Hydrologic complexity and classification: a simple data reconstruction approach

Technological advances, by facilitating extensive data collection, better data sharing, formulation of sophisticated methods, and development of complex models, have brought hydrologic research to a whole new level. Despite these obvious advances, there are also concerns about their general use in practice. On the one hand, it is natural to develop more complex models than perhaps needed (i.e. representations having too many parameters and requiring too much data); on the other hand, it is often difficult to ‘translate’ results from one specific situation to another. Recent studies have addressed these concerns, albeit in different forms, such as dominant processes, thresholds, model integration, and model simplification. A common aspect in some of these studies is that they recognize the need for a globally agreed upon ‘classification system’ in hydrology. The present study explores this classification issue further from a simple phase‐space data reconstruction perspective. The reconstruction involves representation of the given multidimensional hydrologic system using only an available single‐variable series through a delay coordinate procedure. The ‘extent of complexity’ of the system (defined especially in the context of variability of relevant data) is identified by the ‘region of attraction of trajectories’ in the phase space, which is then used to classify the system as potentially low‐, medium‐ or high‐dimensional. A host of river‐related data, representing different geographic and climatic regions, temporal scales, and processes, are studied. Yielding ‘attractors’ that range from ‘very clear’ ones to ‘very blurred’ ones, depending on data, the results indicate the usefulness of this simple reconstruction concept for studying hydrologic system complexity and classification. Copyright © 2007 John Wiley & Sons, Ltd.