In spite of strenuous efforts over the last half century, the engineering prediction of flow and sediment transport over bedforms, in general, and dunes, in particular, still presents a major obstacle in the solution of sedimentation problems in alluvial channels. Fig. 1 illustrates the presence of dunes in the field ~the Rhine River in the Netherlands!. The basic problem of sediment transport can be stated in a deceptively simple manner: given sediment and fluid characteristics, channel geometry, and discharge, what is the expected flow depth and the amount of sediment that will be transported? Even for the simplest case of a well-sorted sediment in a straight prismatic channel with rigid banks, a general answer to these two questions ~to be referred to hereafter as the basic questions! can only be given with an uncomfortably high degree of uncertainty, and much of the difficulty can be traced to the development of bed forms.