Design of a hybrid ANN multi-objective whale algorithm for suspended sediment load prediction.

Accurate and reliable suspended sediment load (SSL) prediction models are necessary for planning and management of water resource structures. More recently, soft computing techniques have been used in hydrological and environmental modeling. The present paper compared the accuracy of three different soft computing methods, namely, artificial neural networks (ANNs), adaptive neuro-fuzzy inference system (ANFIS), coupled wavelet and neural network (WANN), and conventional sediment rating curve (SRC) approaches for estimating the daily SSL in two gauging stations in the USA. The performances of these models were measured by the coefficient of correlation (R), Nash-Sutcliffe efficiency coefficient (CE), root-mean-square error (RMSE), and mean absolute percentage error (MAPE) to choose the best fit model. Obtained results demonstrated that applied soft computing models were in good agreement with the observed SSL values, while they depicted better results than the conventional SRC method. The comparison of estimation accuracies of various models illustrated that the WANN was the most accurate model in SSL estimation in comparison to other models. For example, in Flathead River station, the determination coefficient was 0.91 for the best WANN model, while it was 0.65, 0.75, and 0.481 for the best ANN, ANFIS, and SRC models, and also in the Santa Clara River, amounts of this statistical criteria was 0.92 for the best WANN model, while it was 0.76, 0.78, and 0.39 for the best ANN, ANFIS, and SRC models, respectively. Also, the values of cumulative suspended sediment load computed by the best WANN model were closer to the observed data than the other models. In general, results indicated that the WANN model could satisfactorily mimic phenomenon, acceptably estimate cumulative SSL, and reasonably predict peak SSL values.

Suspended sediment load (SSL) estimation is a required exercise in water resource management. This article proposes the use of hybrid artificial neural network (ANN) models, for the prediction of SSL, based on previous SSL values. Different input scenarios of daily SSL were used to evaluate the capacity of the ANN-ant lion optimization (ALO), ANN-bat algorithm (BA) and ANN-particle swarm optimization (PSO). The Goorganrood basin in Iran was selected for this study. First, the lagged SSL data were used as the inputs to the models. Next, the rainfall and temperature data were used. Optimization algorithms were used to fine-tune the parameters of the ANN model. Three statistical indexes were used to evaluate the accuracy of the models: the root-mean-square error (RMSE), mean absolute error (MAE) and Nash-Sutcliffe efficiency (NSE). An uncertainty analysis of the predicting models was performed to evaluate the capability of the hybrid ANN models. A comparison of models indicated that the ANN-ALO improved the RMSE accuracy of the ANN-BA and ANN-PSO models by 18% and 26%, respectively. Based on the uncertainty analysis, it can be surmised that the ANN-ALO has an acceptable degree of uncertainty in predicting daily SSL. Generally, the results indicate that the ANN-ALO is applicable for a variety of water resource management operations.

Suspended sediment load prediction modelling based on artificial intelligence methods: The tropical region as a case study

The suspended sediment load (SSL) prediction is one of the most important issues in water engineering. In this article, the adaptive neuro-fuzzy interface system (ANFIS) and support vector machine (SVM) were used to estimate the SLL of two main tributaries of the Telar River placed in the north of Iran. The main Telar River had two main tributaries, namely, the Telar and the Kasilian. A new evolutionary algorithm, namely, the black widow optimization algorithm (BWOA), was used to enhance the precision of the ANFIS and SVM models for predicting daily SSL. The lagged rainfall, temperature, discharge, and SSL were used as the inputs to the models. The present study used a new hybrid Gamma test to determine the best input scenario. In the next step, the best input combination was determined based on the gamma value. In this research, the abilities of the ANFIS-BWOA and SVM-BWOA were benchmarked with the ANFIS-bat algorithm (BA), SVM-BA, SVM-particle swarm optimization (PSO), and ANFIS-PSO. The mean absolute error (MAE) of ANFIS-BWOA was 0.40%, 2.2%, and 2.5% lower than those of ANFIS-BA, ANFIS-PSO, and ANFIS models in the training level for Telar River. It was concluded that the ANFIS-BWOA had the highest value of R2 among other models in the Telar River. The MAE of the ANFIS-BWOA, SVM-BWOA, SVM-PSO, SVM-BA, and SVM models were 899.12 (Ton/day), 934.23 (Ton/day), 987.12 (Ton/day), 976.12, and 989.12 (Ton/day), respectively, in the testing level for the Kasilian River. An uncertainty analysis was used to investigate the effect of uncertainty of the inputs (first scenario) and the model parameters (the second scenario) on the accuracy of models. It was observed that the input uncertainty higher than the parameter uncertainty.

