Hello-
Other than modifying "dcrit" in the *.in file, is there any modification needed in the grid file to use wetting and drying? If a certain cell should be available for wetting/drying, is it identified as a "1" in the masks? Thanks in advance...
WET_DRY specifications
Moderators: arango, robertson, rsignell
Neil-
As an early user of roms 2.3, here is some guidance on the wetting/drying. The grid file you develop should use land sea masking.
Cells with a rho mask = 0 will always be dry.
Cells with a rho mask = 1 can be wet (active) or dry (inactive) depending on the depth of water on that cell.
In the present formulation for wetting/drying:
At initialiation the model computes for each cell the total depth of water
D(i,j)=h(i,j)+zeta(i,j,1)
If the total depth is less than a user specified minimum depth (Dcrit), then the water level is increased to = Dcrit - h(i,j).
This 'drapes' the water level over all the land cells.
During computation the water level in each cell is compared to Dcrit. If the water level is less than Dcrit than no flux is allowed out of that cell. Water can always flow into a cell.
As an early user of roms 2.3, here is some guidance on the wetting/drying. The grid file you develop should use land sea masking.
Cells with a rho mask = 0 will always be dry.
Cells with a rho mask = 1 can be wet (active) or dry (inactive) depending on the depth of water on that cell.
In the present formulation for wetting/drying:
At initialiation the model computes for each cell the total depth of water
D(i,j)=h(i,j)+zeta(i,j,1)
If the total depth is less than a user specified minimum depth (Dcrit), then the water level is increased to = Dcrit - h(i,j).
This 'drapes' the water level over all the land cells.
During computation the water level in each cell is compared to Dcrit. If the water level is less than Dcrit than no flux is allowed out of that cell. Water can always flow into a cell.
I just want to say that some compilers (well, optimizers) have issues with the WET_DRY code. If it isn't working for you, try backing off from say -O3 to -O2 in your compilation. This has worked for me for both IBM's xlf and for Pathscale. Next will be to figure out which routine(s) are the culprits and report it to the vendors (may take time).