open TELEMAC-MASCARET The mathematically superior suite of solvers

I'm trying to run my first simulation but some problems have arisen. The model is a coastal lagoon forced with a discharge and tide. Since I have some part of the ocean, I had put 2 open boundaries (free both of them) and one prescribed elevation boundary (the one where the tide comes into the domain). But the simulation crashed after a few time steps and I was getting a message saying that the problem was ill posed due to supercritical flow.

I found that one of the solutions might be to use Thompson BC, so I did it. However I get negative depths (>-2m, around the lagoon) and large velocities at those boundaries. I'm not clear on the meaning of negative depths, so I have been testing by deactivating or activating the tidal flat option. As for now, the situation become worst when I deactivate it. I know that there are options to deal with negative depths in TELEMAC, but I'm more concerned on the meaning and the reason of such negative depths.

I've used before other schemes, FV, and I know that negative depths are a common situation when modelling wetting and drying. But I'm not clear with TELEMAC, because after a few time steps, the whole area that was initially dry, is now with a negative depth. Can someone give me some light on this topic? And advise me about the best way to deal with it, please (later I'll have to add some tracers like salinity and temperature, in case this is relevant).

Ps. I already read the TELEMAC user documentation, but still is not clear to me which is the best way to deal with it.
I add my steering file, in case that someone may take a look at it. I'm using v7p0.

Thanks