Reza Barati

Parameter Estimation of Nonlinear Muskingum Models Using Nelder-Mead Simplex Algorithm

The linear form of the Muskingum model has been widely applied to river flood routing. However, a nonlinear relationship between weighted-flow and storage volume exists in most rivers, making the use of the linear Muskingum model inappropriate. On the other hand, the application of the nonlinear Muskingum model suffers from hydrologic parameters estimation. The current study aims at presenting the objective approach of the Nelder-Mead simplex (NMS) algorithm for the purpose of estimating the parameters of the nonlinear Muskingum model. The performance of this algorithm is compared with other reported parameter estimation techniques together with a historical example. Results of the implementation of this procedure indicate that the NMS algorithm is efficient for the estimating parameters of the nonlinear Muskingum models. This algorithm is easy to be programmed, and it is quite efficient for finding an optimal solution very quickly. Although this technique requires an initial guess for the parameter estim...

Discussion of “Parameter Estimation of the Nonlinear Muskingum Model Using Parameter-Setting-Free Harmony Search” by Zong Woo Geem

First of all, the discusser would like to thank the author for using parameter-setting-free harmony search (PSF-HS) to estimate the parameters of the nonlinear Muskingum model and would like to draw attention to some points and suggest a practical way for parameter estimation of the nonlinear Muskingum model. In the original paper, the available methods for the parameter estimation of the nonlinear Muskingum model are impliedly classified as two groups. This division consists of (1) mathematical techniques, and (2) phenomenon-mimicking algorithms. However, it is notable that the procedures such as the neuro-fuzzy approach (Chu 2009) that cannot be classified in these two groups can be utilized for the purpose of estimating the parameters of the nonlinear Muskingum model. The discusser is unanimous with the author that the methods of the former group have weaknesses of the complex derivative requirement and/or initial vector assumption of design parameters, and the trouble of the methods of the latter group is that these methods need the determination of the algorithm parameters. However, it is notable that the latter group is randomly searched for the optimal solution, and some uncertainties that may cause different solutions on different runs existed when using the phenomenon-mimicking algorithms to estimate the parameters of the nonlinear Muskingum model. In this discussion, an alternative way was proposed for the parameter estimation to avoid the sensitivity analyses of either algorithm parameters or initial values assumption of the hydrologic parameters. In this procedure during two stages the methods of the mathematical techniques and phenomenon-mimicking algorithms are used. In this way, at first stage the phenomenonmimicking algorithm is run, and after a few seconds in the middle of the operation, the model can be stopped. In the second stage, the values obtained for the hydrologic parameters from the phenomenon-mimicking algorithm are considered as the initial guess of the hydrologic parameters for the mathematical technique. In this condition, the sensitivity analyses of either algorithm parameters or initial values assumption are not necessary. Therefore, finding the optimal solution can be done in the shortest operating time. If the near-optimal result is achieved in the first run of the mathematical technique, it is possible that the optimal result is reached with one extra run of the mathematical technique at the second stage. If infeasible, divergent, or faraway-optimal results are achieved, the procedure must be repeated from the first stage, but these results do not happen with a high probability. For this procedure, two combinations of methods are offered. These combinations consist of (1) genetic algorithm (GA; Mohan 1997) and Nelder-Mead simplex algorithm (NMS; Barati 2011) as GANMS, and (2) evolutionary (EV; Premium Solver Platform 2010) and Broyden–Fletcher–Goldfarb–Shanno technique (BFGS; Geem 2006; Premium Solver Platform 2010) as EV-BFGS. The GA-NMS and EV-BFGS procedures can be modeled by the Optimization Tool add-in of MATLAB software (Yang et al. 2005) and by the Excel Solver function add-in of Microsoft Excel software (Premium Solver Platform 2010), respectively. Therefore, these two procedures are widely available for engineers, and hydrologists

Prediction of longitudinal dispersion coefficient in natural rivers using a cluster-based Bayesian network

