Neural Networks for Hydrological Modelling

Robert Abrahart, P.E. Kneale, Linda M. See905809619X, 9789058096197, 9780203024119

Neural network hydrological modeling has reached the end of the beginning: strong technological barriers and steep learning curves have both been defeated – such that harder scientific questions can henceforth be tackled from an informed position of accumulated skills and reported facts. The revised challenge is thus twofold: to establish significant questions that will help push back the boundaries of science; and to exploit, or delimit, the various benefits and drawbacks that are on offer. This book is intended to ease the transition into a data-driven paradigm and to provide a stimulus for further research on neural network hydrological modeling. Each chapter is the work of one or more independent experts in a different field of hydrological modeling – which demonstrates the wide potential of such tools, for solving real world problems throughout all areas of hydrological interest, and at all levels of methodological investigation. The initial material provides a basic introduction to the concepts and technologies involved; subsequent chapters use hydrological examples and case studies to illustrate the practical application of different mechanisms and approaches. The final chapter suggests some useful research directions, based on lateral thinking and alternative strategies or mindsets.