-
Notifications
You must be signed in to change notification settings - Fork 36
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
incorporate csawMG update and pass both intel and GNU RT tests #181
Conversation
@dustinswales @Qingfu-Liu Please review this again when you get a minute. It should be easier to follow now. |
@grant Firl ***@***.***> please add back.
…On Thu, Mar 7, 2024 at 2:35 PM Grant Firl ***@***.***> wrote:
***@***.**** commented on this pull request.
------------------------------
In physics/CONV/Chikira_Sugiyama/cs_conv.F90
<#181 (comment)>
:
> ISTS, IENS ) ! input
!
IMPLICIT NONE
- INTEGER, INTENT(IN) :: IM, IJSDIM, KMAX, NTR !! DD, for GFS, pass in
+ INTEGER, INTENT(IN) :: IM, IJSDIM, KMAX, NTR, nctp !! DD, for GFS, pass in
!
! [MODIFY]
REAL(kind_phys) GTR (IJSDIM, KMAX, NTR) !! tracer tendency
@AnningCheng-NOAA <https://github.com/AnningCheng-NOAA> The code in #171
<#171> wouldn't compile
without adding the NTR dimension back to GTR since the third dimension of
GTR is used in this subroutine. Please double-check that this is OK. This
is the only *CODE* change in #171
<#171> that was
reverted.
—
Reply to this email directly, view it on GitHub
<#181 (review)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ALQPMILHVKKX3X4TAPNCMILYXC6RPAVCNFSM6AAAAABELTBAW2VHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMYTSMRTGM3DKMZRHE>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
|
Either way is fine with me. Those variables are for MG usage and have been
tuned for best performance. It seems like w_upi=0 and not using cf_upi had
been chosen although I tend to keep the code by uncommenting cf_upi in
case for future reference.
…On Thu, Mar 7, 2024 at 2:39 PM Grant Firl ***@***.***> wrote:
***@***.**** commented on this pull request.
------------------------------
In physics/CONV/Chikira_Sugiyama/cs_conv.F90
<#181 (comment)>
:
> @@ -458,8 +466,8 @@ subroutine cs_conv_run( IJSDIM , KMAX , ntracp1 , NN, &
! CNV_PRC3(i,k) = 0.0
CNV_NDROP(i,k) = 0.0
CNV_NICE(i,k) = 0.0
- cf_upi(i,k) = max(0.0, min(1.0, 0.5*(sigma(i,k)+sigma(i,kp1))))
- CLCN(i,k) = cf_upi(i,k) !downdraft is below updraft
+! cf_upi(i,k) = max(0.0,min(0.01*log(1.0+500*ud_mf(i,k)),0.1))
One of the CCPP rules is that all intent(OUT) variables are given values.
These comments prevent cf_upi from being given a value. If cf_upi is no
longer needed, we can remove it from the argument list and CCPP metadata,
although see the comment below about the calculation of w_upi when do_aw
= F.
—
Reply to this email directly, view it on GitHub
<#181 (review)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ALQPMIKZM5JHHD2F5AA6IK3YXC67RAVCNFSM6AAAAABELTBAW2VHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMYTSMRTGM3TEOJRGQ>
.
You are receiving this because you were mentioned.Message ID:
***@***.***>
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The code is fine with me now.
! gciz(i,k) = tem1 * gciz(i,k) | ||
! enddo | ||
! endif | ||
! enddo |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Lines 1245-1277 can be removed
! END DO | ||
! print *,'precip1',pr_tot*86400.,gprcp(1,1)*86400.,gsnwp(1,1)*86400. | ||
! print *,'precip2',pr_tot*86400.,pr_liq*86400.,pr_ice*86400. | ||
! endif |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
lines 1779-1796 can be removed
! dtrdwn(i,k,n) = zero | ||
! enddo | ||
! enddo | ||
! enddo |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
lines 3315-3330 can be removed
! gtt(i,k) = gtt(i,k) + fsigma(i,k)*(dtmelt(i,k) + dtevap(i,k)) + condtermt(i,k) | ||
! gtq(i,k,1) = gtq(i,k,1) + fsigma(i,k)*dqevap(i,k) + condtermq(i,k) | ||
! gtq(i,k,itl) = gtq(i,k,itl) - (condtermq(i,k) + prectermq(i,k) + tem) | ||
! gtq(i,k,iti) = gtq(i,k,iti) + tem |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
lines 1586-1589 can be removed
! | ||
! sigmai(i,k,ctp) = lamdai(i,k,ctp) / lamdaprod(i,k) | ||
! sigma(i,k) = max(zero, min(one, sigma(i,k) + sigmai(i,k,ctp))) | ||
! vverti(i,k,ctp) = sigmai(i,k,ctp) * wcv(i,k,ctp) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
lines 1132-1142 can be removed.
! do i=ISTS,IENS | ||
! GPRCC( I,1 ) = GPRCP( I,1 ) | ||
! GSNWC( I ) = GSNWP( I,1 ) | ||
! enddo |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
lines 1806-1809 can be removed
@Qingfu-Liu @dustinswales @mzhangw regression testing is complete on UFS-WM PR #2180 , and ready for final review. |
@Qingfu-Liu Please let me know if you feel strongly about removing the commented lines in this PR and I'll do that before merge. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
These changes look great to me. Thanks for all the good work!
Ok, I think we have enough approvals to merge at this time. @grantfirl |
This replaces #171. This version started from the top of ufs/dev and only added necessary changes. I've verified that this gives bit-for-bit answers compared to #171 in the SCM.
Description from @AnningCheng-NOAA:
Major updated CSAW deep convective scheme: considering sigma vertical veriantion and reduction on fluxes instead of tendency etc.
• Regression tests are performed on the new version of CSAW for GNU and intel compilers
Here is a EMail from Don describe the changes:
On Thu, Oct 24, 2019 at 3:01 PM Donald Dazlich - NOAA Affiliate <donald.dazlich@noaa.gov> wrote:
Hi Moorthi, Anning,
Dave and I have been using the GFS version of CS to redo our implementation of AW more completely and consistently. Now, you will find that the fix_form implemetation of the tendencies match up very closely with the traditional form. Also you will notice that it is the fluxes for each cloud type (and the downdraft) that are saved and the reduction is done on the fluxes, not the tendency computed from the flux divergence. Since sigma is a function of height the latter approach is incorrect. I’ve also dimensioned several cloud type variables nctp+1 - the nctp+1 slot has downdraft output. Another addition is the sorting of the lamdas after they are computed for all cloud types, then computing the sigma. This approach is detailed in a manuscript Dave is working on. Also, sigma is not applied to the inputs to the downdraft routine, but to the outputs, which are now mass fluxes not tendencies. The water and thermodynamic budgets remain satisfied now.
I’ve attached my version below. It should be pretty close to your version with regards to the interface so you shouldn’t have to do much work with it.
Regards,
Don