Predictions Of Bedload Transport Research Articles

A comparative ensemble approach to bedload prediction using metaheuristic machine learning

An adequate bedload prediction is a challenging task in hydraulic engineering because of the complex sediment transport processes and corresponding environmental factors. The current study introduces a novel comparative ensemble approach using metaheuristic machine learning (ML) models to enhance the accuracy of bedload prediction using data from laboratory flume experiments. To uncover key insights in bedload transport prediction, several models are employed, such as K-Nearest Neighbours (KNN), Extra Trees Regressor (ETR), Linear Regression (LR), Random Forest (RF), Bagging Regressor (BR), and XGBoost (XGB). The coefficient of determination (R2) and root mean square error (RMSE) values vary between various models; however, XGB showed R2 = 0.99 and RMSE = 0.11. The sensitivity analysis emphasizes the crucial role of the Shields parameter in bedload prediction, while the SHAP analysis highlights the substantial influence of the XGB model in enhancing predictive accuracy. The REC curves show that BR, XGB, and RF, outperformed KNN and LR models. Furthermore, a graphical user interface has also been developed to facilitate user interaction with the predictive models, allowing for easier visualization, analysis, and interpretation of bedload transport predictions. Additionally, k-fold cross-validation was performed to assess the performance, consistency and robustness of the ML models. Thus, it can be concluded from the current study that the utilization of ML algorithms can improve accuracy, providing valuable insights for hydraulic engineers and highlighting the importance of ML models in civil engineering practices, particularly in bedload transport prediction.

Prediction of bedload transport inside vegetation canopies with natural morphology

Prediction of bedload transport inside vegetation canopies with natural morphology

Fluvial bedload transport modelling: advanced ensemble tree-based models or optimized deep learning algorithms?

The potential of advanced tree-based models and optimized deep learning algorithms to predict fluvial bedload transport was explored, identifying the most flexible and accurate algorithm, and the optimum set of readily available and reliable inputs. Using 926 datasets for 20 rivers, the performance of three groups of models was tested: (1) standalone tree-based models Alternating Model Tree (AMT) and Dual Perturb and Combine Tree (DPCT); (2) ensemble tree-based models Iterative Absolute Error Regression (IAER), ensembled with AMT and DPCT; and (3) optimized deep learning models Long Short-Term Memory (LSTM) and Recurrent Neural Network (RNN) ensembled with Grey Wolf Optimizer. Comparison of the predictive performance of the models with that of commonly used empirical equations and sensitivity analysis of the driving variables revealed that: (i) the coarse grain-size percentile D 90 was the most effective variable in bedload transport prediction (where D x is the xth percentile of the bed surface grain size distribution), followed by D 84, D 50, flow discharge, D 16, and channel slope and width; (ii) all tree-based models and optimized deep learning algorithms displayed ‘very good’ or ‘good’ performance, outperforming empirical equations; and (iii) all algorithms performed best when all input parameters were used. Thus, a range of different input variable combinations must be considered in the optimization of these models. Overall, ensemble algorithms provided more accurate predictions of bedload transport than their standalone counterpart. In particular, the ensemble tree-based model IAER-AMT performed most strongly, displaying great potential to produce robust predictions of bedload transport in coarse-grained rivers based on a few readily available flow and channel variables.

Hydrological Drivers of Bedload Transport in an Alpine Watershed

AbstractUnderstanding and predicting bedload transport is an important element of watershed management. Yet, predictions of bedload remain uncertain by up to several order(s) of magnitude. In this contribution, we use a 5‐year continuous time series of streamflow and bedload transport monitoring in a 13.4‐km2 snow‐dominated Alpine watershed in the Western Swiss Alps to investigate hydrological drivers of bedload transport. Following a calibration of the bedload sensors, and a quantification of the hydraulic forcing of streamflow upon bedload, a hydrological analysis is performed to identify daily flow hydrographs influenced by different hydrological drivers: rainfall, snowmelt, and combined rain and snowmelt events. We then quantify their respective contribution to bedload transport. Results emphasize the importance of combined rain and snowmelt events, for both annual bedload volumes (77% on average) and peaks in bedload transport rate. A non‐negligible, but smaller, amount of bedload transport may occur during late summer and autumn storms, once the snowmelt contribution and baseflow have significantly decreased (9% of the annual volume on average). Although rainfall‐driven changes in flow hydrographs are responsible for a large majority of the annual bedload volumes (86% on average), the identified melt‐only events also represent a substantial contribution (14% on average). The results of this study help to improve current predictions of bedload transport through a better understanding of the bedload magnitude‐frequency relationship under different hydrological conditions. We further discuss how bedload transport could evolve under a changing climate through its effects on Alpine watershed hydrology.

