The meandering-braided river threshold: A reappraisal

Abstract

The concept of a threshold discharge-slope (Q-s) combination that separates braided from meandering streams is critically re-examined, partly by review of previous data, and partly by analysis of new data from South Island, New Zealand. It is argue d that discriminant functions that use mean annual discharge provide a poor physical basis for interpretations based on natural processes, while bankfull discharge is also deficient because it is not a fully independent variable. The common use of channel (rather than valley) slope adds a further misleading component because channel slope is partly dependent on channel pattern sinuosity. Most previous studies have inadequately treated the role of bed particle size in pattern discrimination, failing to recognize that active gravel streams must plot higher on a Q-s chart than sand-bed channels — irrespective of pattern — because of the greater power requirements for bed material movement. Within any one size class of bed material there is no evidence to indicate a clear discrimination between braiding and meandering, only a weak statistical association between pattern and slope-discharge values. While this association may be explicable in terms of the high shear stress and stream power (that accompany high Qs-values) promoting braiding, the real prerequisite for braiding appears to be high loads of bed-calibre material (producing wide shallow channels), a factor that is only partly controlled by stream power. Accordingly, the search for a pattern threshold based on discharge and slope seems to be a futile exercise. An interpretation of existing pattern discriminant functions from the perspective of the threshold shear stress for particle movement is made. It is concluded that, in effect, these functions merely state that gravel-bed streams are more likely to be braided than channels in finer sediment. Statistically this appears to be valid, and reasons for that difference are offered based on the balance between imposed bed material load and channel capacity.