Murat Cobaner

Comparison of Artificial Neural Network Methods with L-moments for Estimating Flood Flow at Ungauged Sites: the Case of East Mediterranean River Basin, Turkey

A regional flood frequency analysis based on the index flood method is applied using probability distributions commonly utilized for this purpose. The distribution parameters are calculated by the method of L-moments with the data of the annual flood peaks series recorded at gauging sections of 13 unregulated natural streams in the East Mediterranean River Basin in Turkey. The artificial neural networks (ANNs) models of (1) the multi-layer perceptrons (MLP) neural networks, (2) radial basis function based neural networks (RBNN), and (3) generalized regression neural networks (GRNN) are developed as alternatives to the L-moments method. Multiple-linear and multiple-nonlinear regression models (MLR and MNLR) are also used in the study. The L-moments analysis on these 13 annual flood peaks series indicates that the East Mediterranean River Basin is hydrologically homogeneous as a whole. Among the tried distributions which are the Generalized Logistic, Generalized Extreme Vaules, Generalized Normal, Pearson Type III, Wakeby, and Generalized Pareto, the Generalized Logistic and Generalized Extreme Values distributions pass the Z statistic goodness-of-fit test of the L-moments method for the East Mediterranean River Basin, the former performing yet better than the latter. Hence, as the outcome of the L-moments method applied by the Generalized Logistic distribution, two equations are developed to estimate flood peaks of any return periods for any un-gauged site in the study region. The ANNs, MLR and MNLR models are trained and tested using the data of these 13 gauged sites. The results show that the predicting performance of the MLP model is superior to the others. The application of the MLP model is performed by a special Matlab code, which yields logarithm of the flood peak, Ln(QT), versus a desired return period, T.

Comparison of an ANN approach with 1-D and 2-D methods for estimating discharge capacity of straight compound channels

Most natural streams or rivers exhibit a compound or two-stage geometry consisting of a main channel and one or two floodplains. The discharge capacity of compound channels has an importance in flood defence schemes and in the economic development of floodplain areas for agriculture and parks. Therefore, the comprehensive stage-discharge model studies performed and different one or two-dimensional methods have been developed. In this study, the single-channel method (SCM), the divided-channel method (DCM), the coherence method (COHM), the exchange discharge method (EDM) and the Shiono-Knight method (SKM) have been compared with a multilayer perception neural network (MLP) with Levenberg-Marquardt algorithm. The results of the comparisons reveal that the artificial neural network (ANN) model gives slightly better statistical results than those of the COHM, EDM and these three give more accurate results than those of the SCM, DCM, and SKM.

Estimation of Monthly Mean Reference Evapotranspiration in Turkey

Monthly mean reference evapotranspiration (ET0) is estimated using adaptive network based fuzzy inference system (ANFIS) and artificial neural network (ANN) models. Various combinations of long-term average monthly climatic data of wind speed, air temperature, relative humidity, and solar radiation, recorded at stations in Turkey, are used as inputs to the ANFIS and ANN models so as to calculate ET0 given by the FAO-56 PM (Penman-Monteith) equation. First, a comparison is made among the estimates provided by the ANFIS and ANN models and those by the empirical methods of Hargreaves and Ritchie. Next, the empirical models are calibrated using the ET0 values given by FAO-56 PM, and the estimates by the ANFIS and ANN techniques are compared with those of the calibrated models. Mean square error, mean absolute error, and determination coefficient statistics are used as comparison criteria for evaluation of performances of all the models considered. Based on these evaluations, it is found that the ANFIS and ANN schemes can be employed successfully in modeling the monthly mean ET0, because both approaches yield better estimates than the classical methods, and yet ANFIS being slightly more successful than ANN.

Reference evapotranspiration based on Class A pan evaporation via wavelet regression technique

Accurate estimation of reference evapotranspiration (ET0) is important for water resources engineering. Therefore, a large number of empirical or semi-empirical equations have been developed for assessing ET0 from numerous meteorological data. However, records of such weather variables are often incomplete or not always available for many locations, which is a shortcoming of these complex models. Therefore, practical and simpler methods are required for estimating the ET0. In this study, the efficiency of a wavelet regression (WR) model in estimating reference evapotranspiration based on only Class A pan evaporation is examined. The results of the WR model are compared with those of three pan-based equations, namely the FAO-24 pan, Snyder ET0 and Ghare ET0 equations and their calibrated versions. Daily Class A pan evaporation data from the Fresno and Bakersfield stations of the United States Environmental Protection Agency in California, USA, are used in the study. The WR model estimates are compared against those of the FAO-56 Penman–Monteith equation. Results showed that the WR model is capable of accurately predicting the ET0 values as a product of pan evaporation data.

