Sampling non-uniform gravels exhibiting spatial variability

Current methods of sampling spatially heterogeneous deposits of non uniform gravels require either enormous amounts of material to be sampled or result in biased estimates of a river reach's overall grain size distribution. Such techniques do not provide quantitative information on where significant changes in the sediment grain size occur either. Instead they rely on the ability of the operator to visually identify sudden changes in the sediment grain size distribution such as those that often occur at two distinct facies boundaries. However, when studying the effects of particle size on bed load transport rates across different patches of material or the phenomena of downstream fining, one may be interested in detecting gradual changes in a grain size parameter that can not be visually located. Therefore, a recently proposed sampling technique using statistical hypothesis testing and a moving window approach has been developed to provide specific guidelines on how grid samples can be used to quantitatively locate specified changes in the makeup of bed material. The utility of this proposed method in evaluating downstream fining and helping quantify bed-load transport rates of individual patches of material is briefly expanded upon here.