Effect of Sediment Supply on Cyclic Fluctuations of the Disequilibrium Ratio and Threshold Transport Discharge, Inferred From Bedload Transport Measurements Over 27 Years at the Swiss Erlenbach Stream

AbstractBedload transport in natural channels typically shows large fluctuations even for constant hydraulic forcing. It has been shown that the threshold shear stress for initiation of motion depends, for example, on the bed morphology and on channel slope and that it may be influenced by flood history effects. This study is based on a unique data set of continuous field observations of bedload transport acquired during 27 years with impact plate measurements and involving more than 500 sediment‐transporting flood events. A disequilibrium ratio is defined which represents the observed bedload mass per event divided by the bedload mass calculated via a bedload transport equation integrated over the time of a flood event. The disequilibrium ratio varies in a cyclic behavior around equilibrium conditions defined by its median value. The mean duration of a cycle depends on upstream sediment supply or availability on the bed, and longer cycles are associated with a larger active layer on the bed. A memory effect is evident regarding the temporal evolution of both disequilibrium ratio and bed state as reflected by threshold transport conditions, determined from the observations as the critical discharge at the start and at the end of a bedload‐transporting event. The disequilibrium ratio and the threshold transport discharge are inversely correlated with each other, likely providing a feedback mechanism governing the fluctuations around an equilibrium state. Accounting for the memory effect, that is, for the bed state after the previous event, improves the prediction of bedload transport for the following event.

Bedload transport rate prediction: Application of novel hybrid data mining techniques

The accurate prediction of bedload transport in gravel-bed rivers remains a significant challenge in river science. However the potential for data mining algorithms to provide models of bedload transport have yet to be explored. This study provides the first quantification of the predictive power of a range of standalone and hybrid data mining models. Using bedload transport data collected in laboratory flume experiments, the performance of four types of recently developed standalone data mining techniques - the M5P, random tree (RT), random forest (RF) and the reduced error pruning tree (REPT) - are assessed, along with four types of hybrid algorithms trained with a Bagging (BA) data mining algorithm (BA-M5P, BA-RF, BA-RT and BA-REPT). The main findings are four-fold. First, the BA-M5P model had the highest prediction power (R2 = 0.943; RMSE = 0.061 kg m−1 s−1; MAE = 0.040 kg m−1 s−1; NSE = 0.945; PBIAS = −1.60) followed by M5P, BA-RT, RT, BA-RF, RF, BA-REPT, and REPT. All models displayed ‘very good’ performance except the BA-REPT and REPT model, which were ‘satisfactory’. Second, the M5P, BA-RT, and RT models underestimated, and the BA-M5P, BA-RF, RF, BA-REPT and REPT models overestimated, bedload transport rates. Third, flow velocity had the most significant impact on bedload transport rate (PCC = 0.760) followed by shear stress (PCC = 0.709), discharge (PCC = 0.668), bed shear velocity (PCC = 0.663), bed slope (PCC = 0.490), flow depth (PCC = 0.303), median sediment diameter (PCC = 0.247), and relative roughness (PCC = 0.003). Fourth, the maximum depth of tree was the most sensitive operator in decision tree-based algorithms, and batch size, number of execution slots and number of decimal places did not have any impact on model’ prediction power. Overall the results revealed that hybrid data mining techniques provide more accurate predictions of bedload transport rate than standalone data mining models. In particular, M5P models, trained with a Bagging data mining algorithm, have great potential to produce robust predictions of bedload transport in gravel-bed rivers.

