Correction to the effective layer top and bottom calculation. #358

EdwardColon-NOAA · 2021-08-06T00:56:16Z

Correction to the effective layer top and bottom calculation. The key modifications are constrained to the following subprograms

MISCLN.f
UPP_PHYSICS.f
TTBLEX.f

Code description::
(1) TTBLEX.f (line 120 ~ 240): a new routine TTBLEX1D was added (specifically used for subroutine CALCAPE1D), based on original routine TTBLEX which is CALCAPE for air parcels in 2-D plane.
(2) UPP_PHYSICS.f (line 2580 ~ 3014): adding a new routine CALCAPE1D
(3) MISCLN.f (line 3419 ~ 3749):
adding code to search for the effective layer.
scheme 1: line 3438 ~ 3491. Using CALCAPE to compute CAPE/CIN for all air parcels level by level
scheme 2: line 3492 ~ 3688. Finding air-parcel with max theta-e along the sounding profile/column on each grid box for sanity check first, then compute cape/cin of the all 1-D air parcels along that each profile if the sanity check is passed for this column/profile.
line 3689 ~ 3749: additional output to print out the profiles, positions of EL, etc for code debugging.

line:3934 ~ 3943: Array LLOW and LUPP are used to store the base and top of EL for other related variables (like helicity).

!--- IF USSING EL BASE & TOP COMPUTED BY NEW SCHEME FOR THE RELATED VARIABLES
IF ( EL_SCHEME > 0 ) THEN
!$omp parallel do private(i,j)
DO J=JSTA,JEND
DO I=1,IM
LLOW(I,J) = EL_BASE(I,J)
LUPP(I,J) = EL_TOPS(I,J)
ENDDO
ENDDO
END IF
So if set EL_SCHEME =1 or 2 in the beginning of code, then the EL found by new scheme is used, otherwise the EL of old scheme is used.

Usage:
In MISCLN.f:
line 199: debugprint = .TRUE. (if set to .FALSE., then no additional output. Those outputs are in text format and very huge, better turned off after further testing.)
line 200: EL_SCHEME = 1
any negative number , or 0--> OLD scheme;
1 --> scheme 1;
2 --> scheme 2
Note: if set EL_SCHEME =1 or 2 in the beginning of code, then the EL found by new scheme is used, otherwise the EL of old scheme is used.

…r top and bottom.

sorc/ncep_post.fd/ALLOCATE_ALL.f

sorc/ncep_post.fd/MISCLN.f

GangZhao-NOAA · 2021-08-13T01:19:42Z

sorc/ncep_post.fd/MISCLN.f

+
+        END DO ! L
+!
+      ELSE IF ( EL_SCHEME == 2 ) THEN


When I developed the code to search for Effective-Layer, I made two schemes. So the two schemes could compare to each other and verify against each other. After the testing in HWT SFE 2021, it seems that Scheme 1 had been working well and SPC was satisfied with it.
So to make the code management and maintenance simple and easy, I suggest to set Scheme 1 as the only scheme for Effective-Layer, and remove the code related to Scheme 2 (in this MISCLN.f, and in UPP_PHYSICS.f and in TTBLEX.f). Just my suggestion. These code still can be kept.

@EdwardColon-NOAA and @GangZhao-NOAA If scheme 2 is used only by 3DRTMA, I would agree with Gang that keeping scheme 1 only.

sorc/ncep_post.fd/TTBLEX.f

GangZhao-NOAA · 2021-08-13T01:24:34Z

sorc/ncep_post.fd/UPP_PHYSICS.f

+!
+!
+
+      SUBROUTINE CALCAPE1D(P_A,T_A,Q_A,ZINT_A,LPAR0,PSFC,LMASK,CAPE,CINS,    &


Same as what I comment in MISCLN.f for scheme 2 and in TTBLEX.f for subroutine TTBLEX1D,
"After the testing in HWT SFE 2021, it seems that Scheme 1 had been working well and SPC was satisfied with it. So to make the code management and maintenance simple and easy, I suggest to set Scheme 1 as the only scheme for Effective-Layer, and remove the code related to Scheme 2 (in this MISCLN.f, and in UPP_PHYSICS.f and in TTBLEX.f). Just my suggestion. These code still can be kept."
This subroutine CALCAPE1D is used only for scheme 2 in MISCLN.f for searching for the Effective-Layer. It could be removed for code clean if scheme 2 is not necessary.

