HEC-RAS for MR

[BenjaminHartman]

Hey everyone,

I've been teaching myself HEC-RAS for the last 2 months. I am about to graduate and I have a model of the Mississippi River (pieced together by graduate students) that spans from Baton Rouge (River Mile 228.4) to Head of Passes (River Miles 0). I spend most of the time trouble shooting the model and have a particularly annoying problem. From RM 98.1 to 42.7 I am noticing drastic changes in velocity, flow, froude number, and shear. There are no changes in the model on these points that would instigate these effects. The manning's n has been changed to match modeled stage height to USACE observed data. At river mile 38.6-25.2, Bohemia Spillway is installed as a weir that with a height of 1.25 m. With these parameters, it is capturing up to 17,000 cms compared to the 0-3500 cms that is acceptable. I made rating curves to control the flow leaving and adjusted the weir height so water wouldn't spill over. The model already shows flooding in approximately 1/3 of the cross sections but if I raise the height it floods way worse. If I break the diverted flow into individual lateral structures that span the spillway length it still floods. Anyway, I figured the flooding, the drastic changes(which are patterns that occur in the original model), and the spillway might be related. Then, I thought it might be the graduate students who calibrated the stage before checking to see if the flows through lateral structures, cross sections, and reaches matched... maybe if I matched the flows first to observed data then tried to recalibrate it might work. I am so surprised that the stage is calibrated to observed results and the cross sections come from real data and it still floods. Plus there is that crazy pattern in velocity, flow, shear, etc. Has anyone ever seen seemingly inexplicable trends such as these? Does anyone have some tips for me to smooth out my rough edges? I have exhausted parts of the user manual and forums and can probably spend more time there, but spending any more time on this without asking for help seems stupid. I appreciate any feedback.

I'm putting a word document with the graphs pasted on it to show how "drastic" the changes are.

Thanks.

 

[francesca]

Is this a graduate project? Or for a class? I have no doubt that FEMA already has a HEC-RAS model of the lower Mississippi, and I wouldn't be surprised if they don't have a 2D model too of this highly complex river system. Did you know that the lower Mississippi has 70' dunes that migrate down the channel? I don't think that HEC-RAS can accurately model the dynamics in the lower Mississippi.

 

[BenjaminHartman]

Hey Francesca,

It is for graduate work and to add on to the current 2 D model the department already has.

I'll update you on what was happening.

The flooding was occuring because when the work was converted from English to SI units there was some mistake. I went to an earlier version and reconverted it and it ran just like the English units version. But those crazy patterns still exist.

Right now I'm trying to get the flows in the passes to match what I know them to be based on main channel flow and nearby stage data. I am trying to use a lateral structure to route the flow into the passes. For some reason, I don't see any changes in the flow in the passes but the hydro graph of the lateral structure shows water leaving. When you chose "cross section(s) of a river/reach" it lets you set "TW RS". I'm convinced that I must be looking at this wrong. I am setting the "US RS" equal to the point where I want flow to go. The lateral structure is already on the main channel. The point where I want it to go is the beginning of the pass. Am I obviously doing something wrong? See the screenshot!