Abstract—The use of artificial neural network (ANN) modeling for prediction and forecasting variables in water resources engineering are being increasing rapidly. Infrastructural applications of ANN in terms of selection of inputs, architecture of networks, training algorithms, and selection of training parameters in different types of neural networks used in water resources engineering have been reported. ANN modeling conducted for water resources engineering variables (river sediment and discharge) published in high impact journals since 2002 to 2011 have been examined and presented in this review. ANN is a vigorous technique to develop immense relationship between the input and output variables, and able to extract complex behavior between the water resources variables such as river sediment and discharge. It can produce robust prediction results for many of the water resources engineering problems by appropriate learning from a set of examples. It is important to have a good understanding of the input and output variables from a statistical analysis of the data before network modeling, which can facilitate to design an efficient network. An appropriate training based ANN model is able to adopt the physical understanding between the variables and may generate more effective results than conventional prediction techniques. Keywords—ANN, discharge, modeling, prediction, sediment, I. INTRODUCTION ATER resources engineering comprises the study of hydraulics, hydrology, environment and some geological related projects. Engineers frequently faced the difficulties while prediction and estimation of water resources parameters (i.e. sediment discharge, water discharge, rainfall, runoff, water quality etc.). The majority of these variables reveal a highly nonlinear behavior because of spatial and temporal variations. Nonlinear and complex exhibition of these variables is because of spatial and temporal variations which are always difficult to estimate accurately owing to these variations and causes uncertainty in the prediction results. However, water resources engineers attempted to respond these problems arising in design and management of different water resources engineering projects. M. R. Mustafa is with Department of Civil Engineering, Universiti Teknologi Petronas, Tronoh 31750, Malaysia (phone: 006-019-595-7132; e- mail: raza_geo@hotmail.com). M. H. Isa is with Department of Civil Engineering, Universiti Teknologi Petronas, Tronoh 31750, Malaysia (e-mail: hasnain_isa@petronas.com). R. B. Rezaur is with Golder Associates Ltd. 102, 2535-3rd Avenue S.E., Calgary T2A 7W5, Alberta, Canada, (e-mail: Rezaur_Bhuiyan@Golder.com). Their coherent answer to these crisis has somehow produced an effective solution for planning and design of water resources. The one of the most attractive feature is the ANN modeling which has the ability to learn the exact behavior between the inputs and outputs from the examples without any kind of the physical involvement. Artificial neural networks have a wonderful characteristic that it can extract the exact pattern between the input and output variables without any additional explanation. ANNs has been known as to recognize the fundamental behavior between the variables although the data is noisy and containing some errors. All these qualities recommend the applicability of ANNs for the water resources parameters problems regarding prediction and estimation. In this context, a number of applications of ANNs for prediction, forecasting, modeling and estimation of water resources variables (i.e. water discharge, sediment discharge, rainfall runoff, ground water flow, precipitation and water quality etc.) have been found and related to river discharge and sediment are cited here. However, only the ANN applications for river sediment and discharge published in high impact journals since 2002 to 2011 are examined in this review. Therefore, the goal of this study is to examine how effectively ANN has been applied to solve problems in water resources engineering particularly in river sediment and discharge. Furthermore, what kind of infrastructure (input selection criterion, selection and division of the data sets, appropriate structure of the network, activation function and algorithms used for training network etc.) has been utilized for proper modeling to find the best solution of the problems. II. ANN MODELING FOR SEDIMENT ESTIMATION River sediment discharge determination is one of the crucial problems in water resources engineering. Several techniques including ANN have been successfully applied for estimation and prediction of suspended sediments around the world [1- 33]. However, this study is limited to ANN techniques only. A number of attempts made using ANN to solve problems of sediment prediction since 2002 to 2011 are reported here. The review mainly focused on the infrastructural implementation of ANN for successful prediction. Nagy et al. [3] predicted sediment load in rivers by using multilayer feed forward neural network with back propagation training algorithm and compared the results with conventional sediment load formulas. They used eight parameters which include tractive shear stress, velocity ratio, suspension Artificial Neural Networks Modeling in Water Resources Engineering: Infrastructure and Applications M. R. Mustafa, M. H. Isa, R. B. Rezaur W World Academy of Science, Engineering and Technology International Journal of Civil and Environmental Engineering Vol:6, No:2, 2012 128 International Scholarly and Scientific Research & Innovation 6(2) 2012 ISNI:0000000091950263 Open Science Index, Civil and Environmental Engineering Vol:6, No:2, 2012 publications.waset.org/267/pdf