High resolution,basin extent observations and implications for understanding river form and process

Fifty years of fluvial studies have posited a variety of conceptual frameworks for characterizing river forms and processes throughout entire basins, including hydraulic geometry, the river continuum concept, self‐organized criticality, and sediment links. This article uses basin‐extent, high resolution observations of fluvial forms in the Nueces River basin, Texas, and Yellowstone National Park to evaluate the ability of these frameworks to characterize system behavior across a multitude of scales. The Nueces data were collected with remote sensing methods and the Yellowstone data were collected through extensive field surveys. The data resolution, spatial extent, and quality of these data sets allow direct comparison between the two areas. The ‘hyperscale’ comparison supports using of each these frameworks at specific scales, but also indicates an irreducible amount of variation in both datasets across many different scales that is not captured by the conceptual frameworks. Moreover, the scales and locations where one framework, such as hydraulic geometry, works well are often not the same scales and locations where another framework, such as the river continuum concept, works well. Because the conceptual frameworks appear to operate at scales and locations distinct from one another, the measurement approaches necessary to observe them must also be at different scales and locations. For example, ‘seeing’ self‐organized criticality in a river system is difficult without an extensive survey through space, whereas the recognition of sediment links requires quite intense sampling in specific river regions. We suggest that these separations between measurement scales represent an incommensurability issue in river studies, making it very difficult to both communicate among and test between two or more competing theories. Making simultaneous hyperscale observations of the river is one approach to minimizing the theory‐ladeness of observation, as deviations from different predictions can be plotted at every scale. Copyright © 2010 John Wiley & Sons, Ltd.