Numerical simulation of hydrodynamic characteristics and bedload transport in cross sections of two gravel-bed rivers based on one-dimensional lateral distribution method

Accurate evaluation and prediction of bedload transport are crucial in studies of fluvial hydrodynamic characteristics and river morphology. This paper presents a one-dimensional numerical model based on the one-dimensional lateral distribution method (1D-LDM) and six classic bedload transport formulae that can be used to simulate hydrodynamic characteristics and bedload transport discharge in cross sections. Two gravel-bed rivers, i.e. the Danube River located approximately 70 km downstream from Bratislava in Slovakia and the Toltén River in south of Chile are used as examples. In the 1D-LDM, gravity, bed shear stress, turbulent diffusion, and secondary flow are included to allow for accurate predictions of flow velocity and consequently bed shear stress in the cross sections. Six classic formulae were applied to evaluate the non-dimensional bedload transport rate, and the bedload transport discharge through a river cross section is obtained by integrating the bedload transport rate over the width of the cross section. The results show that the root mean square error (RMSE) and mean absolute error (MAE) of velocity and water discharge were less than 8% of the observed magnitude, while the correlation coefficient between model predictions and observations was close to unity. The formulae proposed by Ashida and Michiue (1972), in which particle collision with the bed is taken into account, and by Camenen and Larson (2005), which allows for yielding a non-zero bedload transport rate even when the bed shear stress is smaller than the critical bed shear stress value, appeared to be more appropriate for predicting the observed bedload transport rate in the studied cross sections of two gravel-bed rivers. If non-uniform sediment mixtures were considered, the bedload transport discharge through a cross-section could change considerably by up to 22.5% of the observed magnitude. The relations proposed by Ashida and Michiue (1972) and Egiazaroff (1965) for parameterizing the hiding factor yielded more realistic model predictions in comparison with observations for the measured data set collected for the Toltén River, while the one proposed by Wilcock and Crowe (2003) performs the best for the data set measured for the Danube River.

Bed load prediction in gravel-bed rivers using wavelet kernel extreme learning machine and meta-heuristic methods

The present study aims at developing a wavelet kernel extreme learning machine (WKELM) and meta-heuristic method, known as particle swarm optimization (PSO). PSO algorithm is employed in order to provide a desirable modeling by optimal determination of parameters attributed to WKELM. In order to confirm the ability of employed PSO-WKELM approach in solving the problem, a well-known kernel-based support vector machine (SVM) is applied to compare the obtained results. 890 data points from 19 gravel-bed rivers located in the USA were used to feed the utilized heuristic models. Three different scenarios were proposed; in the scenario 1, different combinations of parameters based on hydraulic characteristics were prepared, scenario 2 was developed using both hydraulic and sediment properties as model inputs of bed load transport, and lastly, the performance of employed PSO-WKELM approach for prediction of bed load transport with different range of median particle size was investigated. The obtained results confirmed the higher predictive potential of PSO-WKELM in comparison with SVM. Also, it was found that prediction of bed load transport with median particles size ranging from 1 to 1.4 mm led to more valid outcome. Performing the sensitivity analysis demonstrated the remarkable impact of the ratio of average velocity (V) to shear velocity (U*) in modeling process.

Uncertainty analysis in bed load transport prediction of gravel bed rivers by ANN and ANFIS

River flow is a complex dynamic system of hydraulic and sediment transport. Bed load transport have a dynamic nature in gravel bed rivers and because of the complexity of the phenomenon include uncertainties in predictions. In the present paper, two methods based on the Artificial neural networks (ANN) and adaptive neuro-fuzzy inference system (ANFIS) are developed by using 360 data points. Totally, 21 different combination of input parameters are used for predicting bed load transport in gravel bed rivers. In order to acquire reliable data subsets of training and testing, subset selection of maximum dissimilarity (SSMD) method, rather than classical trial and error method, is used in finding randomly manipulation of these subsets. Furthermore, uncertainty analysis of ANN and ANFIS models are determined using Monte Carlo simulation. Two uncertainty indices of d factor and 95% prediction uncertainty and uncertainty bounds in comparison with observed values show that these models have relatively large uncertainties in bed load predictions and using of them in practical problems requires considerable effort on training and developing processes. Results indicated that ANFIS and ANN are suitable models for predicting bed load transport; but there are many uncertainties in determination of bed load transport by ANFIS and ANN, especially for high sediment loads. Based on the predictions and confidence intervals, the superiority of ANFIS to those of ANN is proved.

