reduction in bed shear stress available for sediment transport that results from the hydraulic resistance of large wood and longitudinal irregularities in channel dimensions (form roughness). Based on dimensional analysis, we derived an adjustment to bankfull shear stress by multiplying the bankfull hydraulic radius (R-bf) by the one-third power of the ratio of particle-derived resistance to total hydraulic resistance (C-p/C-t)(1/3), where both resistances are empirically based calculations. We computed C, using a Keulegan equation relating resistance to relative submergence of bed particles. We then derived an empirical equation to predict reach-scale hydraulic resistance C-t from thalweg mean depth, thalweg mean residual depth, and large wood volume based on field dye transit studies, in which total hydraulic resistance C, was measured over a wide range of natural stream channel complexity, including manipulation of large wood volumes. We tested our estimates of Ct and RBS* by applying them to data from a summer low flow probability sample of 104 wadeable stream reaches in the Coastal Ecoregion of Oregon and Washington, USA. Stream discharges calculated using these Ct estimates compared favorably with velocity-area measurements of discharge during summer low flow, and with the range of 1 to 2-year recurrence floods (scaled by drainage area) at U.S.Geological Survey gauged sites in the same region. Log [RBS*] ranged from -4.2 to +0.98 in the survey region. Dgm ranged from silt to boulders, while estimated bankfull critical diameter, D*(cbf), ranged from very fine gravel to large boulders. The median value of D*cbf (adjusted for form roughness influences) averaged 40% (inter quartile range 28 to 59%) of the unadjusted estimate D-cbf. Log[RBS*] was consistently negatively related to human disturbances likely to produce excess sediment inputs or hydrologic alteration. Log [RBS*] ranged from -1.9 to +0.5 in the streams within the lower quartile of human disturbance in their basin and riparian areas and was substantially lower (-4.2 to -1.1) in streams within the upper quartile of human disturbance. The synoptic survey methods and designs we used appear adequate to evaluate regional patterns in bed stability and sedimentation and their general relationship to human disturbances. Although the RBS concept also shows promise for evaluating sediment and bed stability in individual streams, our approach is relatively coarse, so site-specific assessments using these rapid field methods might prudently be confned to identifying severe cases of sedimentation or channel alteration. Greater confidence to discern, subtle differences in site-specific assessments could be gained by calculating RBS* using more precise field measurements of channel slope, bed particle size and bankfull dimensions, and by refining our adjustments for energy loss from channel form roughness. Published by Elsevier B.V.