Description
Neural network hydrological modelling has reached the end of the beginning: strong technological barriers and steep learning curves have both been defeated – such that hard(er) scientific questions can henceforth be tackled from an informed position of [i] accumulated skills and [ii] 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 modelling. Each chapter is the work of one or more independent experts in a different field of hydrological modelling – 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.
