Characterizing the spatial structure of river morphology and hydraulics: Remote mapping and geostatistical modeling of dynamic fluvial systems

A river’smorphologic evolution reflects a complex suite of interactions between the form of the channel and the flow and sediment transport processes acting to modify that form. Efforts to better understand these interactions thus require an ability to measure river morphology and characterize its spatial structure. The primary objectives of this dissertation are to develop such a capacity and to apply these methods to the study of channel change. The feasibility of passive optical remote sensing of river bathymetry was examined via radiative transfer modeling, field spectroscopy, and image processing. These analyses indicated the robust performance of a simple, ratio-based depth retrieval algorithm; the conditions under which this technique is appropriate were defined by considering the relativemagnitudes of various radiance components. This physics-based approach allowed the utility of remotely sensed data to be evaluated with a forward image model, which highlighted the influence of both channel and sensor characteristics on depth retrieval accuracy and precision. The potential to obtain a more complete topographic representation of the fluvial environment by combining spectrally-based bathymetry with LiDAR data from bars and floodplains was also demonstrated. Remote sensing thus provides essentially continuous, high resolution data on river morphology, but improved, spatially explicit analytical methods are needed to fully capitalize on this information. This need is addressed by a geostatistical framework consisting of a channel-centered coordinate system and flexible modeling tools for describing reach-scale spatial patterns. Kriging-based spatial prediction techniques can yield geomorphic insight by revealing departures from an a priori expectation of channel form. Similarly, the spatial structure of river morphology and hydraulics can be quantified using a geostatistical metric called the variogram. Survey data from a recently restored channel and three reaches of a dynamic gravel-bed river in Yellowstone National Park documented various styles and degrees of channel change, and variogram models developed from these data were used to examine the relationship between geomorphic context, disturbance history, and the balance between sediment supply and transport capacity. Quantitatively characterizing the variability and spatial organization of channel form will advance our understanding of river morphodynamics