Hi,
In your case, the best practice, from my point of view, is to follow the next steps:
1. If the very downstream sector is important for your study, extend your model in the downstream part with 2-3 Km (with real data, if you have it, or even with some plausible date). The reason is that the last sector upstream the downstream boundary condition is strongly influenced by this condition.
2. Open the Simulation file (mandatory), open from here the editor of the boundary condition file -> Select the row with downstream boundary condition -> than go to Tools (menu) -> "Auto Calculation of H-Q Table..." -> Choose Manning Formula and fill in the fields Slope and Manning's M values with average values used in the downstream sector.
3. Now, a calculated Q-h relationship for your downstream boundary condition is set up.
Please, see the attachment.
Kind Regards,
Aurelian