Abstract The longitudinal dispersion coefficient is a key element in determining the distribution and transmission of pollution, especially when cross-sectional mixing is completed. However, the existing predictive techniques for this purpose exhibit great amounts of uncertainty. The main objective of this study is to present a more accurate model for predicting longitudinal dispersion coefficient in natural rivers and streams. Bayesian network (BN) approach was considered in the modeling procedure. Two forms of input variables including dimensional and dimensionless parameters were examined to find the best model structure. In order to increase the performance of the model, the clustering method as a preprocessing data technique was applied to categorize the data in separate groups with similar characteristics. An expansive data set consisting of 149 field measurements was used for training and testing steps of the developed models. Three performance evaluation criteria were adopted for comparison of the results of the different models. Comparison of the present results with the artificial neural network (ANN) model and also well-known existing equations showed the efficiency of the present model. The performance of dimensionless BN model 30% is more than dimensional ones in terms of the root mean square error. The accuracy criterion was increased from 70 to 83% by performing clustering analysis on the BN model. The BN-cluster model 43% is more accurate than ANN model in terms of the accuracy criterion. The results indicate that the BN-cluster model give 16% better results than the best available considered model in terms of the accuracy criterion. The developed model provides a suitable approach for predicting pollutant transport in natural rivers.

Performance comparison of four turbulence models for modeling of secondary flow cells in simple trapezoidal channels

Abstract Due to the importance of channel flow characteristics in the water conveyance, the study of it is a noteworthy problem for hydraulics experts and much attempts has been accomplished for the modeling of the flow characteristics. One significant problem in this respect is the secondary flow cells and their effect on flow specifications. Widespread experimental and analytical investigations have been accomplished on this phenomenon. However, researchers are trying to replace the expensive and time-consuming experimental approaches and ad hoc analytical models with numerical simulation procedures using computational fluid dynamics (CFD). Selection of the proper turbulence model is one of the most important problems for this type of the numerical modeling. In the present study, after evaluating several turbulence models including k–ε, shear stress transport (SST) and three versions of the Reynolds stress model (RSM) (i.e. LRR-IP, LRR-QI, and SSG models), for the numerical simulation of the secondary flow cells and their effects on trapezoidal channels flow, the more efficient model was selected. Available experimental data and theoretical model was used to validate the selected turbulence model. The results were validated in terms of the free water surface, depth-averaged velocity, and boundary shear stress. The results confirmed the performance and efficiency of SSG version of the Reynolds stress model for the numerical modeling of the secondary flow cells.

Discussion of “Study of the spatial distribution of groundwater quality using soft computing and geostatistical models” by Saman Maroufpoor, Ahmad Fakheri-Fard and Jalal Shiri (2017)

This paper is about the discussion of “Study of the spatial distribution of groundwater quality using soft computing and geostatistical models” by Saman Maroufpoor, Ahmad FakheriFard, and Jalal Shiri (ISH Journal of Hydraulic Engineering, 07th Dec 2017, doi: 10.1080/09715010.2017.1408036). The original study was focused on the application of data-driven approaches for the estimation of the spatial distribution of the groundwater electrical conductivity (EC) for a case study in Iran. The results were compared with the outputs of the geostatistical methods using several performance evaluation criteria. As the conclusion of the original study, the authors mentioned that the performance of the artificial neural network (ANN) model is better than the other applied models (i.e. ANFIS model and simple Kriging and Co-Kriging methods) in estimating the groundwater EC. The work by the authors is really appreciated. The discusser, however, would like to add a few points. During recent years, applications of the soft computing approach have been widely considered in the field of the hydraulic and water resources engineering (Harasami et al. 2017; Roushangar et al. 2017; Talebi et al. 2017; Najafzadeh et al. 2017; Haghiabi et al. 2017; Londhe and Narkhede 2017; Adamala et al. 2017; Londhe and Shah 2017; Alizamir et al. 2018; Barzegar et al. 2018; Niu et al. 2018). However, the modeling of the natural phenomenon with such approach should be done by the careful consideration of the nature of problems in real-field applications (Barati 2011; Barati 2013; Barati et al. 2014a, 2014b; Barati 2017; Shahheydari et al. 2015; Hosseini et al. 2016; Alizadeh et al. 2017). In the original study, two scenarios were used for the estimation of the spatial distribution of the groundwater EC. The scenarios are: (1) the triple input model by considering the longitude, latitude, and number of months as modeling input, and (2) the quadruple input model by using three aforementioned input as well as chlorine (Cl) ion. For the first point, the use of the groundwater level (GWL) as an input of the models can increase the performance of the developed models for the estimation of the groundwater EC. The groundwater flow direction can be shown by the groundwater level pattern. As a practical point of the groundwater movement, it is well known that the groundwater quality decreases during groundwater movement from the aquifer entrance to its outlet. Therefore, it is anticipated that the value of the groundwater EC increases in the direction of the groundwater movement. By considering that the groundwater level decreases from the aquifer entrance to its outlet in an unconfined aquifer, the groundwater level is a reliable input for the estimation of the groundwater EC. In order to show the performance of above-mentioned discussion in a real case study, Davarzan plain with an unconfined aquifer has been considered in Razavi Khorasan province, Iran (a case study with almost similar hydrological and hydrogeological conditions of the study area of the original paper). The sampling points for the hydrochemical analysis of the groundwater are corresponded to 15 operation deep wells. The groundwater level in the locations of the operation wells are estimated using 18 observation wells data. The inverse distance weighted (IDW) geostatistical method is used for the spatial interpolation of the groundwater level. The locations of the operation and observation wells of Davarzan plain were depicted in Figure 1. The groundwater directions were also showed in this figure. The relationships between groundwater EC (μmho/cm) and latitude, longitude, and GWL (m) were, respectively, illustrated in Figures 2–4. The best linear fit along with the correlation coefficient value were also presented in these figures. As it is clear, the relationship between the groundwater EC and GWL is more significant than others (i.e. the groundwater EC with the latitude or longitude). It should be noted that the relationships between the groundwater EC and the latitude and longitude is not linear, but it is considered linear for the comparison purpose. From Figure 4, it can be seen that the groundwater EC decreases with increasing GWL. Moreover, from Figure 1, the groundwater directions are from the northeast and east to west. In other words, the entrance of the aquifer is the northeast and east and the aquifer outlet is the west. Therefore, the value of the groundwater EC is larger for the points in the west. The groundwater EC increases by decreasing the longitude for this case study, because the west points have lower values of the longitude. However, the relationship between EC and the latitude is more complicated. These results can be seen in Figures 2 and 3. These evidences uphold that the value of the groundwater EC increases in the direction of the groundwater movement.