The Mississippi River delta is facing severe land loss. An urgent need exists to examine sediment transport and channel morphology dynamics along this highly engineered alluvial river that has shaped and will continue to shape its delta. This dissertation research focuses on investigating channel morphology dynamics and sediment transport in the recent three decades along a 327-km reach (from about 492 to 167 km upstream from the river’s outlet in the Gulf of Mexico) of the Lowermost Mississippi River (LmMR). The specific objectives of this research were to: 1) analyze riverbed adjustment, i.e., channel-bed aggradation or channel erosion at seven locations along the 327-km LmMR reach over the last three/four decades; 2) determine suspended sand availability under various discharge regimes at Tarbert Landing (the uppermost location of LmMR) during the period between 1973 and 2013; and 3) quantify bedload at Tarbert Landing, St Francisville and Baton Rouge (three uppermost locations of LmMR) and suspended load at St Francisville and Baton Rouge over the last one to four decades. This research found that the first 20–25 km LmMR reach below its diversion to the Atchafalaya River and the reach from ~ 80 to 140 km experienced significant riverbed aggradation, while the reach in between (i.e. from ~ 25 to 80 km) experienced riverbed degradation over the last three/four decades. The lower 187-km reach (i.e. from 140 to 327 km) showed higher sediment outflow and negligible sediment trapping. Furthermore, the LmMR discharged an average annual sand load (SLs) of 27 million tons (MT) during 1973 and 2013, at Tarbert Landing, varying largely from 3.4 to 52.3 MT. Also, during the four decades, the LmMR at Tarbert Landing carried about 71% of the total annual sand load in about 120 days each year, when the discharge was ≥ 18000 cubic meters per second (cms). The bedload transport rates along the LmMR gradually increased from Tarbert Landing [83 million tons (MT) during 2004-2015 for grain size of 0.125 mm] to Baton Rouge (at 367.5 rk) (96 MT during 2004-2014 for the same grain size). However, the total sediment supply (bedload + suspended sediment load) was substantially higher at Tarbert Landing (931 MT) and lower and nearly equal at the other downstream locations (550 MT at St Francisville and 544 MT at Baton Rouge) during 2004-2010 (the matching period of availability for both bedload and suspended load). These findings have relevant implications for the management

Suspended sediment load prediction using non-dominated sorting genetic algorithm II

Suspended sediment load in northwestern South America (Colombia): A new view on variability and fluxes into the Caribbean Sea

Suspended sediment load (SSL) is essential to river and dam engineering. Due to the complexity and stochastic nature of sedimentation, SSL prediction is a challenging task and conventional methods often fail to generate accurate results. Aiming to provide an improved estimation, this paper contributes to a new forecasting framework by integrating the seasonal adjustment (SA) and Bayesian optimization (BOP) into a machine learning (ML) model (denoted as BMS). The SA is used for de-seasonalisation and trend extraction; the BOP is to optimize the ML architecture. The BMS is evaluated using the daily SSL records from the Yangtze River. Its performance is appraised by statistical criteria of Nash-Sutcliffe efficiency (NSE), correlation coefficient (CC), root mean squared error (RMSE) and mean absolute error (MAE). With the de-noising and hyper-parameter tuning modules, the BMS effectively heightens the accuracy of the standard ML models. The most significant improvement occurs in the Boosting model, with its augments in NSE and CC by 4.3% and 1.6%, and reductions in RMSE and MAE by 24.9% and 24.2%. The BMS gains by comparison even under flood conditions, where it remarkably reduces the errors of the constituent models by up to 47.9% for RMSE and 48.3% for MAE.

Concern on alteration of sediment natural flow caused by developments of water resources system, has been addressed in many river basins around the world especially in developing and remote regions where sediment data are poorly gauged or ungauged. Since suspended sediment load (SSL) is predominant, the objectives of this research are to: 1) simulate monthly average SSL (SSLm) of four catchments using artificial neural network (ANN); 2) assess the application of the calibrated ANN (Cal-ANN) models in three ungauged catchment representatives (UCR) before using them to predict SSLm of three actual ungauged catchments (AUC) in the Tonle Sap River Basin; and 3) estimate annual SSL (SSLA) of each AUC for the case of with and without dam-reservoirs. The model performance for total load (SSLT) prediction was also investigated because it is important for dam-reservoir management. For model simulation, ANN yielded very satisfactory results with determination coefficient (R2) ranging from 0.81 to 0.94 in calibration stage and 0.63 to 0.87 in validation stage. The Cal-ANN models also performed well in UCRs with R2 ranging from 0.59 to 0.64. From the result of this study, one can estimate SSLm and SSLT of ungauged catchments with an accuracy of 0.61 in term of R2 and 34.06% in term of absolute percentage bias, respectively. SSLA of the AUCs was found between 159,281 and 723,580 t/year. In combination with Brune’s method, the impact of dam-reservoirs could reduce SSLA between 47% and 68%. This result is key information for sustainable development of such infrastructures.

One of the most undesirable consequences induced by blasting in open-pit mines and civil activities is flyrock. Furthermore, the production of oversize boulders creates many problems for the continuation of the work and usually imposes additional costs on the project. In this way, the breakage of oversize boulders is associated with throwing small fragments particles at high speed, which can lead to serious risks to human resources and infrastructures. Hence, the accurate prediction of flyrock induced by boulder blasting is crucial to avoid possible consequences and its’ environmental side effects. This study attempts to develop an optimized artificial neural network (ANN) by particle swarm optimization (PSO) and jellyfish search algorithm (JSA) to construct the hybrid models for anticipating flyrock distance resulting in boulder blasting in a quarry mine. The PSO and JSA algorithms were used to determine the optimum values of neurons’ weight and biases connected to neurons. In this regard, a database involving 65 monitored boulders blasting for recording flyrock distance was collected that comprises six influential parameters on flyrock distance, i.e., hole depth, burden, hole angle, charge weight, stemming, and powder factor and one target parameter, i.e., flyrock distance. The ten various models of ANN, PSO–ANN, and JSA–ANN were established for estimating flyrock distance, and their results were investigated by applying three evaluation indices of coefficient of determination (R2), root mean square error (RMSE) and value accounted for (VAF). The results of the calculation of evaluation indicators revealed that R2, values of (0.957, 0.972 and 0.995) and (0.945, 0.954 and 0.989) were determined to train and test of proposed predictive models, respectively. The yielded results denoted that although ANN model is capable of anticipating flyrock distance, the hybrid PSO–ANN and JSA–ANN models can anticipate flyrock distance with more accuracy. Furthermore, the performance and accuracy level of the JSA–ANN predictive model can estimate better compared to ANN and PSO–ANN models. Therefore, the JSA–ANN model is identified as the superior predictive model in estimating flyrock distance induced from boulder blasting. In the final, a sensitivity analysis was conducted to determine the most influential parameters in flyrock distance, and the results showed that charge weight, powder factor, and hole angle have a high impact on flyrock changes.

The sediment load transported in a river is the most complex hydrological phenomenon due to a large number of obscure parameters and the existence of both spatial variability of the basin characteristics and temporal climatic patterns. In this paper two artificial neural network (ANN) models were developed for semidistributed modeling of the suspended sediment load process of the Eel River watershed located in California. The first model was an integrated ANN model trained by the data of multiple stations inside the watershed. In the second model, a geomorphology-based ANN model, space-dependent geomorphologic parameters of the subbasins, extracted by geographic information system tools, accompanied by time-dependent meteorological data, were imposed on the network. In both models, three-layer perceptron neural networks were trained considering various combinations of input and hidden layers’ neurons, and the optimum architectures of the models were selected according to the computed evaluation cr...

