There are def ways to do it, but IMO they are just as complex and carry more uncertainty than creating an unsteady state model of the 2 reaches or representing the reach with either a 1D or 2D storage area. Depends how deeply annoyed you want to get with numerical randomness.
A very talented hydraulic engineer once told me that sometimes the difference between a stable and unstable hec ras model is just finding the perfect variables and numerical values entered for stability.