open TELEMAC-MASCARET The mathematically superior suite of solvers

Hi everyone,

I'm running a big 3D model of the Scheldt estuary. We use salinity as an active tracer. Upstream we have flooding areas in case of storm surges. These areas are separated from the river by lower dikes (1 line of points in bathymetry)Only with a storm surge water flows over these dikes into these areas.

When simulating a normal spring-neap tidal cycle I noticed that the tracer moves into these flooding areas. This shouldn't be happening. Can anyone explain me what I might do wrong or why this is happening. I attach a figure of part of the mesh with tracer result and my steering file. the dike is represented in the bathymetry by only 1 point in width.

Sven

Hello,

This is a well known problem when a dyke is represented by only one line of points. Depth at the top is influenced by nearby wet points and get a small quantity of water, that may then flow on the other side. A partial solution has been implemented in 2D (THRESHOLD DEPTH FOR RECEDING PROCEDURE) but not in 3D.

A quick and dirty solution to avoid building a new mesh would consist in clipping the tracers at the end of cvdf3d.f, when the depth is less than a threshold value. This would spoil a bit mass conservation, but it is not too much of a problem in your case. Something like (not tested):

IF(.NOT.VELOCITY) THEN
DO I=1,NPOIN3
IF(H%R(I).LT.1.D-2) FD%R(I)=0.D0
ENDDO
ENDIF

and well, doing this also to velocity can work.

With best regards,

Jean-Michel Hervouet