Cross Section Orientations

[krista1223]

Hi,
We are building a model for a small tributary for the Ohio River. Our trib is perpendicular to the Ohio and has two bridge crossing approximately 100 ft upstream of the Ohio's normal level banks. Our concern is mostly with the high level, 100-500 year events, which puts the Ohio River very high and our trib all in backwater. But, in order to create cross sections that capture the 500-year event, we've found that we need to either cross the roadway or run our downstream cross sections parallel to the contours and never reach the max level. The area downstream of the roadway crossing is much lower than the maximum flood level. Any thoughts on what we should be doing?

I attached a rough drawing that shows our model area. Green lines are our cross section locations, shaded regions are the roadways, the blue lines are the 500-year floodlevels for the Ohio.

Thanks
Krista

[jarvus lunt]

It is a little tough to see what is going on even with the attachment, but it is an interesting problem. It looks like at low flow, the junction is downstream of the bridges, but at high flow the bridges are essentially inside of the junction? Where you have drawn the two question marks, do you think this flow is parallel to the tributary and still entering the Ohio, or is this flow already in the Ohio and is parallel to the Ohio?

In either case, I can't imagine that "run our downstream cross sections parallel to the contours and never reach the max level" is a good idea. I would be more inclined to either dogleg the trib cross sections lines to the right to capture the high ground (if this is trib flow) or extend the Ohio cross sections over the roadway (if this is Ohio flow).

In general, you want the flow to stay perpendicular to the cross section lines, even if you have to bend the cross sections to do that. If the flow direction changes radically between low flow and high flow, RAS may not be able to capture both behaviors with the same data set. Or, to model it with a single model, might require a more ad hoc approach.

[Paul M]

You may be past needing an answer to this post, it was a month old...but...anyway, you have it set up correctly. Make your cross-sections just like you have them, and then enter the floodplain elevation as a known water surface elevation at the downstream boundary and run a subcritical flow analysis. Don't know if you will like the answer if the stream flow then comes up over the bridge deck, but it should give you the correct water surface profile for the stream and bridge. You don't need to worry about the flow quantity and direction in the river, just the river water surface as the downstream boundary condition. Good luck!