Comment on Lewin and Brewer ( 2001 ) : ‘ ‘ Predicting channel patterns ’ ’ , Geomorphology 40 , 329 – 339

With the aim of assessing basic alluvial channel planforms as a function of the main determining parameters, a number of stability diagrams have been published during the past several decades, starting with the well-known plots of channel slope versus bankfull discharge of Leopold and Wolman (1957) and Lane (1957). In more recent versions of these stability diagrams, a parameter representing flow energy is plotted against some geometric or grainsize parameter. In one way or another, the diagrams indicate that straight, meandering and braided patterns represent a trend of increasing flow energy (sensu Ferguson, 1987, and Knighton and Nanson, 1993). However, the discriminators could not be used in a truly predictive way, as the value of one or both of the ‘‘independent’’ variables was predicated upon a priori knowledge of one or more geometric properties of the pattern that was to be predicted, such as the bankfull width, depth or slope of the channel (Parker, 1976; Fredsøe, 1978; Struiksma and Klaassen, 1988). Therefore, a diagram was proposed by the first author using the parameters potential stream power (based on valley gradient as opposed to channel gradient) and median grainsize, variables that can be considered almost independent of channel pattern (Van den Berg, 1995). Potential specific stream power, x, was defined as: