Modeling hydrologic processes at different scales - difficulties and possible solutions

Geoscientists face a host of well-known problems when attempting to model physical processes at different scales. As scale increases, the spatial and temporal resolution of the available data generally decrease and the ratio of the area over which the predictions apply to that for which ground-truthing data is available skyrockets. From our perspective, a primary problem is that with the availability of satellite remote sensing data, the scale at which geophysical models could be applied leapfrogged way beyond our ability to provide matching field-based data. Efforts to improve satellite-based remote sensing systems should therefore be accompanied by detailed studies of the physical controls on the spatial variability of various ecosystem processes, primarily through large-scale field studies combined with low-altitude, as well as high-altitude, remote sensing data. As an introduction to how these, and other, problems can artificially limit the value of remote sensing and other large-scale data sets, we illustrate how these issues have affected a variety of hydrological problems of interest to the Geoscience Research group at the Idaho National Laboratory, including estimation of infiltration rates for large-scale groundwater flow models and examination of the climatic sensitivity of alpine glaciers and basin-scale water balance. The lessons learned in dealing with such problems speak to some of the work that must be done to facilitate application of remote sensing data to a wider variety of geophysical problems. Finally, we propose that modeling studies at a variety of scales would be more useful to society through implementation of what we term the automated modeling-monitoring feedback loop, in which numerical models with simplified interfaces are coupled with environmental monitoring systems to permit manipulation by non-modelers, and which could thereby permit continued calibration as new monitoring data are obtained.