Simplified Channel Evolution Model (CEM)

The Global CEM Option is Not Working in Version 6.6

The Global Simplified Channel Evolution Model is not working in version 6.6.  This will be fixed in the second 6.7 beta release and the final 6.7 release.  But as a work around, you can specify the Simplifed CEM and the parameters in the XS Specific bed change option


The veneer method is the default method for channel erosion and is available (though rarely appropriate) for floodplain erosion.  HEC-RAS includes one additional channel erosion method: the Simplified CEM method.  The Simplified Channel Evolution Model is also designed for reservoir applications, in particular dam removals and reservoir flushing models.

The veneer method can under predict erosion in reservoir deposits.  When a river scours reservoir deposits, following either a dam removal (Echevarria, 2012) or a reservoir flushing drawdown, it does not remove sediment uniformly from the flat reservoir deposits.  Rivers tend to form one or more distinct channel through reservoir deposits during drawdowns or dam removals.  The shear and hydraulic radius of a developing channel will erode more sediment than removing an even veneer from the top of a reservoir deposit.  But these channels will also reach an equilibrium slope faster than a veneer approach and leave sediment embankments behind.  The Simplified CEM model approximates these processes by eroding sediment in a defined, expanding, channel rather than a uniform veneer.   

HEC-RAS translates parameters from the Erosion Bed Change Option into a simplified channel evolution model, eroding sediment in the shape of a trapezoidal channel with the specified parameters (after Cui et al., 2005 and Cantelli et al., 2004).

HEC will only apply this method to cross sections with Max Width and Side Slope data.  The program will erode all other cross sections with the veneer method. Therefore, in a dam removal or reservoir flushing model, users can specify these variables for the cross sections with reservoir deposits to cut through and leave the rest to respond with the veneer method.

Max Width: The Simplified Channel Evolution Model assumes that the channel will expand until it reaches a maximum width (after Morris et al., 2008).  Once it scours to the Max Width it incises, maintaining that trapezoidal bottom width, until it reaches the bottom of the sediment volume (Max Depth or Min Elevation from the Initial Conditions and Transport Parameters Tab).  If the river still has capacity to erode when the channel reaches the bottom of the sediment control volume, the trapezoidal channel will expand laterally, maintaining the bottom elevation and side slopes, but increasing the bottom width.

Selecting a maximum bottom width is difficult and uncertain.  Quantitative tools can help estimate this variable.  Dimensionless analysis (Parker, 2008), regime theory, and other empirical geomorphic equations can guide bottom with estimates.  Adkinson (1996) fit a regression to the bottom widths of channels scoured during several reservoir flushing events, and the equation is included in HEC-RAS (press the Width Calculator… button).  However, all of these relationships have a lot of scatter.  Selecting a bottom width requires good, regional, soil specific, geomorphic intuition.  Even the best estimates are still uncertain and are candidates for calibration parameters or sensitivity analysis (Echevarria, 2012).

Side Slope: The side slopes define the shape of the trapezoidal channel.  While empirical equations can guide Max Width estimation, there is little guidance for Side Slope selection.  Angle of repose is a good place to start, but sensitivity analysis should also quantify the impact of this uncertain parameter.

Center Station (opt) - XS Specific Only: HEC-RAS centers the trapezoidal channel between the movable bed limits.  Because channels often form randomly in reservoir sediment and, often, a single channel is a numerical surrogate for multiple developing channels, this assumption is often good enough, given the other uncertainties.  However, reservoirs that are flushed regularly, often form channels in the general location.  If it is advantageous to capture that lateral channel position either computationally or, simply visually, specify the channel center station here.

Eroding Valley Walls

The Simplified Channel Evolution Model removes sediment from the cross section at the specified side slope without considering physical limits of channel morphology.  For example, if the side slope extends through a bedrock cliff or reinforced valley walls, the algorithm will still clip the cross section at the specified slope, ‘eroding’ the bluff or bank.  Monitor the method carefully to assure physically reasonable results.

Eroding to Bedrock

The Channel Evolution Model will often erode to the bottom of the sediment control volume, particularly if the reservoir was built on a previously bed-rock channel, or if the model is designed to return the channel to the historic grade (e.g. if there is a coarse, historic, cobble layer buried underneath the reservoir silt deposits, and the modelers decide to limit scour to that historic control).  Previous versions of the Simplified CEM model would oscillate once HEC-RAS scoured to the bottom of the sediment control volume, because there was no bed sediment available to compute capacity.  Recent versions of HEC-RAS partition capacity based on the initial bed gradation if the cross section erodes to the Max Depth or Min Elev (by either the CEM or the Veneer Method).  HEC-RAS will also throw a runtime error, if the model scours to the bottom of the control volume:

This is often a sign that the model is eroding too aggressively (because the bed is too fine or the transport function to powerful) or that the alluvial control volume was too small.  But in cases where the model is expected to erode to a known vertical control, this error is expected and acceptable.