The need to estimate velocity and discharge indirectly in gravel-bedded rivers is a commonly-encountered problem. Semilogarithmic friction equations are used to estimate mean velocity using a friction factor obtained from depth and grain size information. Although such equations have a semi-theoretical basis, in natural gravel-bed channels, an empirical constant (6.8 or 3.5) has to be introduced to scale-up the characteristic grain size (Ds0 or DS4) to represent the effective roughness length. In this paper, two contrasting approaches are used to suggest that the multiplier of characteristic grain size is attributable to the effect of small-scale form resistance, reflecting the occurrence of microtopographic bedforms in gravel-bedded environments. First, spatial elevation dependence in short, detailed bed profiles from a single gravel-bedded river is investigated using semivariogram and zero-crossing analyses. This leads to objective identification of two discrete scales of bed roughness, associated with grain and microtopographic roughness elements. Second, the autocorrelation structure of the three-dimensional near-bed velocity field is examined to identify regularities associated with eddy shedding and energy losses from larger grains and microtopographic bedforms. Apart from improving the capacity to determine friction factors for velocity and discharge estimation, the findings have implications in general for the initial motion of gravelly bed material.

No abstract

No abstract

No abstract

ABSTRACT1. Improved linkage between physical characteristics of rivers and biological performance or potential is a recurrent theme in contemporary river survey, management and design. This paper examines the degree to which flow biotopes and functional habitats may be differentiated with respect to physical habitat delimiters, i.e. flow depth, velocity and Froude number.2. Re-examination of published data demonstrates only very broad associations between biotopes, functional habitats and 'low', 'medium' and 'high' bands in the chosen physical habitat delimiter. The associations are also not consistent between different delimiters.3. Re-analysis is complicated by considerations of research design and methodology. Further studies are required with greater control on the circumstances of observation (particularly flow stage and seasonality), which explicitly control for variance within and between different river cases, and which adopt more consistent terminology.4. Field data for a single river reach at low and high flows indicate that use of the Froude number, in particular, requires careful interpretation. Very different velocity and depth combinations can exhibit similar Froude numbers. This may obscure important contrasts, such as those between channel margin and channel centreline environments.5. Field data also reveal that the flow characteristics of even basic (riffle or pool) units of channel morphology exhibit strong stage-dependence. There are also significant variations when channel margins are isolated from the definitions of the bedforms or analysis of the data. Current practice which seeks to simplify field survey to channel cross-section transects is therefore likely to be misleading.6. Given these difficulties, attempts to link biotopes with ecological response appear premature. Further research might, however, be directed to identifying possible associations between combinations of flow types and bedforms or functional habitats. In this sense, the biotope concept may be more profitably employed as one of several surrogate measures for potential biodiversity.