ABSTRACT Vegetative filter strips (VFS) intercept sediment, but most models assess performance under a single storm, ignoring legacy deposits from successive events that may alter trapping efficiency. A 10-m grass-lined flume underwent paired runoff sequences (low-to-high and high-to-low flow/concentration) with sediment left undisturbed between runs; real-time sensors recorded sediment flux, Darcy–Weisbach resistance factor f, unit stream power P, and shear stress τ. Grain-size distributions of effluent and deposits were measured, regression and non-parametric tests evaluated the data. Cumulative retention across storms proved independent of event sequence; however, within any given storm, a larger antecedent sediment load caused a pronounced decline in instantaneous trapping efficiency. Low-flow, low-concentration events produced a finer, narrower effluent size spectrum with peak grain size Dp appearing earlier than under high-flow conditions; antecedent deposits delayed Dp emergence. Dp was a stronger linear predictor of outlet flux than D50. Power-law relations between f or P and sediment flux held only on the bed without previously deposited sediment; legacy sediment weakened or eliminated them, whereas τ never showed a reliable fit. Explicit inclusion of legacy sediment in VFS models is essential to prevent overestimating long-term retention and to guide design for repeated storms.
Luo et al. (2026) studied this question.