updates to MYNN-EDMF and Thompson #120
Conversation
1. adding a heat source from fires. It is icontrolled by a flag and not turned on yet. Needs more testing. 2. Adding output of surface paramteres: leaf area index (LAI), wilting point and soil field capacity.
joeolson42
requested review from
grantfirl,
Qingfu-Liu,
dustinswales,
gthompsnWRF and
RuiyuSun
as code owners
|
Here's some supporting documentation for this PR: |
1. Removed commented lines in fire feedback to RUC LSM. 2. The coordinates of a test point are passed to the RUC LSM subroutines to make debug printing more manageable. 3. Reverted change in the TRANSF subroutine to original version. The non-linear root distribution needs scientific evaluation in the retro runs.
…ing a limit to protect against the ridiculously large z0s when using thin first model levels in module_sf_mynn.F90
…hermal conductivity is made in module_sf_ruclsm.F90.
…hard snow slab thermal conductivity.
…void FPE errors in some tests
| @@ -2652,7 +2652,7 @@ subroutine snowwater_glacier (nsnow ,nsoil ,imelt ,dt ,sfctmp , & !in | |||
|
|
|||
| !to obtain equilibrium state of snow in glacier region | |||
|
|
|||
| if(sneqv > mwd) then ! 100 mm -> maximum water depth | |||
| if(sneqv > mwd .and. isnow /= 0) then ! 100 mm -> maximum water depth | |||
There was a problem hiding this comment.
@barlage @HelinWei-NOAA Are you both OK with adding this to avoid a floating point exception that is sometimes triggered (e.g. in the rrfs_v1beta_debug_intel test for this PR and in previous PRs: #114)?
There was a problem hiding this comment.
@grantfirl yes, I'm good with this change. I was going to suggest bumping the mwd up to a larger number again, but this should fix the issue. It's still puzzling to me as to why this happens. Will look into it more to track down the root cause.
EDITED! See 2nd paragraph This is all wrong. Actually, this line of code gets the drop and ice sizes from the existing arrays computed by gthompsn's method. However, since the fractional clouds have no cloud water or ice, they will default to the smallest sizes. Hence, what I'm pointing out is that the assumed minimum size of 2.5 microns is being used for the fractional clouds away from places with grid-explicit cloud water. So as I typed above, the chunk of code in |
|
@gthompsnWRF @AndersJensen-NOAA @grantfirl No thanks for bringing this up on the final days of the PR, but I think this is worth looking into for a follow-up PR. This will require more testing an evaluation. I never saw any 2.5 micron effective radii over water. It was more like 3.5 to 6 range, but a lot of them. Anyway, let's handle this off line. Anders and I are willing to look into this very soon. I'm marking this as resolved for this PR. |
|
Ruiyu and Anning,Could you please take a look at this. I am confident when Thompson MP is called in gfs/gefs, the corresponding radii from Thompson NO are used for radiation calculation.
|
|
@yangfanglin Yes. In the GFS, when Thomspon MP is used, the radii from the Thompson MP are used in the radiation calculation. |
|
Hi, Fanglin:
Effective radius (e.g., re_cloud) is calculated in calc_effectRad
of module_mp_thompson.F90, passed into progcld_thompson (e.g., rew
assigned to cld_reliq) of radiation_clouds.f. cld_reliq etc. are used for
radiation calculation in rrtmgp_lw_main.F90 and rrtmgp_sw_main.F90.
Anning
On Thu, Nov 9, 2023 at 4:14 PM Fanglin Yang - NOAA Federal <
***@***.***> wrote:
… Ruiyu and Anning,
Could you please take a look at this. I am confident when Thompson MP is
called in gfs/gefs, the corresponding radii from Thompson NO are used for
radiation calculation.
Fanglin
Sent from my iPhone
On Nov 9, 2023, at 10:30 AM, gthompsnWRF ***@***.***> wrote:
@gthompsnWRF <https://github.com/gthompsnWRF> @AndersJensen-NOAA
<https://github.com/AndersJensen-NOAA> @grantfirl
<https://github.com/grantfirl> Sorry to be slow to respond, but yes, we
use progclduni.
This is all wrong. progclduni uses the GFDL microphysics radii for cloud
water and ice. This is all wrong, especially for Thompson microphysics. I
have a slide deck showing the problems of GFDL's assignment of ice and
water sizes. I thought the pre-calculated Thompson MP radii were carried
into this subroutine if not module.
—
Reply to this email directly, view it on GitHub
<#120 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AKY5N2K7UYTYKM652RFMRDDYDUOLLAVCNFSM6AAAAAA6CZ25KCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQMBUGM2DGOBZHE>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
|
|
The so-called "progclduni" formally designed for MG2/3 make the code
difficult to read, but the principle is the same. Greg used icloud=3 to
make the code easy to read though.
On Thu, Nov 9, 2023 at 5:54 PM Anning Cheng - NOAA Affiliate <
***@***.***> wrote:
… Hi, Fanglin:
Effective radius (e.g., re_cloud) is calculated in calc_effectRad
of module_mp_thompson.F90, passed into progcld_thompson (e.g., rew
assigned to cld_reliq) of radiation_clouds.f. cld_reliq etc. are used for
radiation calculation in rrtmgp_lw_main.F90 and rrtmgp_sw_main.F90.
Anning
On Thu, Nov 9, 2023 at 4:14 PM Fanglin Yang - NOAA Federal <
***@***.***> wrote:
> Ruiyu and Anning,
>
> Could you please take a look at this. I am confident when Thompson MP is
> called in gfs/gefs, the corresponding radii from Thompson NO are used for
> radiation calculation.
>
> Fanglin
>
> Sent from my iPhone
>
> On Nov 9, 2023, at 10:30 AM, gthompsnWRF ***@***.***>
> wrote:
>
>
>
> @gthompsnWRF <https://github.com/gthompsnWRF> @AndersJensen-NOAA
> <https://github.com/AndersJensen-NOAA> @grantfirl
> <https://github.com/grantfirl> Sorry to be slow to respond, but yes, we
> use progclduni.
>
> This is all wrong. progclduni uses the GFDL microphysics radii for cloud
> water and ice. This is all wrong, especially for Thompson microphysics. I
> have a slide deck showing the problems of GFDL's assignment of ice and
> water sizes. I thought the pre-calculated Thompson MP radii were carried
> into this subroutine if not module.
>
> —
> Reply to this email directly, view it on GitHub
> <#120 (comment)>,
> or unsubscribe
> <https://github.com/notifications/unsubscribe-auth/AKY5N2K7UYTYKM652RFMRDDYDUOLLAVCNFSM6AAAAAA6CZ25KCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQMBUGM2DGOBZHE>
> .
> You are receiving this because you were mentioned.Message ID:
> ***@***.***>
>
>
|
Also, added the fraction of grid cell that is burned out.
One more variable is added: fraction of grid cell burned by the fire. This fraction is used to take into account fire's effect on surface albedo. Also, added some debug prints. These changes will change the results only when fire feedback is turned on. The default is .false.
I think you mean radlw_main and radsw_main? |
|
I agree about not dragging anything else into this PR. Too much has already been dragged in. |
|
@dustinswales Please re-review when you get a chance since you requested changes. |
|
All tests are done at ufs-community/ufs-weather-model#1945. @grantfirl @dustinswales Please, go ahead to merge this pr. |
The following PR consists of updates to 3 schemes and related modules: