Sediment load in fluvial systems is one of the critical factors shaping the river geomorphological and hydraulic characteristics. A detailed understanding of the total sediment load (TSL) is required for the protection of physical, environmental, and ecological functions of rivers. This study develops a robust methodological approach based on multiple linear regression (MLR) and support vector regression (SVR) models modified by principal component analysis (PCA) to predict the TSL in rivers. A database of sediment measurement from large-scale physical modelling tests with 4759 datapoints were used to develop the predictive model. A dimensional analysis was performed based on the literature, and ten dimensionless parameters were identified as the key drivers of the TSL in rivers. These drivers were converted to uncorrelated principal components to feed the MLR and SVR models (PCA-based MLR and PCA-based SVR models) developed within this study. A stepwise PCA-based MLR and a 10-fold PCA-based SVR model with different kernel-type functions were tuned to derive an accurate TSL predictive model. Our findings suggest that the PCA-based SVR model with the kernel-type radial basis function has the best predictive performance in terms of statistical error measures including the root-mean-square error normalized with the standard deviation (RMSE/StD) and the Nash–Sutcliffe coefficient of efficiency (NSE), for the estimation of the TSL in rivers. The PCA-based MLR and PCA-based SVR models, with an overall RMSE/StD of 0.45 and 0.35, respectively, outperform the existing well-established empirical formulae for TSL estimation. The analysis of the results confirms the robustness of the proposed PCA-based SVR model for prediction of the cases with high concentration of sediments (NSE = 0.68), where the existing sediment estimation models usually have poor performance.
Because pollutant dispersion in rivers is strongly influenced by the longitudinal dispersion coefficient ( ), its accurate estimation is critical in the field of environmentally sound hydraulic engineering. In this study, a granular computing (GC) model was explored for the first time to overcome problems in accurately estimating . Because GC is a black-box model that is not user friendly, an appropriate nonlinear regression (NLR) method was also applied to precisely predict . The inclusion of the generally ignored parameter of river curvature in estimation significantly improved NLR model performance. In so doing, both GC and NLR model estimations of achieved high linear coefficients of determination ( ) and small error indices [root mean square error (RMSE) and mean absolute error (MAE)] with respect to measured values. The same analysis showed that the GC model (with , RMSE, and MAE values equal to 0.997, 8.11, and 2.18, respectively), outperformed the NLR model, particularly for extreme high values of . Similarly to previous studies, it was also found that the most effective parameters on were the channel aspect ratio, friction term, and river curvature, respectively, in descending order of importance. Moreover, a comparison between some well-known models and the developed GC and NLR alternative presented here showed the latter to have outperformed the former, indicating that the GC and NLR models are a good choice for prediction.
1 October 2020 in Journal of Contaminant Hydrology
Successful application of one-dimensional advection–dispersion models in rivers depends on the accuracy of the longitudinal dispersion coefficient (LDC). In this regards, this study aims to introduce an appropriate approach to estimate LDC in natural rivers that is based on a hybrid method of granular computing (GRC) and an artificial neural network (ANN) model (GRC-ANN). Also, adaptive neuro-fuzzy inference system (ANFIS) and ANN models were developed to investigate the accuracy of three credible artificial intelligence (AI) models and the performance of these models in different LDC values. By comparing with empirical models developed in other studies, the results revealed the superior performance of GRC-ANN for LDC estimation. The sensitivity analysis of the three intelligent models developed in this study was done to determine the sensitivity of each model to its input parameters, especially the most important ones. The sensitivity analysis results showed that the W/H parameter (W: channel width; H: flow depth) has the most significant impact on the output of all three models in this research.
6 June 2018 in Journal of Water Supply: Research and Technology—AQUA
The fate of pollutants in rivers is mainly affected by the longitudinal dispersion coefficient (Kx). Thus, improved Kx estimation could greatly enhance the water quality management of rivers. In this regard, evolutionary polynomial regression (EPR) was used to accurately predict Kx in rivers as a function of flow depth, channel width, and average and shear velocities. The predicted Kx by EPR modelling was compared with results obtained by more conventional Kx estimation formulas. Initial data analyses using general linear models of variance revealed that all input variables were statistically significant for Kx estimation. The calibrated EPR model showed good performance with coefficient of determination and root mean square error of 0.82 and 79 m2/s, respectively. This is better that other more conventional estimation methods. Application of sensitivity analysis for the EPR model indicated that channel width, average velocity, shear velocity, and flow depth were the main variables in descending order that affected Kx variability. The introduced EPR estimation model for Kx can be incorporated in one-dimensional water quality models for improved simulation of solute concentration in natural rivers.