ABSTRACT
The nature of river flow has recently been the subject of debate in the study of gravelbedded rivers (e.g. Yalin, 1992; Biron et al., 1996; Bradbrook et al., 1998; Lane et al., 1999). Buffin-Belanger et al. (2000) set out to confirm the existence of largescale flow structures and develop a technique for improving the visualization of these forms. From this technique, they suggest that the high-speed ‘wedges’ of flow display a complex organization. This paper illustrates a number of points relevant to this chapter. The construction of the entity of study – the high-speed wedges – relies upon the use of specific detection methods and a specific theoretical framework. The new technique of analysis actively constructs the new entity of analysis, the velocity fluctuations in a space-time matrix. Lastly, the outcome of their analysis provides a basis for further discussion of their constructed entities. The entities can be treated as if they were real and so debated amongst other researchers in the field. This does not mean to say that they are ‘real’ or understandable isolated from the measurement system, only that they can be treated as such for discussion.
