Bank Retreat Research Articles

Bank Retreat Mechanisms Driven by Debris Flow Surges: A Parameterized Model Based on the Results of Physical Experiments

AbstractLateral erosion is a critical factor that influences the formation and amplification of debris flows. However, our understanding of the bank retreat process in debris flow channels is limited, which limits the evaluation of debris flow magnitudes and the prediction of their activity trends. Herein, we conduct physical experiments to investigate bank retreat mechanisms using five types of bank soil and multiple debris flow surges. The bank retreat process is categorized into two stages: toe cutting and bank collapse. Toe cutting is mainly caused by hydraulic erosion, bank collapse includes gravity erosion in the form of toppling failure. Notably, the bank retreat process exhibits a significant negative feedback loop. Bank erosion widens the channel bed, subsequently decreasing the flow depth. In turn, this reduction in flow depth mitigates bank erosion. Moreover, we discover a concise pattern in the complex coupling of hydraulic erosion and toppling failure: erosion efficiency is linearly and negatively correlated with the bed widening width. We develop a new parameterized model for describing the bank retreat process and provided empirical values for the model parameters. Furthermore, we observe that the initial erosion efficiency first increases and then decreases with an increase in the fine particle content of the bank soil. Additionally, we report a negative correlation between the maximum bed widening width and the fine particle content in the bank soil that follows a power function relationship.

Bank erosion and erosion processes from dendrogeomorphology in southern U.S. prairie streams

AbstractStreambank erosion processes influence the amount of soil material contributed to rivers and sedimentation rates in receiving reservoirs. However, the amount of data on bank erosion rates is limited both in range and extent affecting planning for mitigation and watershed management. Dendrogeomorphology is used to determine the date of wood anatomy changes in annual growth increments of roots exposed by erosion of stream banks that when coupled with measurement of the distance of roots to the channel side can be used to calculate the bank retreat rate. Erosion rates derived from dendrogeomorphology are important because these provide erosion data over longer time scales (decades). Here, we use this method to quantify erosion for three different sized watersheds (4 to 3781 km2) located in the water‐scarce southern U.S. prairies that are heavily reliant on surface water and reservoir storage. From 49 roots from the two smaller drainages, erosion ranged from 1.5 to 25.4. For 19 roots collected from the larger subbasin erosion rates were larger ranging from 7.4 to 325.0 cm/years with the larger values and variance associated with two high‐flow events that occurred a year before sampling. We also found differences in straight and meandering portions of the streams where the distance to bank was strongly and positively correlated with the years since root exposure in straight sections. In contrast, meandering bends also showed a positive but low correlation for root exposure date and distance collected. We attribute this difference to erosional processes (i.e., scour and mass wasting) occurring at these channel locations. When compared with other erosion studies across the southern U.S. prairie, our values were similar in magnitude but with low correlation to drainage area indicating site specificity of erosion mechanisms, and watershed landcover influence for different drainages, despite being in the same ecological region.

Bank erosion under the impacts of hydraulic erosion, river stage change and revetment protection in the Middle Yangtze River

Bank failures in alluvial rivers are a typical soil-water interaction problem, which is related to many factors including the direct action of flow, river stage change, and human actions (such as bank revetment). To investigate the failure mechanism of protected riverbanks and possible factors affecting their stability, we analyzed data measured from a typical reach of the Middle Yangtze River. Furthermore, we performed numerical simulations of seepage and stress variation inside the riverbank. The field observation and simulated results indicated that: (1) Hydraulic erosion by near-bank flow remains the primary factor influencing the erosion of the protected riverbank. However, the bank protection works effectively limit the lateral bank retreat but increase the incision of the nearby riverbed, with the largest erosion depth of 10.6 m during August to November in 2020. (2) The initial damage in protected banks may be triggered by local tensile stress concentration during the water-rising period, under the combined actions of hydrostatic confining force, pore water pressure and gravity. This initial damage will progress into more severe bank failure events, particularly during the flood period. (3) After the regulation of the Three Gorges Project, the increased changing rate of river stage (∼1.6–2.5 fold) could potentially increase the risk of damage to protected riverbanks in the Middle Yangtze River.