As a major component of the hydrologic cycle, rainfall runoff plays a key role in water resources management and sustainable development. Conceptual models of the rainfall-runoff process are governed by parameters that can rarely be directly determined for use in distributed models, but should be either inferred through good judgment or calibrated against the historical record. Artificial neural network (ANN) models require comparatively fewer such parameters, but their accuracy needs to be checked. This paper compares a Hydrologic Engineering Centre-Hydrologic Modeling System (HEC-HMS) conceptual model and an ANN model based on the conjugate gradient method for streamflow prediction. Daily precipitation, temperature, and streamflow data of the Upper Indus River for a period of 20 years (1985–2004) are used as input for calibrating in the case of the HEC-HMS, and for training in case of the ANN. Ten years of data (2005–2014) are used to validate the HEC-HMS model and test the ANN. The performance of the models is assessed using different statistical indicators such as the Nash–Sutcliffe efficiency (NSE), root mean square error (RMSE), mean bias error (MBE), and the coefficient of determination (R2). The results show good predictions for streamflow in the case of both HEC-HMS and ANN models. A parametric study is conducted using Monte Carlo analysis and finds that the most important parameters for HEC-HMS models are the storage coefficient and the time of concentration; while for ANN models, input combinations are the most important. This study investigates the sensitivity of these parameters, which can be used to determine preliminary estimation ranges of their values for future modeling. Finally, evaluating the impact of the simulated streamflow’s accuracy on the flow duration curve shows that the curve is significantly affected by any streamflow simulation inaccuracy.

River suspended sediment modelling using the CART model: A comparative study of machine learning techniques

Coarse sediment transport in Himalayan rivers exerts a first-order control on river morphology, flood hazard, and the longevity of hydropower infrastructure. Variations in gravel or sand export from the mountain front govern downstream channel incision and aggradation rates, directly modifying channel conveyance capacity and flood risk on the Gangetic Plains. These highly mobile gravel-bed rivers draining the Himalayas are prone to abrupt channel switching driven by the mobility of coarse bedload, exacerbating flood hazard. Despite their importance, current flood models and hydropower development plans remain limited by uncertainties in bedload flux estimates. To understand the controls on sediment yield from the mountain front, we monitor both suspended sediment load through direct sampling, and bedload through acoustic and seismic monitoring at the mountain front of the Karnali and West Rapti Rivers in Nepal. Two Aquarian hydrophones connected to AudioMoth data loggers are installed at each site to monitor bedload transport for the duration of an entire monsoon season. This acoustic dataset is combined with seismic data from a DiGOS geophone at each site, water level data from the Department of Hydrology and Meteorology in Nepal, and manual suspended sediment concentration sampling. These complementary approaches independently approximate the sand and gravel fractions respectively, allowing us to evaluate the accuracy of current approximation methods for bedload in Himalayan rivers. The results indicate that suspended sediment loads are primarily supply-limited, showing a pronounced seasonal hysteresis in suspended sediment concentration with peaks that record local events such as landslides and storms in the catchment. In contrast, the bedload flux appears to be transport-limited, closely tracking the river discharge. Acoustic data show initial bedload movement at lower discharges, with an abrupt increase in bedload transport signal at a threshold discharge. At higher discharges there is a saturation of the signal, which may reflect boundary conditions of bedload transport or data clipping. These contrasting signals indicate independent transport mechanisms for suspended load and bedload, with a high degree of variability in suspended load relative to a much more predictable bedload flux. Seasonal fluctuations in this bedload-to-suspended-load ratio demonstrate that bedload estimates based on suspended sediment measurements can substantially misrepresent total sediment fluxes, with major implications for flood early warning systems and urban planning.

Anthropogenic climate change and changes to land use and land management practices can have significant impacts on streamflow and sediment transport. In this study, we investigated long-term precipitation, streamflow, and suspended sediment load patterns within the Hangman Creek watershed, draining from the Rocky Mountains in Idaho to Washington, to identify the magnitude of changes with the goal of better understanding the links between these processes and the potential effects of agricultural conservation practices (ACPs) implemented since the 1990s. Comparing the study periods of 1991 to 2020 with 1961 to 1990, 1991 to 2020 had lower streamflow/precipitation ratios in the highest flow months such as February and March. Most streamflow occurred during winter and spring, so did suspended sediment. In addition, 2018 had much lower suspended sediment load compared to earlier years (1999 and 2000) during high flow seasons (January to April) given that streamflow was higher in 2018 than in 1999 and 2000. These changes may be attributed to the adoption of agricultural conservation practices because land cover remained almost unchanged from 2001 to 2021 and ACP adoption increased. Finally, the flow frequency analysis showed a strong linkage between higher streamflow events and increased suspended sediment load, with between 81% and 96% of total annual suspended sediment loads transported during the highest 10% of flows.