AI techniques for hydrological modeling and management. I: Simulation

In recent times, a significant amount of research has been carried out into the development and application of artificial intelligence (AI) based techniques for hydrological modeling. The common feature of these methods is that they employ computer agents to perform tasks which require intelligent behavior, such as learning, problem solving and decision making under uncertainty. This can be particularly beneficial in hydrological modeling, where systems are usually characterized by complex, dynamic and nonlinear systems, where the underlying physical relationships are not fully understood and the available data are noisy, incomplete and/or unquantifiable. As such, AI techniques may provide a promising alternative, or complement, to traditional process-based or statistical approaches used in hydrological modeling. This two-part review aims to highlight the strengths and weaknesses of the various AI technologies and identify where they are most applicable through a brief discussion of their advantages and disadvantages together with previous and/or potential applications in the field of hydrology. The first part of this series focuses on prediction and simulation techniques including artificial neural networks, Bayesian networks, fuzzy systems and agent-based models, while the second paper of the series focuses on evolutionary optimization methods and their application to hydrological problems.