The effects of Sesarma reticulatum (L.) herbivory and sea level rise on creek expansion in Cape Cod salt marshes

High densities of herbivorous purple marsh crabs (Sesarma reticulatum) have caused major vegetation loss in salt marshes across Cape Cod (Massachusetts, USA). As creekbanks are a preferred habitat for this species, much of the damage is concentrated along these edges, resulting in an elevated potential for creek widening through erosion and bank slumping. The presence and/or spatial distribution of S. reticulatum is variable across this region with impacts ranging from minor to severe, and this provides an opportunity to distinguish between background vs. S. reticulatum-enhanced widening. To accomplish this, GIS tools were used to delineate portions of tidal creek systems in 18 marshes, half of which have experienced substantial creekbank vegetation losses from crab-driven herbivory. Delineations were based on georeferenced aerial photography between 2005 and 2021. In addition, creek widths at discrete locations in the delineated systems were measured in each photo year using GIS tools to estimate rates of bank retreat. A smaller number of narrow <2.0 m mosquito ditches were analyzed in the same way to further stratify the data by initial width. Between 2010 and 2021, all tidal conduits ranging from the larger, natural creeks to mosquito ditches widened at a rate that was significantly higher in marshes with no Sesarma damage. While all marshes are under threat from SLR, creek-widening in those experiencing intense crab herbivory may have a greatly shortened lifespan.

Study on the influencing factors and evolution of loess bank collapse with physical modelling

BackgroundReservoir bank collapse in loess areas may lead to the siltation of reservoir and bank retreat. Therefore, the study of reservoir bank collapse has practical significance. Almost of the bank collapse width prediction method were based on the classical graphical method which do not consider the process of bank collapse. But practice shows that this method can overestimate the width of bank collapse. Meanwhile, there are few studies specifically focused on the collapse of loess bank slopes.MethodsTo improve the prediction method of loss bank collapse width, the influence of water depth, dry density and bank slope angle on loess bank collapse was studied by physical modelling. The bank collapse width and the morphology of the bank slope were recorded during the experiment.ResultsThe bank collapse width increases with the increase of water depth, increases with the increase of slope angle, and decreases with the increase of dry density. The modeling process shows that the loess bank collapse occurs firstly underwater, the erosion niche will be formed underwater, and then the above water slope is damaged. This process is repeated until the underwater accumulation slope reaches the stable state, and then bank collapse stops. After the bank collapse, the above water slope is polyline, while the underwater slope is curved. When the slope angle is less than 27°, the bank collapse will not occur, and when the slope angle is between 27° and 40°, the bank collapse type is abrasion type. When the slope angle is greater than 40°, the bank collapse type is dumping type or shear type. Based on this, the improved balanced alluvial accumulation approach was proposed, which considers the mechanical equilibrium of above water bank slope and the morphology of underwater slope. The new method can reflect the stage characteristics of loess bank collapse, which is more reasonable than the empirical graphical method.ConclusionsThe experimental results indicate that when predicting the width of loess bank collapse, it is necessary to combine the bank collapse width and process of bank collapse. The relevant conclusions have a certain role in exploring the mechanism of loess bank collapse and bank collapse prediction methods.