Estimation of bedload discharge in sewer pipes with different boundary conditions using an evolutionary algorithm

Sediment transport is a complex phenomenon due to the nonlinearity and uncertainties of the process. The present study applies Gene Expression Programming (GEP) to develop bedload transport models in sewer pipes. In this regard, two types of bedload were considered: loose bed (deposition state) and rigid bed (limit of deposition state). In order to develop the models, two scenarios with different input combinations were considered: Scenario 1 considers only hydraulic characteristics and Scenario 2 considers both hydraulic and sediment characteristics as inputs for modeling bedload discharge. The results proved the capability of GEP in prediction of sediment transport and it was found that for prediction of bedload transport in sewer pipes Scenario 2 performed more successfully than Scenario 1. According to the outcome of sensitivity analysis, Frm (Modified Froude number) and d50/y (relative particle size) for rigid boundary and Frm for loose boundary had key roles in the modeling. The outcome of the GEP models also was compared with existing empirical equations and it was found the GEP models yielded better results. It was also found that pipe diameter affected the transport capacity of the sewer pipe. Increasing pipe diameter caused an increase in model efficiency. A pipe with a diameter of 305mm yielded to the best results.

Representative sediment sizes in predicting the bed-material load for nonuniform sediments

This paper evaluates the applicability of representative sediment size in terms of median grain size and the effective sediment sizes for the transport of non-uniform natural sediment mixtures. Natural sediment from the Krasak River, Yogjakarta was used. Bed samples were collected from three different river sections, i.e. upstream (U), midstream (M) and downstream (D). The median diameters, d50 and sediment geometric standard deviation, σg for sections U,M and D were 1.63mm, 0.91mm and 0.69mm, and 4.8, 2.3 and 1.5 respectively. Particular attention was paid to the effect of varying representative sediment sizes on the threshold sediment motion, where four threshold equations were employed. The threshold value varies when σg=4.8 and has little variation for sediment mixtures with σg<2.4. Analysis is extended to prediction of bedload transport using eight selected formulae which account for both with and without the threshold criteria of sediment motion. To account for the non-uniformity of the bed materials, a comparison between the calculated bedload discharge using d50 was made with the bedload prediction when a representative sediment diameters were used. Analysis shows that an improvement of the statistical analysis was only observed for predictions calculated using Brownlie (1981) and Camenen and Larson (2005) expressions when effective sediment sizes were employed for non-uniform sediment mixtures. Employing the effective sediment sizes de or dem to calculate the bedload discharge does not necessarily improve the bedload prediction and the use of d50 is seen as sufficient.

An analysis of nonlinearity effects on bed load transport prediction

Because bed load equations are nonlinear and because parameters describing the flow and the bed can have large variance, different results are expected when integrating bed load over a cross section with respect to spatially variable local data (2D), or when computing bed load from cross‐section‐averaged data, which reduces the problem to uniform conditions (1D). Evidence of these effects is shown by comparing 1D (flume‐derived) equations with 2D field measurements, and by comparing a 2D (field‐derived) equation with 1D flume measurements, leading to the conclusion that different equations should be used depending on whether local or averaged data are used. However, whereas nonlinearity effects are considerable for low‐transport stages, they tend to disappear for higher‐flow conditions. Probability distribution functions describing the variance in flow and bed grain size distribution (GSD) are proposed, and the width‐integrated bed load data (implicitly containing the natural variance in bed and flow parameters) are used to calibrate these functions. The method consists of using a Monte Carlo approach to match the measured 2D bed load transport rates with 1D computations, artificially reproducing the natural variance associated with the mean input parameters. The Wilcock and Crowe equation was used for the 1D computation because it was considered representative of 1D transport. The results suggest that nonlinearity effects are mostly sensitive to the variance in shear stress, modeled here with a gamma function, whose shape coefficient α was shown to increase linearly with the transport stage. This variance in shear stress suggests that even for very low flow conditions, shear stress can locally exceed the critical shear stress for the bed armor, generating local armor breakup. This could explain why the bed load GSD is usually very similar to subsurface GSD, even in the presence of complete armor.