Photocatalytic removal of cadmium (II) and lead (II) from simulated wastewater at continuous and batch system

Aims: The aim of this study was to evaluate the photocatalytic processes for cadmium (Cd 2+ ) and lead (Pb 2+ ) removal at continuous and batch system. Materials and Methods: This study was performed at laboratory scale. The reactors used in this study consisted of three parts: Ultraviolet (UV) source, reaction cell, and mixing chamber. The experiments were carried out in a batch and continuous reactor for synthetic wastewater. The concentration of Cd 2+ and Pb 2+ was constant (25 mg/L) in all experiments and effect of titanium dioxide (TiO 2 ) dose, pH, and air dispersion was investigated on the removal efficiency. Results: The results showed that with increasing TiO 2 dose and pH, the cadmium and lead removal increase. The maximum removal of cadmium and lead was obtained in TiO 2 dose 0.9 g/L and pH: 11 that were equal to 99.8 and 99.2% respectively. Furthermore, when air dispersion increased, the removal efficiency increased; while in the air dispersion 2 cm 3 /L the removal efficiency was maximum (88 and 93.2% at the contact time 56 min for Cd 2+ and Pb 2+ , respectively). Conclusion: According to these results the TiO 2 has been considered as photocatalyst is the separable and recyclable, so UV/TiO 2 process is an environment friendly process for toxic metal removal.

Discussion of ‘Modeling water table depth using adaptive Neuro-Fuzzy Inference System by Umesh Kumar Das, Parthajit Roy and Dillip Kumar Ghose (2017)

ABSTRACT This paper is about the discussion of ‘Modeling water table depth using adaptive Neuro-Fuzzy Inference System by Umesh Kumar Das, Parthajit Roy and Dillip Kumar Ghose (ISH Journal of Hydraulic Engineering, 29 December 2017, DOI: 10.1080/09715010.2017.1420497). The authors of the original paper presented the application of Back Propagation Neural Network (BPNN) and Adaptive Neuro-Fuzzy Inference System (ANFIS) models to predict the water table depth of an aquifer by considering the precipitation, temperature, humidity, surface runoff and evapotranspiration loss as the input of the models. Five different input combinations were examined in the modeling. The authors concluded that the performance of ANFIS is better as compared to BPNN by considering several evaluation criteria. The discusser appreciates this type of study, and the results could be extremely useful for engineers or other audiences. Although the methodology and findings observed in the study are reasonable, this discussion calls attention to add a few issues that may require further clarification.