GangZhao-NOAA

I have one question and one comment:

one question is about the reason of changing "im to lm" in some code. This is same question as Wen Meng already asked Ed for the reason.
One comment is about whether to keep Scheme 2 (searching for Effective-Layer) in MISCLN.f or remove Scheme 2.
As what I comment in MISCLN.f, in TTBLEX.f and in UPP_PHYSCIS.f for scheme 2,
"After the testing in HWT SFE 2021, it seems that Scheme 1 had been working well and SPC was satisfied with it. So to make the code management and maintenance simple and easy, I suggest to set Scheme 1 as the only scheme for Effective-Layer, and remove the code related to Scheme 2 (in this MISCLN.f, and in UPP_PHYSICS.f and in TTBLEX.f). Just my suggestion. These code still can be kept if you like."

The subroutine TTBLEX1D (in TTBLEX.f) , subroutine CALCAPE1D (in UPP_PHYSCIS.f) and code for scheme 2 in (MISCLN.f) are used only for scheme 2 to search for the Effective-Layer. It could be removed for code clean if scheme 2 is not necessary.

If you would like to remove these code related to Scheme 2, I can help to clean it.

Thank you!

Gang

sorc/ncep_post.fd/MISCLN.f

WenMeng-NOAA · 2021-08-13T17:30:34Z

sorc/ncep_post.fd/MISCLN.f

@@ -2910,7 +2938,8 @@ SUBROUTINE MISCLN
               EGRID1(I,J) = LOG(PMID(I,J,LM)/EGRID2(I,J))   &
                           / LOG(PMID(I,J,LM)/PMID(I,J,LM-1))

-        IF (MODELNAME == 'GFS' .OR. MODELNAME == 'FV3R') THEN
+        IF (MODELNAME == 'GFS' .OR. MODELNAME == 'FV3R' &
+           .OR. MODELNAME == 'FV3R') THEN


@EdwardColon-NOAA The model name "FV3R" is duplicated here.

This duplication was resolved.

sorc/ncep_post.fd/MISCLN.f

i loop indices corrected.

Removed EL_SCHEME flag and the EL_SCHEME 2 as it was intended as a method of validating the output of EL_SCHEME effective layer top and bottom calculations. The HWT/SFE confirmed that the EL_SCHEME 1 computations were viable making EL_SCHEME 2 superfluous. The debugprint flag was also set to false as it currently not needed in the execution of the MISCLN subprogram, however, it may be switch on at any point it becomes necessary to trace a problem.

Removed the definition of CALCAPE1D which was invoked in MISCLN.f as the CAPE/CIN computational routine used in EL_SCHEME 2. It is now a defunct option.

WenMeng-NOAA · 2021-08-17T13:11:22Z

@EdwardColon-NOAA Could you sync your branch with upstream/develop and solve the conflicts in parm/ppst_avbflds.xml so I can conduct the tests?

…/EMC_post into feature/3DRTMA8

EdwardColon-NOAA · 2021-08-17T14:24:56Z

Hi Wen, Sorry, I had some issues in setting up personal access tokens needed to push updates back to the repository after syncing my branch and resolving the conflict in parm/post_avbflds.xml.

…

-Edward

On Tue, Aug 17, 2021 at 9:11 AM WenMeng-NOAA ***@***.***> wrote: @EdwardColon-NOAA <https://github.com/EdwardColon-NOAA> Could you sync your branch with upstream/develop and solve the conflicts in parm/ppst_avbflds.xml so I can conduct the tests? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#358 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACFXXQDZJ4K7PDC33HJKHITT5JNYLANCNFSM5BU5DEBQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email> .