Influence of sediment supply on mountain streams bedload transport

Sediment transport and storage in mountain stream channels are closely dependent on hillslope activities and bank erosion. This paper analyses bedload measured in 13 mountain streams (989 values with slope higher than 5%) in an attempt to investigate how much the sediment supply condition impacts bedload transport rates. First, the sediment supply condition of each stream was evaluated qualitatively using information available in the original studies. In the second step, the data set was analysed by comparing bedload transport rates considering a Shields stress ratio and by comparing transport rate efficiency with consideration of a dimensionless stream power. The two analyses indicate higher sediment transport rates for the streams identified as “connected” to an active source. The consequences for bedload transport prediction are discussed with a nonthreshold bedload equation. Even though predicting the exact sediment supply condition and the associated transport rate is not yet possible, this study suggests that defining an envelope delimiting minimum and maximum transport for a given river reach should be possible.

Start and end of bedload transport in gravel-bed streams

[1] The threshold of incipient bedload motion, expressed either as a critical force or as a critical water discharge, is a key parameter in bedload transport prediction. Measuring the threshold of motion is difficult, and reliable data from natural streams are rare. By recording the vibrations triggered by bedload particles when moving over a steel plate mounted in the channel bed, we determined the time at start and end of bedload transport in four streams, where discharge is continuously monitored. The threshold discharge scatters over approximately one order of magnitude for each stream, reinforcing previous observations that critical discharge is characterized by a distribution of values. We interpret a strong correlation between the discharge at the start of transport and the discharge at the end of transport of the previous event to reflect temporal changes in bed structure and consequent effects on the driving and resisting forces acting on the bed.

MS485 PARAMETERS AFFECTING PROGNOSIS OF PATIENTS WITH ACUTE MYOCARDIAL INFARCTION. AN AMI REGISTRY

This paper evaluates the applicability of representative sediment size in terms of median grain size and the effective sediment sizes for the transport of non-uniform natural sediment mixtures. Natural sediment from the Krasak River, Yogjakarta was used. Bed samples were collected from three different river sections, i.e. upstream (U), midstream (M) and downstream (D). The median diameters, d50 and sediment geometric standard deviation, σg for sections U,M and D were 1.63 mm, 0.91 mm and 0.69 mm, and 4.8, 2.3 and 1.5 respectively. Particular attention was paid to the effect of varying representative sediment sizes on the threshold sediment motion, where four threshold equations were employed. The threshold value varies when σg=4.8 and has little variation for sediment mixtures with σg<2.4. Analysis is extended to prediction of bedload transport using eight selected formulae which account for both with and without the threshold criteria of sediment motion. To account for the non-uniformity of the bed materials, a comparison between the calculated bedload discharge using d50 was made with the bedload prediction when a representative sediment diameters were used. Analysis shows that an improvement of the statistical analysis was only observed for predictions calculated using Brownlie (1981) and Camenen and Larson (2005) expressions when effective sediment sizes were employed for non-uniform sediment mixtures. Employing the effective sediment sizes de or dem to calculate the bedload discharge does not necessarily improve the bedload prediction and the use of d50 is seen as sufficient.

A comparison between flume and field bed load transport data and consequences for surface‐based bed load transport prediction

The ability of simple equations to predict bed load transport with limited knowledge of the bed surface material was investigated. This was done using a data set consisting of 7,636 bed load transport values from the flume (1,317 data) and from 84 river reaches (6,319 field data). It was possible to collapse field and flume data by correcting the ratio between the Shields number and its critical value with a very simple hiding function proposed as a power law of the D84/D50 ratio. In so doing, a surface‐based bed load transport formula was proposed. It was successfully tested on an independent data set (comprising sand and gravel bed rivers with slope in the range 0.0002–0.08), with 86% of the values predicted within a precision of 1 order of magnitude. Moreover, the formula reproduced the low transport rates well, contrary to the usual surface‐based formulas also tested, and is particularly well suited for estimating low transport rates associated with near‐bankfull flow discharge. This new formula is neither time consuming (no fractionwise calculation) nor data consuming (the required parameters are the flow discharge, the active width, the slope, and the surface grain diameters D50 and D84).

