Artificial Neural Networks (ANNs) are being used increasingly to predict and forecast water resources variables. In this paper, the steps that should be followed in the development of such models are outlined. These include the choice of performance criteria, the division and pre-processing of the available data, the determination of appropriate model inputs and network architecture, optimisation of the connection weights (training) and model validation. The options available to modellers at each of these steps are discussed and the issues that should be considered are highlighted. A review of 43 papers dealing with the use of neural network models for the prediction and forecasting of water resources variables is undertaken in terms of the modelling process adopted. In all but two of the papers reviewed, feedforward networks are used. The vast majority of these networks are trained using the backpropagation algorithm. Issues in relation to the optimal division of the available data, data pre-processing and the choice of appropriate model inputs are seldom considered. In addition, the process of choosing appropriate stopping criteria and optimising network geometry and internal network parameters is generally described poorly or carried out inadequately. All of the above factors can result in non-optimal model performance and an inability to draw meaningful comparisons between different models. Future research efforts should be directed towards the development of guidelines which assist with the development of ANN models and the choice of when ANNs should be used in preference to alternative approaches, the assessment of methods for extracting the knowledge that is contained in the connection weights of trained ANNs and the incorporation of uncertainty into ANN models.

Neural networks for the prediction and forecasting of water resources variables: a review of modelling issues and applications

Communities of microscopic plant life, or phytoplankton, dominate the Earth's aquatic ecosystems. This important new book by Colin Reynolds covers the adaptations, physiology and population dynamics of phytoplankton communities in lakes and rivers and oceans. It provides basic information on composition, morphology and physiology of the main phyletic groups represented in marine and freshwater systems and in addition reviews recent advances in community ecology, developing an appreciation of assembly processes, co-existence and competition, disturbance and diversity. Although focussed on one group of organisms, the book develops many concepts relevant to ecology in the broadest sense, and as such will appeal to graduate students and researchers in ecology, limnology and oceanography.

/pdf/the-ecology-of-phytoplankton-4ev70n3jsq.pdf

The Ecology of Phytoplankton

An artificial neural network (ANN) is a flexible mathematical structure which is capable of identifying complex nonlinear relationships between input and output data sets. ANN models have been found useful and efficient, particularly in problems for which the characteristics of the processes are difficult to describe using physical equations. This study presents a new procedure (entitled linear least squares simplex, or LLSSIM) for identifying the structure and parameters of three-layer feed forward ANN models and demonstrates the potential of such models for simulating the nonlinear hydrologic behavior of watersheds. The nonlinear ANN model approach is shown to provide a better representation of the rainfall-runoff relationship of the medium-size Leaf River basin near Collins, Mississippi, than the linear ARMAX (autoregressive moving average with exogenous inputs) time series approach or the conceptual SAC-SMA (Sacramento soil moisture accounting) model. Because the ANN approach presented here does not provide models that have physically realistic components and parameters, it is by no means a substitute for conceptual watershed modeling. However, the ANN approach does provide a viable and effective alternative to the ARMAX time series approach for developing input-output simulation and forecasting models in situations that do not require modeling of the internal structure of the watershed.

Artificial Neural Network Modeling of the Rainfall‐Runoff Process

In this two-part series, the writers investigate the role of artificial neural networks (ANNs) in hydrology. ANNs are gaining popularity, as is evidenced by the increasing number of papers on this ...

Artificial Neural Networks in Hydrology. I: Preliminary Concepts

Review and comparison of methods to study the contribution of variables in artificial neural network models

Rainfall forecasting in space and time using a neural network

http://secure.tutorsglobe.com/Atten_files/54_325534_2_Artificial-neural-network-approach-for-modelling-and-prediction.pdf

Artificial neural network approach for modelling and prediction of algal blooms

A hybrid nonlinear regression (NLR) model and a neural network (NN) model, each designed to forecast next-day maximum 1-hr average ground-level O3 concentrations in Louisville, KY, were compared for two O3 seasons--1998 and 1999. The model predictions were compared for the forecast mode, using forecasted meteorological data as input, and for the hindcast mode, using observed meteorological data as input. The two models performed nearly the same in the forecast mode. For the two seasons combined, the mean absolute forecast error was 12.5 ppb for the NLR model and 12.3 ppb for the NN model. The detection rate of 120 ppb threshold exceedances was 42% for each model in the forecast mode. In the hindcast mode, the NLR model performed marginally better than the NN model. The mean absolute hindcast error was 11.1 ppb for the NLR model and 12.9 ppb for the NN model. The hindcast detection rate was 92% for the NLR model and 75% for the NN model.

https://www.tandfonline.com/doi/pdf/10.1080/10473289.2000.10464228

A comparison of nonlinear regression and neural network models for ground-level ozone forecasting.

A physically based rainfall forecasting model for real-time hydrologic applications is developed with emphasis on utilization of remote sensing observations. Temporal and spatial scales of interest are lead times of the order of hours and areas of the order of 10 km2. The dynamic model is derived from conservation of mass in a cloud column as defined by the continuity equations for air, liquid water, water vapor, and cloud water. Conservation of momentum is modeled using a semi-Lagrangian frame of reference. The model state is vertically integrated liquid water content in a column of the atmosphere. Additionally, laws of thermodynamics, adiabatic air parcel theory, and cloud microphysics are applied to derive a basic parameterization of the governing equations of model dynamics. The parameterization is in terms of hydrometeorologic observables including radar reflectivity, satellite-infrared brightness temperature, and ground-level air temperature, dew point temperature, and pressure. Implementation and application is described by French et al. (this issue) and involves incorporation of uncertainty analysis and a two-dimensional spatial domain, where the dynamics of the continuous space-time rainfall process are discretized onto a rectangular grid.

A model for real-time quantitative rainfall forecasting using remote sensing: 1. Formulation

In the past, flash floods were characterized by rainfall and limited effort was made to examine them from a runoff perspective. For example, the question of how flash floods differ from other floods remains to be investigated. In order to satisfy this need, runoff hydrographs of 30 flood events from four watersheds in eastern Kentucky are selected and analyzed. In flash flood situations, certain portions of the hydrograph, such as the rising limb, time to peak, and magnitude of the peak, are important. To recognize this, runoff hydrograph characteristics such as the rising curve gradient, K, flood magnitude ratio, M, and flood response time, TP, are evaluated directly from observed floods. Using these characteristics, relative severity factors RK, RM, and RT are determined and summed to obtain a flash flood index, RF. Results from this study show that a reported flash flood event of May 18, 1995, has values RK = 3, RM = 16, and RT = 6 and a flash flood index, RF = 25. Overall, half of the flood events hav...

Mark N. French

Papers

Rainfall forecasting in space and time using a neural network

Artificial neural network approach for modelling and prediction of algal blooms

A comparison of nonlinear regression and neural network models for ground-level ozone forecasting.

A model for real-time quantitative rainfall forecasting using remote sensing: 1. Formulation

Characterization of flash floods in eastern kentucky