What type of outflow boundary condition should I use?

Hello i have 1 question. What type of boundary condition shoud I use (i have 2D HD FM model), when i have known discharge as an input boundary, but unknow discharge or water level at outflow? The input boundary is time varying and i dont know water level,Q-H curve or discharge at outflow.  In Hec-Ras there is normal velocity where i can specify energy slope, but i dont know how to deal whit this problem in Mike.  In examples there is known time series of discharge at inflow and known time series of  water level at outflow. But what to do when water level is unknown? 
Thank you very much.

Hi,

there are some possibilities:

Firstly: does your outflow occur mainly at the river channel ?

If yes, you can manually calculate a q-H-relation with any 1D HN - software otherwise you can use the so called "Normalabfluss" (normal discharge) after GAUCKLER, MANNING, STRICKLER. This is usually a good assumption.

If no: you should calibrate your outflow boundary due to observations like flooding areas oder other flood land marks.

Generally: you should extend your model to the downstream so far that the downstream boundary does never effect the water levels in your area of interest. This could be very easily done with FM, because you can increase the element sizes off the AOI very strong and the extension of the model to the safe side dows not cost very many nodes and elements.

With this approaches, a Q-h-relation is in my opinion the best controllable and most stable boundary condition.
You define the Q-h-relation as rating curve. You will find the explanation in the MIKE21FM documentation.

Hope that helps,
Christian