This picture shows velocity vectors for the first and the last 10 kilometers at the time 3600s of the simulation, it is obvious that the maximum speed is obtained in the center of the valley. At this time, the calculation has converged to the steady state.
The second image
presents the interaction between the upper flow and the lateral flow at
the top of the hill. The speed of the water coming from the top the hill
is more important than the speed of the water coming from the sides hence
the flow direction is quickly oriented downhill.
On the first picture are presented the water height at time, from rigth to left, 0s, 400s, 900s, 1800s and 3600s. This picture shows the evolution towards a situation where the altitude is constant for a given Y value. This point is shown on the second graph.
Here are presented the water heights at X=0 for the next simulation times : 0s in red, 400s in blue and 3600s hatched, the ground is represented in plain green. We can see on this graph that in middle part of the domain there is little, or no, variation of the water height whereas in the entering zone, the height decreases and in the exiting zone the height increases. We think that the increase of the water height is linked to the choice of the downstream boundary conditions, indeed in this case we impose the flow rate at constant value and it seems to act like there were a flat portion of the ground whereas we don't know what happens after the end of the domain.