StreamLab Collaboratory: Experiments, Data Sets, and Research Synthesis
Field studies have documented that the average bedload transport rate in gravel bed streams depends on the sampling time (or integration time) over which this average is computed. In this paper we use sediment transport data from a controlled laboratory experiment to document and quantify the dependence on sampling time not only of the mean but of the whole probability density function (pdf) of sediment transport rates in a gravel bed stream. We demonstrate that the higher moments (variance and skewness) scale differently than the mean and we provide a concise parameterization of this statistical scale-dependence. The results indicate that the mean sediment transport rate in moderate flows decreases with increasing sampling time, in agreement with results reported in field studies. The proposed methodology provides a framework within which to seek universal scaling relationships, compare results of different samplers, and also interpret extremes.