Bridge afflux analysis through arched bridge constrictions using artificial intelligence methods

Although many studies have been carried out for estimating the afflux through modern straight deck bridge constrictions, little attention has been given to medieval arched bridge constrictions. Hydraulic Research Wallingford in the UK (Brown, P.M., 1988. Afflux at arch bridges. Report SR 182. Wallingford, UK: HR Wallingford) recently published a major coverage of both experimental and field afflux data obtained from arched bridge constrictions. The report pointed out that the present day formulas developed for estimating the bridge afflux are inadequate to apply to ancient arched structures. Therefore, this study aimed at developing new afflux methods for arched bridge constrictions using multi-layer perceptrons (MLP) neural networks, radial basis function-based neural networks (RBNN), generalised regression neural networks (GRNN) and adaptive neuro-fuzzy inference system (ANFIS) model. Multiple linear and multiple nonlinear regression analyses were also used for comparison purposes. Mean square errors, mean absolute errors, mean absolute relative errors, average of individual ratios between predicted and actual values, and determination coefficients were used as comparison criteria for the evaluation of model performances. The test results showed that MLP, RBNN, GRNN, and ANFIS models gave reasonable accuracy when applied to both the field and experimental data collected by Hydraulic Research Wallingford.

Artificial neural network approaches for prediction of backwater through arched bridge constrictions

This paper presents the findings of laboratory model testing of arched bridge constrictions in a rectangular open channel flume whose bed slope was fixed at zero. Four different types of arched bridge models, namely single opening semi-circular arch (SOSC), multiple opening semi-circular arch (MOSC), single opening elliptic arch (SOE), and multiple opening elliptic arch (MOE), were used in the testing program. The normal crossing (@f=0), and five different skew angles (@f=10^o, 20^o, 30^o, 40^o, and 50^o) were tested for each type of arched bridge model. The main aim of this study is to develop a suitable model for estimating backwater through arched bridge constrictions with normal and skewed crossings. Therefore, different artificial neural network approaches, namely multi-layer perceptron (MLP), radial basis neural network (RBNN), generalized regression neural network (GRNN), and multi-linear and multi-nonlinear regression models, MLR and MNLR, respectively were used. Results of these experimental studies were compared with those obtained by the MLP, RBNN, GRNN, MLR, and MNLR approaches. The MLP produced more accurate predictions than those of the others.

ANN approaches for the prediction of bridge backwater using both field and experimental data

This paper presents the findings of laboratory model testing of arched bridge constrictions in a rectangular open-channel flume whose bed slope was fixed at zero. Four different types of arched bridge models, namely single-opening semi-circular arch, multiple-opening semi-circular arch, single-opening elliptic arch, and multiple-opening elliptic arch, were used in the testing program. The normal crossing (φ  =  0) and five different skew angles (φ  =  10°, 20°, 30°, 40°, and 50°) were tested for each type of arched bridge model. Recently, a major coverage of backwater field data obtained from the medieval arched bridge constrictions was published by the Hydraulic Research Wallingford in the UK (Brown, P.M., 1985. Hydraulics of bridge waterways: Interium report. Wallingford, UK: Hydraulic Research Wallingford, Report SR 60; Brown, P.M., 1987. Afflux at arch bridges: second interium report. Wallingford, UK: Hydraulic Research Wallingford, Report SR 115; Brown, P.M., 1988. Afflux at arch bridges. Wallingford, UK: Hydraulic Research Wallingford, Report SR 182). These data were also used in the analysis. The main aim of this study is to develop a suitable model for estimating backwater through arched bridge constrictions with normal and skewed crossings using both experimental and field data. Therefore, different artificial intelligence approaches, namely multi-layer perceptron (MLP), radial basis neural network (RBNN), generalized regression neural network (GRNN), and multi-linear and multi-nonlinear regression models, MLR and MNLR, respectively were used. The comparison between these developed models and one of the most commonly used traditional methods (Biery, P.F. and Delleur, J.W., 1962. Hydraulics of single span arch bridge constrictions. ASCE Journal of the Hydraulics Division, 88, 75–108) has been made. The test results showed that the MLP model gave highly accurate results than those of Biery and Delleur, MLR, MNLR, and GRNN and gave similar results with the RBNN model when applied to both field and experimental data.

Feasibility of Hydropower Plant Installation to Existing Irrigation Dams

Abstract Because the cost of energy has risen considerably in recent decades, the addition of a suitable capacity hydropower plant (HPP) to the end of the pressure conduit of an existing irrigation dam may be economically feasible. First, a computer program capable of realistically calculating total losses from the inlet to outlet throughout any pressure conduit for many discharges (Qs) from the minimum to the maximum at small increments is coded, the outcome of which enables the determination of the C coefficient of the Total Head Loss = C·Q2. Next, a computer program is used to determine the hydroelectric energy produced at monthly periods, the present worth (PW) of their monetary gains. The average annual energy produced by a HPP is then coded. Inflows series, irrigation water requirements, evaporation rates, turbine running time ratios, and the C coefficient are the input data of this program. Running the program with a synthetically generated Mnumber of in-year-long infows series, histograms of both the average annual energies and the PWs of energy incomes are determined to which suitable probability distributions are then fitted. This model has been applied to ten randomly chosen irrigation dams in Turkey, and a regression equation is obtained to estimate the average annual energy as a function of gross head available and the annual volume of irrigation water, which should be useful for reconnaissance studies.