Evolution characteristics of mid‐channel bars with a typical high/low bank structure in the Middle Yangtze River

AbstractAdjustments in mid‐channel bars (MCBs) often influence the stability of channel regime and the safety of river navigation in anabranching reaches. This study aims to investigate the evolution characteristics of the low and high parts of MCBs (L‐MCBs and H‐MCBs) in the Middle Yangtze River (MYR) and reveal the main factors influencing the deformation of these two components. Results show that the total exposed area of L‐MCBs decreased from 305 to 288 km2 in 2003–2019, whereas the H‐MCBs remained relatively stable. Taking the Nanyang bar with a typical high/low bank structure as an example, the deformation of the L‐MCB mainly occurred in the head zone, while the deformation of the H‐MCB was characterised by bank retreat in the tail zone and floodplain deposition on the top of the H‐MCB. In terms of influencing factors, the impacts of the altered flow–sediment conditions and the MCB's properties (soil composition and morphology characteristics) on the Nanyang bar deformation were investigated. On the one hand, the reduction in sediment concentration was the most direct factor causing the shrinkage of L‐MCBs, and the decrease in flood occurrence frequency made the deformation of H‐MCB less significant. On the other hand, the soil composition of the L‐MCB primarily consisted of fine sand with weak erosion resistance, while the large content of cohesive sediment made the H‐MCB difficult to be eroded by overbank flows; moreover, the riverbank of the H‐MCB in the bar‐tail zone was much steeper than that in the bar‐head zone, causing the deformation of the high part to mainly occur in the tail zone. Furthermore, the deformation of the high part was simulated using a bank erosion model at section scale, and the calculated safety factors of riverbank at a typical cross‐section show that the riverbank of the H‐MCB was the most unstable during the receding period.

Water level fluctuations drive bank instability in a hypertidal estuary

Abstract. Hypertidal estuaries are very dynamic environments characterized by high tidal ranges (&gt; 6 m) that can experience rapid rates of bank retreat. Whilst a large body of work on the processes, rates, patterns, and factors driving bank erosion has been undertaken in fluvial environments, the process mechanics affecting the stability of the banks with respect to mass failure in hypertidal settings are not well-documented. In this study, the processes and trends leading to bank failure and consequent retreat in hypertidal estuaries are treated within the context of the Severn Estuary (UK) by employing a combination of numerical models and field-based observations. Our results highlight that the periodic fluctuations in water level associated with the hypertidal environment drive regular fluctuations in the hydrostatic pressure exerted on the incipient failure surfaces that range from a confinement pressure of 0 kPa (at low tide) to ∼ 100 kPa (at high tide). However, the relatively low transmissivity of the fine-grained banks (that are typical of estuarine environments) results in low seepage inflow/outflow velocities (∼ 3 × 10−10 m s−1), such that variations in positive pore water pressures within the saturated bank are smaller, ranging between about 10 kPa (at low tide) and ∼ 43 kPa (at high tides). This imbalance in the resisting (hydrostatic confinement) versus driving (positive pore water pressures) forces thereby drives a frequent oscillation of bank stability between stable (at high tide) and unstable states (at low tide). This transition between stability and instability is found not only on a semidiurnal basis but also within a longer time frame. In the spring-to-neap transitional period, banks experience the coexistence of high degrees of saturation due to the high spring tides and decreasing confinement pressures favoured by the still moderately high channel water levels. This transitional period creates conditions when failures are more likely to occur.

Quantifying the Impact of Model Selection When Examining Bank Retreat and Sediment Transport in Stream Restoration

The objective of this study was to assess the performance of form-based and process-based models, and of local-scale and reach-scale models, used to examine bank retreat and sediment transport in stream restoration. The evaluated models were the Bank Erosion Hazard Index (BEHI), Bank Assessment for Nonpoint Source Consequences of Sediment (BANCS), Bank Stability and Toe Erosion Model (BSTEM), and HEC River Analysis System (HEC-RAS 1D). Model-to-model assessments were conducted to quantify the impact of model selection when predicting applied stress and geomorphic change in a restored stream in North Carolina, USA. Results indicated that the mobility of the bed dictated model selection at the reach-scale. The process-based HEC-RAS 1D was needed to accurately analyze the sand-bed stream, predicting amounts of geomorphic change comparable to measured data and up to three orders of magnitude higher than those from local-scale models. At the local-scale, results indicated that the bank retreat mechanism and flow variability constrained model selection. The form-based BEHI and BANCS did not directly account for geotechnical failure nor capture severe floods, underpredicting amounts of geomorphic change by an order of magnitude when compared to the process-based BSTEM, and failing to characterize erosion potential and applied stresses after short-term morphodynamic adjustments.

