Relative submergence of macroroughness elements such as boulders and bedrock outcrops, or large bed elements (LBEs), collectively, is a primary control on hydraulics and morphodynamics in steep, coarse-bedded rivers. However, in practice, the property is typically represented by singular, often reach- or cross-section-averaged values that mask bed-surface heterogeneity and joint distributions of local flow depths. By coupling sub-meter resolution two-dimensional (2D) hydrodynamic modeling with spatially explicit mapping of LBEs from a 13.2 km segment of a boulder-bedded mountain river, we present complete distributions of LBE relative submergences at multiple spatial scales and explore their dynamism across discharges. Through distribution fitting and statistical analysis of resultant discharge-dependent LBE relative submergence datasets, it was confirmed that segment- and reach-scale datasets exhibited similar statistical properties and were able to be drawn from the same type of distribution. Further, the rate at which statistical and parametric properties changed between discharge-dependent datasets were statistically equivalent between spatial domains, which we term “process-based similarity”. Commonality in distribution type and the uniform between discharge–scaling relationships suggest mutual self-organizing processes associated with the size-frequency distribution, spatial arrangement, and submergence of LBEs were present between most domains.

Wind vectors, watershed discharge, and subestuarine water levels were monitored in a vegetated delta at the head of the Bush River, an upper Chesapeake Bay tributary in Maryland, during an El Niño/La Niña cycle 1995-1996 to investigate hydrometeorological processes that affect the tidal freshwater ecosystem located there. Time series of these processes were analyzed in both the time and frequency domains using such methods as flood frequency analysis, harmonic analysis, averaged and evolutionary power spectral analysis, and coherency spectral analysis. Wind speed variations with periods of 3-4 and 7 days were found to have both high spectral power and high statistical significance. The frequencies of these variations fluctuated over weeks to months and the amplitudes modulated seasonally, but the variations persisted interannually. Significantly greater subtidal wind speed variations in the prinicipal wind direction occurred during the cold and stormy La Niña winter of 1996 relative to the warm and dry El Niño winter of 1995. Data from five hurricanes occurring in the region during the study provided high-resolution snapshots of the mechanisms revealed by the time series analyses. Water level quickly responded to south-north directed wind speed fluctuations during the aperiodic hurricanes, illustrating the strong coupling between wind and water levels in this system. The magnitude of the response was large enough to determine the extent and duration of flooding over tens of hectares in important intertidal marsh habitats. Subtidal water level variations were greater during the La Niña period. During El Niño conditions, the east-west wind component played a larger role than during the La Niña period. Variations in local watershed discharge as well as Susquehanna River outflow had no measurable impact on water levels in the upper reaches of the Bush River tributary during the study. © 2003 Elsevier Ltd. All rights reserved.