open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi open-telemac,

I'm working on a triple coupling model (GAIA, TOMAWAC, TELEMAC), the simulation works with TOMAWAC and TELEMAC but I'm still trying to make GAIA working with TELEMAC (without TOMAWAC to accelerate the processus in the first time).
The database for the inlet (maritim border) is provided by PREVIMER.
There are three classes of sediments, one cohesive (7%, d1=3.5E-2 mm) and two non-cohesive (d2=0,15 mm and d3=0,375 mm).
When the simulation doesn't allow the advection of NCO, the bed evolution is close to zero but I've got this message : "WARNING, WITH COHESIVE SEDIMENT IN SUSPENSION
AND ONLY BEDLOAD FOR SAND:
IT IS ASSUMED THAT:
1- BEDLOAD OF SAND IS NOT TOTAL BEDLOAD (QSS = 0).
2- EROSION FLUX OF SAND USED TO COMPUTE EROSION FLUX
OF MIXED SEDIMENT IN SUSPENSION IS EQUAL TO 0."

But, when I'm activating the suspension of NCO sediment, the model is diverging along the boundary condition and a high evolution of the bed is observed in the harbour.
The boundary condition is far from the coast (more than 10 km). The distance between nodes varies from 1km in the english channel to 8 m approching the harbour.
I'm giving graphs and the steering file for GAIA is also attached.

Many thanks in advance,

Arnaud