Add CLM Lake Model and update Flake #26
Conversation
Updated Flake physics and modified related files
|
The automatic github actions are failing because they're trying to access a version of python that doesn't exist. |
I think that this bug is fixed in #18, so it should be fixed once merged and pulled into your branch. |
There was a problem hiding this comment.
@SamuelTrahanNOAA The changes make sense to me. Most of them are the generalization for the use of lake model. There are important changes in the GFS_surface_composites_post.F90 for non-fractional grid that introduce frac_ice option to make fractional ice work correctly.
| else ! islmsk(i) == 2 | ||
| !-- ice | ||
| call composite_icy(.false.) | ||
| call composite_wet_and_icy |
There was a problem hiding this comment.
@SamuelTrahanNOAA Do we need to have on line 289:
if (frac_ice) call composite_wet_and_icy
or
if(cice(i) < 1) call composite_wet_and_icy
There was a problem hiding this comment.
I haven't tested fractional ice on ocean points, so I don't know if that would work. It is set up to only use fractional ice on lake points.
|
|
||
| public :: clm_lake_run, clm_lake_init, LAKEDEBUG | ||
|
|
||
| integer, parameter, public :: kind_lake = kind_dbl_prec |
There was a problem hiding this comment.
Why does CLM need its own real precision? I.e. Why can't it use kind_phys to be controlled like the rest of the physics?
There was a problem hiding this comment.
It was double precision in the WRF, and we're uncertain of whether it needs to be double precision. For safety's sake, we're leaving it as double precision for now. The kind_lake lets us experiment with kind_lake=kind_phys without having to rewrite the whole file.
There was a problem hiding this comment.
I've added a comment explaining this.
|
|
||
| end subroutine ShalLakeHydrology | ||
|
|
||
| subroutine QSat (T, p, es, esdT, qs, qsdT) |
There was a problem hiding this comment.
@grantfirl note to self: another independent saturation algorithm
There was a problem hiding this comment.
This one is run at double precision (kind_lake=kind_dbl_prec)
There was a problem hiding this comment.
Very minor comments. CLM seems like a very well-written scheme. My comments RE: redefined physical constants do not need to be fixed before merging because there are so many other schemes that still have this, but it would be nice to see it fixed at some point.
|
Apart from 273.15, it isn't immediately apparent to me what those constants are. If someone (@tanyasmirnova?) can tell me, I'll put them at the top of the file. |
|
I can't find 0.61, but I found these: 0.622 = con_eps There is no entry for con_eps in GFS_typedefs.meta, so I'll have to calculate it as con_rd/con_rv. |
con_eps is imported into GFS_typedefs.F90 and has metadata. |
0.61 is con_fvirt or ratio_of_vapor_to_dry_air_gas_constants_minus_one |
…A/ccpp-physics into clm_lake_revert.v9
SamuelTrahanNOAA
requested review from
wd20xw,
XuLi-NOAA,
JongilHan66 and
mzhangw
as code owners
|
Testing is finished on PR #1509, and we are ready to begin merge process. @grantfirl are you able to merge this PR? |
|
@grantfirl @dustinswales @Qingfu-Liu tests are all done at ufs-community/ufs-weather-model#1509. Please, go ahead to merge this pr. |
This PR adds the CLM Lake Model, updates the Flake This PR adds the CLM Lake Model, updates the Flake model, and adds an experimental feature to run both NSST and a lake model on lake points. The FV3_HRRR suite uses clm_lake by default, and the long-forgotten control_flake test now exists. Any test with LKM=1 that validates a restart file will change results.
Issue requesting this: ufs-community/ufs-weather-model#1510