Removal of Riprap within Channelized Rivers: A Solution for the Restoration of Lateral Channel Dynamics and Bedload Replenishment?

Riverbank erosion is an essential morphodynamic process for the improvement of river health and the ecohydrogeomorphological functioning of alluvial rivers. Lateral channel dynamics and sediment supply caused by bank erosion largely create and maintain heterogeneous in-channel habitats for fauna and aquatic or riparian plant species. However, humans very early started to stabilize riverbanks in order to favour navigation or to prevent valuable land and infrastructures close to the channel from eroding. During the 20th century, bank protection works such as riprap considerably increased and blocked lateral channel erosion, causing a loss of local sediment supply, which in turn resulted in a decrease in local bedload transport and channel incision. The aim of the article is to evaluate to what extent riprap removal may be an efficient restoration measure in terms of the reactivation of bank erosion and the replenishment of the local bedload in gravel-bed floodplain rivers with a sufficient amount of freedom space. An experimental in situ restoration approach was chosen. First, riprap was removed at two geomorphologically contrasting sites on the Allier River, France. Second, bank retreat was monitored, and the volumes eroded were quantified using photogrammetric and LiDAR surveys. Third, in the case of post-restoration bank erosion, grain size and morphological channel evolution analyses were carried out. Our results suggested that the removal of riprap is an effective measure for certain but not all channelized floodplain reaches. The geomorphological and sedimentary contexts are two criteria that should be considered when selecting sites for restoration. Thus, this study helps river managers to better target the criteria to be taken into account for the selection of sites with high potential for the restoration of lateral channel dynamics.

Longitudinal variation of channel evolution along the middle Yangtze River after the operation of the Three Gorges Project

三峡工程运行后,坝下游河道发生持续冲刷。本文研究了长江中游(955 km)不同河段沿程演变差异及其原因。总体而言,河床形态调整幅度自上而下减弱,这是因为在河床持续冲刷过程中,水流含沙量沿程恢复,故越往下游冲刷相对缓慢。平面形态方面,长江中游岸线崩退及洲滩变形的强度均呈沿程减弱趋势,且在荆江河段最为显著。断面形态方面,河床冲深幅度在宜枝下段＞荆江河段＞宜枝上段＞城汉河段＞汉湖河段。理论上距离三峡工程最近的河段冲刷应最为剧烈,但由于宜枝上段床沙粗化显著,限制了冲刷的进一步发展。过流能力方面,宜枝河段由于距洞庭湖较远,并未受到入汇顶托作用,故其平滩流量的调整基本由进口水沙条件控制,并随着河床冲深下切而增大;对于荆江、城汉和汉湖河段,河床冲刷虽显著,但支流或湖泊的入汇顶托对平滩流量产生的影响大于前者,故平滩流量总体随上下游水位差同步波动。;Significant degradation have occurred in the middle Yangtze River since the operation of the Three Gorges Project. In the current study, longitudinal evolution differences of different reaches along the middle Yangtze River were quantitatively compared, and the main reasons were investigated. In terms of planform geometry, the intensity of bank retreat and central-bar deformation showed a decreasing trend along the reach, and it was most significant in the Jingjiang Reach. With respect to cross-sectional geometry, channel degradation developed gradually from the Yizhi Reach to downstream reaches. It is uncommon that the adjustments of bankfull channel geometry in the upper Yizhi Reach were smaller than those in the downstream reaches. The reason is that the significant coarsening of bed-material composition in the upper Yizhi Reach hindered the development of channel degradation. In terms of discharge capacity, the variations of reach-scale bankfull discharge in the Yizhi Reach were mainly controlled by the incoming flow and sediment conditions, and it increased with the process of channel degradation. In the other subreaches, the variations of bankfull discharge were mainly influenced by the backwater effects owing to the confluences of tributaries and lakes.

