A new methodology for velocity estimation in sheet and rilled overland flow using modified Manning's roughness coefficient

Abstract

Manning's roughness coefficient for flow over bare soil is needed in order to determine other hydraulic parameters such as flow depth and velocity in overland flow, parameters needed to calculate streampower and shear stress in erosion process studies. A modified equation of Manning's n was developed to estimate flow velocity in shallow overland flow in experiments undertaken in a 5.8 by 1 m flume of the Griffith University Tilting flume Simulated Rainfall facility and adopted data from the literature. Even with the flow of water over a soil surface in which roughness elements are well inundated, and in less erosive situations where erosional bed forms are not pronounced, the magnitude of resistance coefficients in equations such as those of Manning, Darcy-Weisbach or Chezy vary with flow velocity (at least). Using both original laboratory and field data, and data from the literature, the paper examines this question of the apparent variation of resistance coefficients in relation to flow velocity, even in the absence of interaction between hydraulics and resulting erosional bed forms. Resistance equations are first assessed as to their ability to describe overland flow velocity when tested against these data sources. The result is that Manning's equation received stronger support than the Darcy-Weisbach or Chezy equations, though all equations were useful. The second question addressed is how best to estimate velocity of overland flow from measurements of slope and unit discharge, recognizing that the apparent flow velocity variation in resistance coefficients is probably a result of shortcomings in all of the listed resistance equations. A new methodology is illustrated which gives good agreement between estimated and measured flow velocity for both well-inundated sheet and rill flow. Comments are given on the predictive use of this methodology.