Author(s): A. G. Tsakiris
Linked Author(s):
Keywords: Cluster; Microforms; Correlation; Spacing; Roughness
Abstract: Cluster microforms, or simply clusters, are organized groupings (patches) of gravel particles, superimposed on the beds of many gravel streams. Once developed, clusters influence the timing and patterns of sediment transport via a non-linear, bilateral interaction with turbulence, the entrainable sediment and the overall plan geometry of the gravel stream. The resulting spatial complexity of the gravel stream bed from this interaction leads to subjective estimates of the roughness length parameter for clustered reaches. More specifically, the roughness length parameter becomes highly dependent on cluster spatial characteristics, such as cluster size and spacing. Subjectivity in the estimates of these cluster spatial characteristics may, thus, introduce considerable errors in numerical models and bedload predictions that account for the presence of clusters. The objective of this research is to develop a methodology for estimating a characteristic cluster size and characteristic cluster spacing in a clustered bed. To address this objective, the 2D correlation function of a cluster density function was considered. This density function accounted for the presence or absence of cluster particles in a plan view image of clustered beds. A series of artificial clustered beds with predetermined cluster size and spacing characteristics were employed for extracting cluster spatial characteristics from the2D correlation function. Results showed that application of the methodology on clustered beds developed in the laboratory yielded an accurate characteristic cluster size. Furthermore, a good estimate of characteristic cluster spacing of the clustered laboratory beds was also feasible. The findings of this study are important for the determination of a representative roughness length parameter of a clustered reach. The latter could, in turn, improve the accuracy numerical models of flows over ubiquitous roughness elements, such as clusters and of bedload predictions in clustered stream reaches.
Year: 2009