HEC-RAS includes multiple methods to vary channel roughness spatially or temporally. HEC-RAS can vary manning n value as a function of flow, stage, lateral station, or season. Calibrating one-dimensional hydraulic models often requires process dependent roughness parameters. Use these features to calibrate a steady or unsteady flow hydraulic model before adding sediment data.

However, bed roughness can also vary based on sediment dynamics. These are discussed in detail in the Technical Reference Manual.

Limerinos:

The Limerinos method is a relatively simple equation based on the d₈₄ and hydraulic radius, with no consideration for bed form mechanics. Therefore, it should be applied mainly to coarse, gravel and cobble systems where the grain roughness is the primary source of channel roughness.

Brownlie:

Brownlie (1983) computed bed roughness based on bed form mechanics in large rivers. In particular, he tried to capture the non-linear drop in roughness when bed form dominated transport shifts from lower to higher regimes (EM 4000). Brownlie, evaluates the bed form regime based on hydraulic parameters, grain size, and the gradational distribution, then applies separate equations for low and high regime transport.

Van Rijn:

Van Rijn (1984) is based on flume and field data. Van Rijn computes bed form dimensions from flow and sediment parameters and converts these into equivalent bed roughness. This method was designed to compute bed roughness in both dune and plane bed regimes. 

Modeling Note: Check Roughness Predictors

1D modelers have had mixed success with automated n-value equations. It can be valuable to calculate a dynamic n-value based on evolving hydrodynamic and sediment conditions, but n-value is a very sensitive parameter in sediment models, and is should be calibrated during the hydraulic modeling. Check the n-values computed by these equations against those selected during the hydraulic calibration. Or run a sediment model without bed-change (by making all the nodes pass-through) to see the range of computed n-values for the hydraulic simulation alone. If the equations are not computing n-values that calibrate the hydraulic model, they are not appropriate for a sediment model.

Reach Averaging Bed Roughness Predictors

Bed roughness predictors can be noisy. Calculating separate roughness parameters for each cross section based on dynamic hydraulic and sediment parameters can make roughness vary erratically both in space and time, sometimes switching regimes from cross section to cross section or time step to time step. The reach average tool smooths these effects.

Group sequential cross sections with similar morphological setting (e.g. flow, slope, gradation) into a 'bed roughness averaging reach' by pushing the button: Select Reach to Average Bed Roughness Predictors. HEC-RAS will average the hydraulic and sediment parameters for the cross sections included in each reach and use the average values to compute a single bed roughness for all included cross sections. This tends to smooth the results in space and time. Any cross section not included in a reach will compute bed roughness locally, based on its own parameters.

To delete a bed roughness averaging reach, double click on the record in the list.