The beds of most gravel-bedded rivers are often organized into patches of similar grain size and sorting.  Although such patchiness probably has a profound influence on bedload transport, bed surface heterogeneities are generally neglected in sediment transport calculations.  Additionally, the spatial organization of the river bed has direct biological implications, since these patches create benthic microhabitats that determine biodiversity patterns and food web structure. 

Theoretical work has been done showing the tendency for bedload sheets to arise through grain sorting interactions, but presently there is no theory for the development of patches in general.  However, we hypothesize that three processes may be important in patch development.

1) Grain to grain interactions, in which large grains are caught in the wakes of other large grains, followed by infilling of their intersticies by smaller particles.  These small particles smooth out the wake effect, increasing the drag on large particles downstream, and mobilize them.

2) Boundary shear stress divergences may be important in that fine patches develop in areas of divergent stress while coarse patches may develop in areas of convergent stress.

3) Finally cross stream slopes and topographic sorting may be important, particularly in channel bends, where a secondary circulation develops.  Fine particles are forced upslope by the secondary circulation, while larger particles are driven by gravity to the bottom of the pool.  This may therefore produce coarse patches in the pool and fine patches up the slope.

We further propose that there are two ‘families’ of patches.  ‘Free’ patches, which freely migrate downstream and which may occur in the form of bedload sheets, are the result of the catch and mobilize process associated with grain to grain interactions.  ‘Fixed’ patches are the result of boundary shear stress divergences or topographic sorting and, since they are a consequence of bed topography, they are more persistent through time.

Thusfar we have focused on ‘free’ patches through experimental work at UC Berkeley’s Richmond Field Station.  There we fed unimodal, sand-free sediment into a 30 meter long, 86 cm wide flume using a low width to depth ratio and we documented patch development and migration in response to stepwise reductions in sediment supply.  The key findings of those experiments were that a) free patches (bedload sheets) developed, b) decreasing sediment supply caused the bed to be less patchy, with coarse patches expanding at the expense of finer patches.  Additionally, data we analyzed from experiments performed in Japan in the 1980’s showed that the migration rate and spacing of bedload sheets is dependent upon sediment supply.  Together, these experiments have given us some insight on the importance of grain-to-grain interactions in patch development and behavior.

Those experiments were primarily 1-dimensional, as we suppressed the development of bed topography by using a low width to depth ratio.  We also participated in the StreamLab06 experiments to observe patch dynamics in 2-dimensional bed topography (resulting from a high width to depth ratio).  In these experiments we expect to see both free patches, resulting from grain to grain interactions, and forced patches, resulting from boundary shear stress divergences due to topography.  We thus should be able to examine the interactions of forced and free patches and work towards a coherent theory of patch development.

Major accomplishments:

Papers
Dietrich, W. E., P. A. Nelson, E. Yager, J. G. Venditti, M. P. Lamb, and L. Collins (2006), Sediment patches, sediment supply and channel morphology, in River, Coastal, and Estuarine Morphodynamics: RCEM 2005, edited by G. Parker and M. Garcia, pp. 79-90, Taylor & Francis Group, London.

Nelson, P.A., J.G. Venditti, and W.E. Dietrich (2005), Response of bed surface patchiness to reductions in sediment supply, Eos Trans. AGU, 86(52), Fall Meet. Suppl., Abstract H51H-04.