The formation of headland/island sandbanks

There are four extensive sandbanks in the vicinity of the Isle of Portland, a headland in the English Channel. The formation and maintenance of the two most prominent of these sandbanks (one on either side of the headland) can largely be explained by net bedload convergence, driven by instantaneous headland eddies generated by tidal flow past the headland. However, there are also two less prominent sandbanks (again, one on either side of the headland), which are not located in zones of bedload convergence. It is suggested here that these latter two sandbanks were formed when the Isle of Portland was isolated from the mainland by a tidal strait. Relative sea-level data and radiocarbon dates indicate that this would have occurred ca. 9–7 ka BP, prior to the closure of the strait by sedimentation. Tidal flow through this strait generated eddy systems in addition to the headland eddies, leading to the formation of associated headland/island sandbanks. At 7 ka BP, sedimentation resulted in closure of the strait, leading to the present-day headland configuration, and subsequent reworking of these now moribund sandbanks formed by the strait. A series of idealised morphological model experiments, parameterised using bedrock depths and glacial isostatic adjustment model output of relative sea level, are here used to simulate this hypothesised sequence of sandbank evolution over the Holocene. The results of the model experiments are corroborated by in situ observations of bedforms and sediment characteristics, and by acoustic Doppler current profiler (ADCP) data applied to predictions of bedload transport over the sandbanks. In addition to demonstrating the mechanism which leads to the formation of sandbanks by tidal flow through a strait, the model results show that upon subsequent closure of such a strait, these sandbanks will no longer be actively maintained, in contrast to sandbanks which are continuously maintained by headland eddies.

Effects of Vanes and W-Weir on Sediment Transport in Meandering Channels

Sediment transport patterns in a meandering channel with instream restoration structures (vane and W-weir) have been studied. Laboratory experiments were conducted in a large-scale mobile-bed channel with graded materials under bank-full and overbank flow conditions. Bed-load samples were collected with a calibrated minisampler. Vanes, constructed against the outer bank in a meander bend, relocated the scour hole toward midchannel, thereby protecting the bank from erosion. The sediment sizes ( d50 , d90 ) in the bend became slightly more coarse and more uniform in the center of the channel. The W-weir installed immediately below a riffle section created two scour holes without affecting the upstream bed or the natural sediment transport of the channel. Predictions of bed-load transport by selected deterministic and stochastic methods showed large deviation from measurements using Helley–Smith sampler in sections downstream of the bend apex. In addition to creating local scour holes, the structures also re...

Unified View of Sediment Transport by Currents and Waves. I: Initiation of Motion, Bed Roughness, and Bed-Load Transport

Attention is given to the properties of sediment beds over the full range of conditions (silts to gravel), in particular the effect of fine silt on the bed composition and on initiation of motion (critical conditions) is discussed. High-quality bed-load transport data sets are identified and analyzed, showing that the bed-load transport in the sand range is related to velocity to power 2.5. The bed-load transport is not much affected by particle size. The prediction of bed roughness is addressed and the prediction of bed-load transport in steady river flow is extended to coastal flow applying an intrawave approach. Simplified bed-load transport formulas are presented, which can be used to obtain a quick estimate of bed-load transport in river and coastal flows. It is shown that the sediment transport of fine silts to coarse sand can be described in a unified model framework using fairly simple expressions. The proposed model is fully predictive in the sense that only the basic hydrodynamic parameters (depth, current velocity, wave height, wave period, etc.) and the basic sediment characteristics (d10, d50, d90, water temperature, and salinity) need to be known. The prediction of the effective bed roughness is an integral part of the model.

The storage of bed material in mountain stream channels as assessed using Optically Stimulated Luminescence dating

A detailed understanding of channel forming and maintenance processes in mountain streams requires some measurement and/or prediction of bed load transport and sediment mobility. Traditional field based measurements of such processes are problematic because of the high formative discharges characteristic of such streams. The application of Optically Stimulated Luminescence (OSL) dating is proposed here as a new way of determining actual residency times of fine sediments and consequently validating selected predictions for the entrainment of sediment in these streams. Model predictions of sediment mobility for selected step-pool and plane-bed channels in a mountain catchment in south eastern Australia are initially calculated using equations of hydraulic competence and the one-dimensional HEC-RAS model. Results indicate that floods exceeding bankfull with recurrence intervals up to 13 years are competent to mobilise the maximum overlying surface grain sizes at both sites. OSL minimum age model results from 7 samples of well bleached quartz in the fine matrix particles indicate general agreement with selected competence equations. The apparent long (100–1400 y) burial age of most of the mineral quartz, however, suggests that competent flows are not able to flush all subsurface fine-bed material. The depth of maximum bed load exchange (flushing) was limited to ≤ twice the depth of the overlying D 90 grain size. Application of OSL in this study provides important insight into the nature of storage and flushing of matrix material in mountain streams.