BC invalid (not related to dry cells)

Question

Hi Chris,

2 questions please:

1) I have a reservoir model with 4 rivers inflowing. Their 4 BCs are defined as paths. These paths run from a cell adjacent to land, all the way to permanently inundated cells. River A and B give "non valid BC..", "inflow ignored" from the first time step. After trying a few things I realised that, starting the BC path a cell away from land, all rivers start flowing in well, but when a big peak flow is supposed to come in, rivers A and B stop flowing. (Note; peak flow>big WL rise and wind speed;cyclone conditions)

2) An outflow cell will give the same (invalid bc) problem sometimes, other don't, depending in where I place it. I have tried cells at different depths and distance to land and cant see any pattern. Any suggestions on what can be the problem?

Thanks in advance.

Answer 1

The inflow ignored message appears when there is no area for the water to flow in/out.

For a bc cell to be valid it must be a wet cell that has a ghost cell on the specified face(xp, yp etc), it also must be within any inflow_mindapeth or inflow_maxdepth range specified.

Comments

Thanks Chris,

We have tried several scenarios, all with permanently inundated cells in the BCs. Would you explain us how to identify or implement a ghost cell please?

Cheers,
Isabel

Answer 2

Ghost cells are created automatically in pre and are just cells that are one cell next to the specified domain.  

So flows can only enter or leave the domain at the boundary between an interior cell and a ghost cell.

Consider the grid at http://www.hydronumerics.com.au/images/aem3d_ghost_grid.png 

Column i=1 is a land column, but column 2 has a bathymetry value in the middle of layer k = 2.  Therefore there is a ghost/interior boundary at the xm face of i=2,k=2 but not on the xm face of i=3,k=2

So if the water level is as shown there will not be a valid bc for an inflow to occur

Comments

sorry, I am still confused... Is that grid a map view or a crossection?

---

It's a cross-section, following the normal AEM3D coordinate conventions i = x , k= z

---

Hello, apologies for reviving such an old thread.  Could you please upload your "aem3d_ghost.png" image again?  The link to this image appears broken now.  These "ghost cells" are not very well documented, and I would like to understand them more in relation to setting boundary conditions, especially for troubleshooting boundary condition errors during pre-processing.  Thanks.

---

The image has been restored