Potential Issue with Velocity Calculation for Compound Channel Flow in module_channel_routing.F

[mrosicky]

While reviewing the channel routing implementation in module_channel_routing.F, I encountered what may be an issue regarding the computation of flow velocity. I believe the current approach might not properly account for the case when flow spills into the floodplain.

The depth variable h, derived from the MC equation, appears to represent the total flow depth, which includes both the main channel and floodplain when there is overflow (i.e., h > bfd where bfd is the bankfull depth) at line 337 and line 354.

However, average velocity is calculated using Manning's equation at line 498, where R is the hydraulic radius based on h at line 497. This equation applies to flow in the main channel only, not the compound flow that includes the floodplain.

If h includes the floodplain depth, the current velocity calculation might be incorrect, as it does not separately account for the floodplain's resistance or hydraulic characteristics.

Expected Behavior

The velocity is calculated with separate components for the main channel and floodplain to ensure accurate modeling of compound channel flow, then combined.

Possible Solution

Calculate separate velocities for the main channel and the floodplain, using their respective hydraulic parameters (e.g., AREAC, WPC). Then, combine the velocities for an accurate representation of the total flow across both the main channel and floodplain.