439 | | |
| 439 | == Some comments on hte vertical discretisation by Jan == |
| 440 | |
| 441 | I worked on the vertical discretization scheme as we outlined it in the meeting. I coded up the scheme we chose and found a few problems which we oversaw during the discussion. |
| 442 | |
| 443 | The list of parameters as I see them today : |
| 444 | - depth_max = OK |
| 445 | - depth_Wmax = OK |
| 446 | - depth_geom = OK |
| 447 | - reason_geom_below = OK |
| 448 | |
| 449 | - depth_lin : is probably a bad choice of words. We decided it to have a region of layers with a constant thickness. So I propose the following name instead : depth_cstthickness. |
| 450 | |
| 451 | - depth_topthickness : this is a parameter we forgot. In order to start the geometric series, without a fixed number of layers, we need to choose the thickness of the top most layer. Furthermore this is an important parameter as it is key in the infiltration of water in CWRR. |
| 452 | |
| 453 | - depth_refinebottom (Boolean) : Aurélien and Tristan have demonstrated that if we want to change the lower boundary condition in CWRR we will need to refine the resolution at the bottom. Using symmetry, we can easily set a refinement at the bottom while respecting all the above parameters. |
| 454 | |
| 455 | There are a few dependences between these parameters. I have tried to put them into the code as well. |
| 456 | |
| 457 | This is all coded up in python for testing purposes( https://forge.ipsl.jussieu.fr/orchidee/attachment/wiki/Meetings/CMIP6/Physic/FinalHydroVert.py). It prints to the screen the levels computed and makes a graphic of the levels for T and W and corresponding layer thickness. So please play around with it and if all are happy we can implement it in the code. |
| 458 | |
| 459 | |
| 460 | |