Application of soft computing and intelligent methods in geophysics (Q721123)

This is certainly a useful book offering novel approaches of computational techniques related to a variety of geophysical applications widely using Matlab and its toolbox algorithms. It can be interesting both for those active in pure computational techniques as well as for professional researchers looking for practical solutions of problems in the fields of geophysics and geodesy. The book is also a highly instructive literature for graduate students as it provides many exercises and practical examples of how to use and apply the described numerical methods in treating concrete geophysical problems. The text written in 17+ 533 pages is conceptually divided into four main parts whose chapters are followed by lists of references. However, there is no subject index at the end of the book. Part I is introductory and its two chapters deal with the essentials of artificial neural networks in computational processes. Thus, Chapter 1gives details about how such networks are designed and constructed, their functioning characteristics including machine learning abilities and testing procedures. The next Chapter 2 is an introduction in numerous possibilities of application of neural networks in geophysics. Among other things, this includes studies of gravitational anomalies related to detection of various geological and archeological buried structures, seismic activity like picking first breaks, sea level predictions, and estimates of surface settlements and displacements encountered in engineering geodesy. Part II is devoted to details of the method of fuzzy logic which includes fuzzy sets, fuzzy relationships and fuzzy numbers as given in Chapter 3, and its application in solving particular geophysical problems as presented in Chapter 4 which includes classification of volcanic activities, earthquake forecasting, analysis of geomagnetic data, remote sensing, mudslide behavior imaging, and shape and locations estimates of micro gravity anomalies. Part III contains computational methods that combine artificial neural networks and fuzzy logic approach. In this sense, Chapter 5 introduces different types of neuro-fuzzy systems with detailed analyses of their properties including example exercises using the corresponding Matlab toolbox. Applications of the described procedures to geophysical problems are given Chapter 6 which includes processing of microgravity data, estimates of ground motions and settlings occurring during various civil engineering activities, determination of the Earth surface electric parameters, identification and differentiation of week earthquakes from recorded disturbances artificially caused by blasts, and prediction of the Earth rotation parameters needed in accurate positioning procedures. The final Part IV with its Chapter 7 is on genetic algorithms and their application in solving inverse problems of recovering the Earth subsurface physical parameters from known measurement data obtained by different methods at the ground. After presenting the basic concept of generic algorithms, two examples of application enlighten the applied computing procedure for specific assumptions regarding the number of involved parameters and type of optimization.