Review of Environmental and Hydrological Systems Modeling by A. W. Jayawardena

This book is quite different from what most books on hydrologic and environmental systems are in that it focuses on modeling techniques, as the title rightly suggests. The author of the book is known for many contributions in the environmental and water resources area, and has brought the book author’s rich and varied experience to bear on the quality of the book.

The subject matter of the book is divided into 14 chapters. Chapter 1 is introductory and defines systems, and discusses general systems theory and equifinality. Chapter 1 concludes with scoping out the book and laying out its organization. It is a short chapter but written nicely, capturing what to expect in succeeding chapters.

Chapter 2 reviews the historical development of hydrological modeling. The material presented in the chapter can be divided into (1) linear systems approach to rainfall-runoff modeling as well as flood routing, (2) hydraulic approach, (3) conceptual watershed models, and (4) physically based models. The chapter also discusses guiding principles and criteria for choosing a model, and challenges in hydrological modeling. The scope of this chapter is so vast that one can easily write a whole book and therefore it is difficult to do justice to its theme. Nevertheless, the book author has attempted to capture the essence relatively well. This is the only chapter in the book where real hydrology is discussed.

Population dynamics constitutes the theme of Chapter 3. Starting with Malthusian growth model, it goes on to discussing the Verhulst growth model, predator-prey model, Gompertz curve, logistic map, cell growth, bacterial growth, and radioactive decay and carbon dating. Hydrologists are familiar with some of the material and environmental engineers are familiar with most of the material presented in the chapter. What is interesting is that there is a great deal of commonality between concepts and models presented, and those that hydrologists routinely use. Thus, Chapter 3 serves as a good connecting link. Chapter 4 deals with reaction kinetics. Discussing what reaction kinetics is, it presents the (1) Michaelis-Menten equation and (2) Monod equation, two of the more commonly used equations. It is a very short chapter.

Chapter 5 is on water quality systems. It begins with a discussion of dissolved oxygen systems and moves on to discussing water quality in completely mixed water bodies, and water quality in rivers and streams. The focus is primarily on chemical aspects, although water quality has a much broader connotation. Nevertheless it is a well-written chapter.

Longitudinal dispersion is discussed in Chapter 6. It first presents governing equations and then discusses the dispersion coefficient, numerical solution, and dispersion through porous media. The last part of the chapter is on general purpose water quality models. Although brief, it is a nicely written chapter.

Chapter 7 discusses time series analysis and forecasting. Beginning with a discussion of basic properties of a time series, it presents statistical parameters, tests for homogeneity, components of a time series, trend analysis, periodicity, stochastic component, residual series, forecasting, synthetic data generation, autoregressive moving average with exogenous input (ARMAX) modeling, Kalman filtering, and parameter estimation. Chapter 7 is well-written and is a good prelude to time series analysis.

Artificial neural networks (ANNs) are discussed in Chapter 8. For nearly 2 decades there has been an explosion of applications of ANNs in the environmental and hydrologic area. Beginning with a discussion of where they originated, the chapter presents basics of ANNs, unconstrained optimization techniques, perceptron, types of activation functions, types of ANNs, learning modes and learning, back propagation (BP) algorithm, feedback systems, and problems and limitations. Chapter 8 is concluded with highlighting several application areas. On the whole the chapter is good and provides a good perspective on ANNs.

Radial basis function (RBF) neural networks are discussed in Chapter 9. It deals with interpolation, regularization, generalized RBFs, normalized RBFs and kernel regression, learning of RBFs, curse of dimensionality, performance criteria, and comparison of multilayer perceptron (MLP) versus RBF networks. Chapter 9 is concluded with a short discussion of applications. Chapter 9 is well-written.

Chapter 10 deals with fractals and chaos. Starting with fractal dimensions, it gives several examples of well-known fractals and perimeter-area relationship of fractals. Then it discusses chaos, presents some definitions, invariants of chaotic systems, examples of known chaotic attractors, and applications of chaos. It provides a good background for those who are not familiar with the subject and is well-written.

Dynamical systems approach of modeling is the focus of Chapter 11. Beginning with a discussion of random versus chaotic deterministic systems, it presents time series as representation of a dynamical system, embedding, phase space reconstruction, phase space prediction, inverse problem, nonlinearity and determinism, noise and noise reduction, and application areas. The discussion is very useful and nicely presented.

Support vector machines constitute the subject matter of Chapter 12. It starts out with a discussion of linearly separable binary classification and moves on to soft-margin binary classification, support vector regression, parameter selection, kernel tricks, quadratic programming, limitations and problems, and application areas. Chapter 12 is a good introduction to support vector machines.

Fuzzy logic systems are presented in Chapter 13. It deals with fuzzy sets and fuzzy operations, membership functions, fuzzy rules, fuzzy inference, neurofuzzy systems, and adaptive neurofuzzy inference systems, and concludes with application areas. Chapter 14, the last chapter, is on genetic algorithms (GAs) and genetic programing (GP). It discusses coding, genetic operators, parameters of GA, and genetic programming, and concludes with application areas. Chapter 14 is a good introduction.

The book is a good introduction to some of the modeling techniques that are being employed these days. There are other techniques, such as copula, copula entropy, operation research techniques, and so on, that have not been dealt with. Of course, no book can cover everything and this book is no exception. It is a well-written treatise and presents modeling techniques under one cover that are scattered in the literature. The book will be of great value to graduate students and faculty.