Modelling of cantilever bank failure for peat-type meander bends in the source region of the Yellow River

Bank retreat involving a combination of fluvial erosion and bank collapse has been found to be a major contributor to sediment transport, lateral migration, and planform evolution of meandering rivers. Previous studies have largely examined the general mechanism of cantilever bank failure. However, the composite process of beam (toppling) failure caused by shear failure of the lower part composed of non-cohesive soil remains poorly understood. The current paper investigates the diversity and coupling of this collapse process, which is the dominant pattern in the peat-type meandering rivers of the Zoige River basin in the source region of the Yellow River. For this purpose, morphologic and hydraulic measurements were done in the summers of 2016 and 2017. Moreover, a combined method of a theoretical model and numerical simulation was developed to be applied for a continuous meandering channel in the Zoige basin, focusing in particular on five typical cross sections of a highly convoluted bend approaching neck cutoff. Specifically, the theoretical model evaluating the stability of overhanging riverbank was based on static equilibrium theory considering the interaction between composite soil layers. In addition, the Bank Stability and Toe Erosion Model was used to simulate the retreat process of the silt layer based on field measurements including bank morphology, soil composition, and hydrological parameters. Thus, a calculation method for the occurrence of cantilever beam failure in the flood period is proposed, and temporal change and spatial variability in bank profiles caused by three consecutive failures is estimated at the five sections. Satisfactory results showed a relatively close agreement between the calculations and field measurements in terms of the width of the overhanging arms and the dimensions of the slump blocks.

A three-dimensional experimental study on bank retreat: The coupled role of seepage and surface flow

The coupled role of seepage and surface flow on bank retreat has long been neglected, partly due to the concealment and complexity of seepage erosion. To fill this gap, we set up a three-dimensional laboratory experiment to explore bank retreat process in response to seepage and surface flow. During each experiment, we measured the changes of total soil stress, matric suction, and water content within the bank, as well as flow velocity and suspended sediment concentration near the bank. Results show that a rapid decrease in matric suction, the bank toe undercutting consequent to seepage erosion, the formation of tension crack on the bank top, and the occurrence of toppling or shear failure is the typical sequence of the observed bank retreat process under seepage flow. The inclusion of surface flow erodes slump blocks and so promotes cantilever formation, leading to additional bank failure. Compared with the case where only seepage is considered, the frequency of toppling failure under the coupled effect of seepage and surface flow decreases, but the contribution to the bank retreat increases by 37 %. The time taken to collapse in three-dimensional experiments is at least 1.5 h earlier than that of two-dimensional experiments, indicating the importance of preferential flow pathways of seepage. Overall, this research illustrates how surface flow interacts with seepage flow to control bank retreat process and is indeed a first step toward a fully understanding of multifactor-driven bank retreat.

Modelling of Stratified River Bank Erosion Due to Undercutting

Theories of soil mechanics and basic hydraulic relations and principles of loose boundary channel hydraulics were combined and integrated to analyze stratified river bank failure due to undercutting (undermining) of a bottom loose layer. A near bank velocity distribution model was developed from which the boundary shear stress acting on the bank surface was determined. The model allows computation of the eroded soil volume from the bottom loose layer and the lateral undercut distance at any time in the flow hydrograph as long as the induced flow shear stress is greater than the particle critical shear stress resistance to hydraulic entrainment. At failure the size of the failure block and the ultimate critical value of the lateral undercut distance are determined. Therefore the annual rate of bank retreat and bank sediment load contribution can be identified for the reach under consideration. The application of the model to the River Nile hydrograph in Northern Sudan State showed a good agreement with field observations.

