[rabrownie]

I am attempting to model an anthropogenically disturbed canyon that does not exhibit properties fundamental to fluvial systems. The main channel in this canyon was pushed to the side of the wall, and a remant channel existings in the valley center. The main channel recieves all of the flow from usptream as it it tied into the stream network of the entire canyon. The remnant channel recieves oveflows from the main channel. I cannot model the entire valley cross section in these areas, because the remnant channel is lower (by 10 feet) than the main channel, Hec-Ras would push all of the water in the low channel. In order to model this situation I have been using lateral weirs, but I have been getting some weird results.

The first thing I noticed is my velocities through the sections with lateral weirs, even for non-overtopping flows, are very high (30-40ft/s in a creek 15' wide at 3-4%). Adding some interpolated sections helps this.

The flow profiles never converge.

Sometimes there are negative Q's leaving the channel, what does that mean? What is coming over the hill? Doesn't make sense.

Any help or guidance on this would be great, thanks.