Convective Reflectivity #129
Conversation
ericaligo-NOAA
requested review from
grantfirl,
Qingfu-Liu,
dustinswales and
haiqinli
as code owners
|
@ericaligo-NOAA Is it possible to apply the following conditions (from cu_gf_driver_post.F90) when applying GF convection? Thanks. line 73 if(cuprate .lt. 0.05) cuprate=0. |
|
Thanks for pointing that out! I made the change and pushed it to my branch
…On Thu, Nov 16, 2023 at 11:34 AM Qingfu Liu ***@***.***> wrote:
***@***.**** commented on this pull request.
------------------------------
In physics/GFS_MP_generic_post.F90
<#129 (comment)>
:
> +! parameterizations.
+ if ( (imp_physics==imp_physics_thompson .or. imp_physics==imp_physics_nssl) .and. &
+ (imfdeepcnv==imfdeepcnv_samf .or. imfdeepcnv==imfdeepcnv_gf .or. imfshalcnv==imfshalcnv_gf) ) then
+ do i=1,im
+ factor(i) = 0.0
+ lfrz = .true.
+ zfrz(i) = phil(i,1) / con_g
+ do k = levs, 1, -1
+ zo(i,k) = phil(i,k) / con_g
+ if (gt0(i,k) >= 273.16 .and. lfrz) then
+ zfrz(i) = zo(i,k)
+ lfrz = .false.
+ endif
+ enddo
+ enddo
+!
Line 132 defined the variable: onebg = one/con_g, so lines 145 and 147 can
be replaced by multiplication since multiplication is faster than division
for some processors
—
Reply to this email directly, view it on GitHub
<#129 (review)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ALQ75MI7NMR66U56ZOU5UILYEY6BXAVCNFSM6AAAAAA7LI3MWCVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMYTOMZUHEYDQNJTGQ>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
|
Thanks for the comments! I had shown previously that with the conditions included there is inconsistency between the precipitation and reflectivity. See this presentation and more specifically the links on slide 12 to see animations of various fields. I attached one example here. In the example here, the lower left plot is the hourly max precipitation rate valid at 15UTC 08 July 2020. The bottom center plot is the hourly max 1km agl reflectivity valid at the same time with those conditions you noted, and the bottom right is the hourly max 1km agl reflectivity without those conditions. We can see here that we have consistency between precip and reflectivity when we leave out those conditions. Additionally, it would make it difficult to compare reflectivity between GF and saSAS with those conditions. |
|
@ericaligo-NOAA Thank you very much for your information! |
|
I think it's a great idea to pull these reflectivity calculations out of the convection scheme, but I'm not sure if GFS_MP_generic_post.F90 is the ideal place for them, since they rely on both the convection and mp parameterizations. May I suggest to add them to GFS_physics_post.F90? (This was recently created as a landing spot for computations on multiple schemes.) |
|
Thanks for the suggestion. The convective reflectivity needs to be included
in the hourly max reflectivity computations (1-km agl, -10C), and that
routine (maxiumum_hourly_diagnostics) is called before GFS_physics_post
<scheme>cu_gf_driver_post</scheme>
<scheme>maximum_hourly_diagnostics</scheme>
<scheme>GFS_physics_post</scheme>
…On Thu, Nov 16, 2023 at 5:04 PM Dustin Swales ***@***.***> wrote:
I think it's a great idea to pull these reflectivity calculations out of
the convection scheme, but I'm not sure if GFS_MP_generic_post.F90 is the
ideal place for them, since they rely on both the convection and mp
parameterizations. May I suggest to add them to GFS_physics_post.F90
<https://github.com/ufs-community/ccpp-physics/blob/ufs/dev/physics/GFS_physics_post.F90>?
(This was recently created as a landing spot for computations on multiple
schemes.)
—
Reply to this email directly, view it on GitHub
<#129 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ALQ75MJ4BIPBEHICSHUG7GLYE2EVTAVCNFSM6AAAAAA7LI3MWCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQMJVGM4DMOBRGE>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
|
|
@ericaligo-NOAA Good point. |
|
@haiqinli Should this same thing be done in the C3 scheme? |
|
@grantfirl Thanks for your good point. Yes, this same thing should also be done for C3. |
I'd suggest we do not include C3 in this PR since we need to get this PR committed asap to meet the deadline of the first round of RRFS code freeze. |
OK. I can add an issue to remember to do this in a later PR. |
@yangfanglin Do you know when the deadline is for the code freeze? As of now, this is scheduled for merge after Thanksgiving. Is this acceptable? If not, we can try to push is earlier. |
|
@grantfirl the target is December 1 for RRFS Beta evaluation soft freeze |
Merge branch 'ufs/dev' into feature/refconv
Merge branch 'ufs/dev' into feature/refconv
Merge branch 'ufs/dev' into feature/refconv
|
Testing is complete on ufs-wm #1996, @grantfirl can you please merge this ccpp-physics PR for us? |
[@RuiyuSun] and I compute convective reflectivity for Thompson or NSSL microphysics and with GF deep, shallow or SAS deep convection. I removed the calculation that was in cu_gf_driver_post.F90 and cu_gf_driver_post.meta, and associated parameters, and added the code to GFS_MP_generic_post.F90. The 3D refl_10cm array will be changed. In the process of creating a new baseline. I had to modify regional.nml.IN and change imfhalcnv and imfdeepconv from 2 to -1 otherwise the code would try to calculate convective reflectivity despite no calls for deep/shallow convection.
RegressionTests_hera.log
RegressionTests_hera_newbaseline.log