Analysis of historical changes in planform geometry of a mountain river to inform design of erodible river corridor

The analysis of changes in planform geometry of the Biała River during the last 130 years is used to diagnose hydromorphological river degradation and thus to verify the need of river restoration through establishing an erodible river corridor. Outcomes from the analysis are then used for delimiting the erodible corridor and predicting the potential for future river widening in the corridor. The study analyses presentations of the active river zone and the structure of river geomorphic units on topographic maps from 1878, 1935 and 1962, aerial photos from 1967, 1977 and 1987, and orthophotos from 1998 and 2009. As a result of river channelization followed by channel incision, during the study period the width of the Biała was reduced to 16%–57% of its width in 1878. The river narrowing was associated with a reduction in the proportion of channel bars in the total river width and nearly complete elimination of islands. All these changes demonstrated the degradation of hydromorphological integrity of the river and the need for its restoration by allowing free channel development in an erodible corridor. The belt of the river migration during the last 130 years identified on the basis of overlays of the river position from all dates was ⁓5 times wider than the contemporary river. This justified delimiting the erodible river corridor on a substantially larger area of the valley floor than the area of the river in 2009. In 2009 the river was narrower than the largest river width recorded between 1878 and 1998 along nearly the whole length of the study river sections. This indicates that the concentration of flood flows in the narrow channel, increasing their unit stream power, is a factor increasing the potential for future bank retreat and channel widening in the erodible corridor of the river.

Russian Sanctions and the Banking Sector

AbstractWestern allies have imposed restrictive measures on Russian banks and threatened harsher sanctions on the finance sector since the annexation of Crimea in 2014. However, this did not impede sanctioned banks reporting record profits, nor deter Russia from invading Ukraine on 24 February 2022. Although the sanctions’ full effectiveness remains in question, the impact of Russia's invasion of Ukraine has not only caused a retreat of foreign banks from mainland Russia, but also ended the international ambitions of the largest Russian banks in Europe and beyond.

Bank Retreat Controls River and Estuary Morphodynamics

Understanding and predicting the geomorphological response of fluvial and tidal channels to bank retreat underpins the robust management of water courses and the protection of wetlands.

A Review on Bank Retreat: Mechanisms, Observations, and Modeling

AbstractBank retreat plays a fundamental role in fluvial and estuarine dynamics. It affects the cross‐sectional evolution of channels, provides a source of sediment, and modulates the diversity of habitats. Understanding and predicting the geomorphological response of fluvial/tidal channels to external driving forces underpins the robust management of water courses and the protection of wetlands. Here, we review bank retreat with respect to mechanisms, observations, and modeling, covering both rivers and (previously neglected) tidal channels. Our review encompasses both experimental and in situ observations of failure mechanisms and bank retreat rates, modeling approaches and numerical methods to simulate bank erosion. We identify that external forces, despite their distinct characteristics, may have similar effects on bank stability in both river and tidal channels, leading to the same failure mode. We review existing data and empirical functions for bank retreat rate across a range of spatial and temporal scales, and highlight the necessity to account for both hydraulic and geotechnical controls. Based on time scale considerations, we propose a new hierarchy of modeling styles that accounts for bank retreat, leading to clear recommendations for enhancing existing modeling approaches. Finally, we discuss systematically the feedbacks between bank retreat and morphodynamics, and suggest that to move this agenda forward will require a better understanding of multifactor‐driven bank retreat across a range of temporal scales, with particular attention to the differences (and similarities) between riverine and estuarine environments, and the role of feedbacks exerted by the collapsed bank soil.

Experimental investigation of permeability and length of a series of spur dikes effects on the control of bank erosion in the meandering channel

Use of spur dikes by modifying the flow pattern is among the common methods for control of erosion at the river banks and their protection. In this research by construction of a laboratory meandering channel and use of five spur dikes, the effects of permeability and length of a series of spur dikes on the erosion control was investigated. The experiments were performed using three effective lengths, four permeability values and three Froude numbers in the non-submerged state. The results showed that the highest amount of bank retreat was achieved in impermeable spur dikes with a length of 9 cm and the lowest amount of bank retreat in permeable spur dike was 54% and a length of 15 cm. In this research use has been made of two criteria R 2 and RMSE for comparing the data obtained from the predicted relationship and the observed data and showed a good compatibility.

