Bedload Sediment Transport Regimes of Semi‐alluvial Rivers Conditioned by Urbanization and Stormwater Management

Abstract

Watershed urbanization and stormwater management (SWM) alter the hydrologic processes of rivers. Although differences have been documented in channel morphology and sediment yield pre-and posturbanization, little is known about how the modified hydrology affects grain-scale bedload transport dynamics. This study aims to characterize the bedload sediment transport regime of three rivers with different hydrologic settings: rural, urban with no SWM, and urban with peak-shaving SWM. The rivers are "semi-alluvial," characterized by an alluvial layer over a cohesive till. Bedload transport was monitored using tracer stones over 3 years. Hydrograph characteristics of the streams fit with what is expected in urban and SWM systems, and the rural stream has an episodic transport regime typical of gravel-bed rivers. Entrainment thresholds are not detectably impacted by the semi-alluvial bed cover, but travel lengths of grains relative to their size are longer than in alluvial gravel-bed streams. Downstream displacement rates of particles up to the D 90 are accelerated in the urban river due to more frequent mobilization rather than increased event-based travel lengths and may explain channel enlargement. SWM decreases the mobility and travel lengths of particles below those in the rural system, which is combined with channel narrowing, and the loss of bed forms suggests a shift toward a competence-limited transport regime. This new regime is a result of reduced shear stresses that are insufficient to transport coarse material. This study presents empirical evidence of the effects of watershed urbanization and SWM on bedload transport and provides recommendations for process-based river management strategies.Plain Language Summary Most studies on the effects of urbanization on river systems focus on linking the extent of urbanization to channel form or total sediment yield. Few studies apply process-level sediment transport theory to urban environments. This study compares how coarse sediment (bedload) is transported through three streams: a rural river, an urban river with no stormwater management, and an urban river with stormwater management. The glacial geology of the area introduces a unique sediment composition (semi-alluvial) that increases the transport distance of mobilized bedload. The speed of bedload sediment displacement was increased in the urban stream because of more frequent floods, which can explain the enlarged channel dimensions observed. Although stormwater management was successful at reducing the displacement speed of coarse sediment, it resulted in a narrow channel, coarser bed material, and a loss of natural bed variation important for aquatic habitats. Results from this study can be used to improve urban river management strategies that aim to reduce erosion and promote ecologic health.

Process‐based assessment of success and failure in a constructed riffle‐pool river restoration project

“…Values of τ Ã ci were chosen by assuming that the peak shear stress during events with lowL i approximates the critical shear stress (τ ci ) and back-calculating τ Ã ci from Equation (2). A value of τ Ã cD50 ≈ 0.047 (Julien, 2010) was found to work well for all study reaches (see also Papangelakis et al, 2019), which fits with what is expected in gravel-bed rivers. The onset of tracer mobilization for the tracers corresponding to the D 75 and D 90 size classes fit well with the τ Ã ci predicted by the Egiazaroff (1965) hiding function (Papangelakis et al, 2019;Raso, 2017):…”

“…The duration of mobilizing events is also decreased in the restored reach for events exceeding the τ c(5 − 6 ψ ) , but not events exceeding the τ c(6 − 7 ψ ) or τ c(7 − 8 ψ) ( Table 1). Event durations remain much smaller compared to the rural stream that experiences less flashy hydrographs (Papangelakis et al, 2019). Figure 6 shows the difference inL i between the two reaches:…”

“…The second wave also included a population of tracers seeded in Ganatsekaigon Creek as part of another study (Papangelakis et al, 2019). The second wave of tracers was designed to improve upon several limitations observed by MacVicar et al (2015) using the F i g u r e 3 (a) Grain-size distribution of study reaches and tracers where ψ = −ϕ = −log 2 D (D = b-axis diameter in mm), and (b) boxplots of tracer b-axis diameters within each tracer size class for each seeded reach and tracer wave.…”

“…When defined in the latter way, it can be used to quantify the mean annual velocity of bedload sediment through the system (Houbrechts et al, 2015). A recent study comparing the mean annual velocity of riverine tracer particles showed that bed particles move farther downstream per year in rivers with urbanized watersheds (Papangelakis et al, 2019). The increase in the virtual velocity of bed particles and the increase in mobility of the coarsest fractions was linked to the channel enlargement observed (Bevan et al 2018;Papangelakis et al, 2019).…”

“…A recent study comparing the mean annual velocity of riverine tracer particles showed that bed particles move farther downstream per year in rivers with urbanized watersheds (Papangelakis et al, 2019). The increase in the virtual velocity of bed particles and the increase in mobility of the coarsest fractions was linked to the channel enlargement observed (Bevan et al 2018;Papangelakis et al, 2019). A reasonable goal of river restoration could, therefore, be to decrease the annual velocity of bed sediment particles to pre-urbanization values.…”