I am currently modeling a creek in an urban downtown area that has several split flows during high flows and is conveyed through several culverts and bridges. A 1D model was originally created for the area by a separate party, but it seemed to be lacking in describing the split flows that were likely. I created a 1D/2D model (1D for the main channel, 2D for the overbanks) for the area, and it seems to offer an improved description of the 100-year event. However, there are still issues.
One issue is that the overtopping “bridges” do not transfer flow to overbank areas at the actual bridge location. Some of these bridges are quite long (several hundred feet) and in fact, are not really acting as bridges. Overtopping flow is likely to leave the main system and be conveyed down streets, etc. I should mention that earlier on in the modeling process, there was a discussion between the invested parties on whether or not to model these as culverts or bridges, and bridges were chosen as the prefered choice. So my question is does anyone have a better method to model these long stretches of underground creek? Or is the combined 1D/2D the way to go while accepting some error at these structures? Most places, the depths directly over bridge and in the adjacent overbanks seem to be reasonably close, but there are one or two locations that are noticeably off.