Increased artificiality trend driven by an inter-basin water transfer on the Zayandeh-rud River floodplain in Iran

The objective is to evaluate the effects of inter-basin water transfers on land use changes in the floodplain buffer area and the geomorphological evolution of a fluvial system on the receiving basin. Remote sensing techniques and machine learning algorithms have been combined with geomorphological indices for this purpose in the Zayandeh-rud Basin located in central Iran, to assess the trajectory of change over the last 50 years (from 1969 to 2020). This basin extremely experienced water crises due to artificialization events which made government transferring water from surplus basin. The outcomes of the present research express a reduction process in the area of the active channel throughout the study period, with an average narrowing rate of 0.92 m per year, i.e., a total reduction of 49.3%. Based on geomorphological indexes, Bank Retreat and River Network Channel Index, the highest sedimentation and narrowing trend was found for the period 1984-1997, while in the most recent period, from 2009 to 2020, a reverted tendency is observed which indicates higher erosion rates. Additionally, the area of natural cover, water bodies, and absence of farming classes have reduced surfaces in the floodplain to the benefit of urbanisation and agriculture. The stationarity conditions of the floodplain cover predominate followed by degradation or artificialisation processes. Given the current perspective of increasing the availability of water in the basin, an integrated planning that aims to consider environmental, socio-political, and economical elements of the system would be advantageous and needed to break the environmental degradation of this fluvial system.

The effects of discharge and bank orientation on the annual riverbank erosion along Powder River in Montana, USA

Annual bank erosion was measured at multiple cross sections along the free-flowing meandering Powder River in the western United States from 1979 through 2019. Bank erosion was separated into two components—above water and underwater erosion. Above water erosion was measured as the annual bank retreat rate (0–15.4 m y−1). Underwater erosion rate (0–47 m3 m−1 y−1) was calculated as the volume eroded below the water level corresponding to the dominant annual peak discharge, Qp. This paper focuses primarily on the underwater erosion. A total of 491 annual erosion rates were calculated for 23 bank sites along a 90-km study reach in southeastern Montana. Sites were not just hotspots for bank erosion but represent the spectra of variables such as the radius of curvature divided by channel width, R/w (2–86), the peak discharge, Qp (22.7–314 m3 s−1), and the bank orientation (0–360°).Local annual bank erosion was extremely variable in time and space. It was episodic and unsynchronized along the study reach with the maximum annual bank erosion occurring in different years at different bank sites. The composite probability distribution of all 491 annual bank erosion rates was best modeled by a zero-adjusted Weibull distribution. Individual probability distributions for each of the 23 sites were all different from each other and from the composite distribution highlighting the extreme variability. The correlation of the annual underwater erosion with channel geometry and bank variables was low (R2 < 0.31) but the correlation was higher for peak discharge with 25% of the sites having R2 > 0.50.Time-averaging reduced the variability at each site and when grouped into five peak-discharge classes each class was correlated with R/w as a power law with an exponent of about −1. Reach-averaging also reduced the variability for each year, and when grouped by bank orientation (north-, east-, south-, and west-facing), bank erosion was linearly related to Qp with south- and west-facing orientations having about twice as much erosion per unit discharge (0.030 m3 m−1 y−1/m3 s−1) than north- and east-facing orientations.Bank erosion was found to be not just a multi-variate complex process with little correlation and high variability that suggests randomness, but also a process that was a function of a different combinations of variables at different sites at the same time. However, this high variability was reduced by time- and reach-averaging, which produced predictable results analogous to the central limit theorem.