Hi Engelbert,
some remarks and hopefully one at least is helpful
You continue the computation from a hotstart file. Maybe you should reset the time to zero if you start the hydrograph in the liq file at time 0.
In case you did: you shouldn't continue the FV simulation from a hotstart file which you simulated with FE, especially if you had TREATMENT OF NEGATIVE DEPTHS =1 in the FE options which allows negative water depths.
By the way, the FV HLLC scheme in Telemac doesn't compute the turbulent stress terms in the SWE equations, therefore the keywords for turbulence model and the velocity diffusivity are obsolete.
For checking what is going on in the first time step you should set LISTING PRINTOUT PERIOD equal to 1. Sometimes this can be very helpful.
Hope this helps and greetings from Graz!
Clemens