Hi Hermjan, 

At a first glance, I do not see a reason for this to happen. I assume that the domain bed slope is equal to the bed slope used for computing the normal flow load distribution.
Be aware that small bed elevation fluctuations are expected and are correct, as the load is computing as if the flow upstream would be normal, while actually you are imposing a hydrograph. How large is your degradation?
Several things come to mind that could help in solving the problem. You could try a simulation with constant flow starting with equilibrium (i.e., also normal). In this case, no changes in upstream bed elevation would be expected, as the normal flow load would be the one maintaining equilibrium without fluctuations. If there is a bug, there would be changes. 
You can also try computing an equilibrium load for your hydrograph externally and set it using the type of boundary condition 1. Finally, I never needed a fixed upstream bed level, but it could be easily implemented. 

Hi Hermjan, 
 
In the figure you sent:

The upstream load seems to be lower than equilibrium, indeed. Be aware that the number in the log file is `qbk` which is sediment transport per size fraction and unit width (i.e., m^2/s). I think the unit should appear in the log file. On the contrary, the input parameter in the morphodynamic boundary condition is `Qbk`, which is the total value per size fraction (i.e., m^3/s). 

Could this be your problem?

Hi Hermjan, 
From your previous email it seems that there was an error in your sediment transport formulation. In any case, you find degragation both upstream and downstream. 

It is not clear to me which type of initial and boundary conditions you are aplying. In any case, when things do not work as expected, it is good to increase in complexity by steps. I would suggest that you do the following. 

1. Simulation starting under equilibrium with a constant discharge (input.ini.initype=4; input.bch.uptype=1; input.bcm.type and input.bch.dotype are overwritten).  I would recommend that you chose as constant discharge the dominant discharge of your hydrograph. In this simulation, you should not observe any degradation neither upstream nor downstream. 
2. Simulation imposing your hydrograph and a constant sediment discharge equal to the equilibrium value computed in Simulation 1 (input.ini.initype=2; input.bch.uptype=1; input.bcm.type=1; and input.bch.dotype=3). In this case, bed level variation at the upstream end is expected, but should not be extreme. At the dowsntream end, you should also expect bed level variation. 
3. Simulation using normal flow load distribution (input.ini.initype=2; input.bch.uptype=1; input.bcm.type=13; and input.bch.dotype=3)