Cross-scale modeling of riparian ecosystem responses to hydrologic management

There is much demand for quantitative models to aid in comparison of policy options and design of adaptive management polities for riparian ecosystems . Such models must represent a wide variety of physical and biological fact ors that can vary on space-time scales from meters-seconds to basin-decades . It is not possible in practice to develop a complete model for all variat ion. Incomplete but still useful models can be developed by using state var iable identification methods that focus scientific attention on causal path ways of most direct policy concern, and by using various analytical methods to provide cross-scale analytical predictions about effects of microscale variation. The main value of such models has not been to provide detailed q uantitative prescriptions, but to help identify robust, qualitative argumen ts about efficacy of various policy choices. However, they have not been su ccessful at representing some important dynamic effects in riparian systems , where small physical changes (such as overtopping dikes) and infrequent e xtreme physical events can cause habitat changes at large Spatial scales an d ecological impacts that last for decadal or even longer time scales.