-- ------------------------------------------------- Edward Colón IMSG at NOAA/NWS/NCEP/EMC 5830 University Research Ct. Rm #2025 Riverdale, MD 20737 Office: (301) 683-3815 Cell: (301) 213-3566

WenMeng-NOAA · 2021-08-17T14:36:49Z

@EdwardColon-NOAA Now your branch is in a good shape. Thanks!

WenMeng-NOAA · 2021-08-17T15:05:59Z

@EdwardColon-NOAA My tests indicate that EFHL in hrrr/rap datasets have changed results like:
163:149027926:EFHL:surface:rpn_corr=0.238913:rpn_rms=37.7674

Do you think what would cause EFHL change?

EdwardColon-NOAA · 2021-08-17T16:17:06Z

Hi Wen, I will need to consult Guoqing since this commit did include FV3R updates related to RRFS integration.

-Edward

On Tue, Aug 17, 2021 at 11:06 AM WenMeng-NOAA ***@***.***> wrote: @EdwardColon-NOAA <https://github.com/EdwardColon-NOAA> My tests indicate that EFHL in hrrr/rap datasets have changed results like: 163:149027926:EFHL:surface:rpn_corr=0.238913:rpn_rms=37.7674 Do you think what would cause EFHL change? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#358 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACFXXQGOYF3WLNDGLU64HNLT5J3GFANCNFSM5BU5DEBQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email> .

-- ------------------------------------------------- Edward Colón IMSG at NOAA/NWS/NCEP/EMC 5830 University Research Ct. Rm #2025 Riverdale, MD 20737 Office: (301) 683-3815 Cell: (301) 213-3566

EdwardColon-NOAA · 2021-08-18T17:01:50Z

Hi Wen, I determined that the problem stems from the new calculation of effective layer top and bottom. The EFHL field is impacted as a result and will not yield a bit-wise identical comparison with the regression test HRRR reference file. There are a couple of ways around this. One would be to maintain the old method of calculating the effective layer top and bottom as an option linked to the selection of the HRRR model input (RAPR) while the new method would only be invoked if the submodel is specified as RTMA. The other is just to update the HRRR regression test reference file with the new calculation of the effective layer top and bottom. The later would impact current UPP stakeholders if the EFHL field is a product needed in their research. Let me know what you think.

…

-Edward On Tue, Aug 17, 2021 at 12:16 PM Edward Colon - NOAA Affiliate < ***@***.***> wrote:

Hi Wen, I will need to consult Guoqing since this commit did include FV3R updates related to RRFS integration. -Edward On Tue, Aug 17, 2021 at 11:06 AM WenMeng-NOAA ***@***.***> wrote: > @EdwardColon-NOAA <https://github.com/EdwardColon-NOAA> My tests > indicate that EFHL in hrrr/rap datasets have changed results like: > 163:149027926:EFHL:surface:rpn_corr=0.238913:rpn_rms=37.7674 > > Do you think what would cause EFHL change? > > — > You are receiving this because you were mentioned. > Reply to this email directly, view it on GitHub > <#358 (comment)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/ACFXXQGOYF3WLNDGLU64HNLT5J3GFANCNFSM5BU5DEBQ> > . > Triage notifications on the go with GitHub Mobile for iOS > <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> > or Android > <https://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email> > . > -- ------------------------------------------------- Edward Colón IMSG at NOAA/NWS/NCEP/EMC 5830 University Research Ct. Rm #2025 Riverdale, MD 20737 Office: (301) 683-3815 Cell: (301) 213-3566

-- ------------------------------------------------- Edward Colón IMSG at NOAA/NWS/NCEP/EMC 5830 University Research Ct. Rm #2025 Riverdale, MD 20737 Office: (301) 683-3815 Cell: (301) 213-3566

WenMeng-NOAA · 2021-08-18T22:43:35Z

@EdwardColon-NOAA The first way would be the one to get this PR merged quickly. We use separate algorithms for hrrr/rap and 3Drtma. For the long term solution, you might prepare the evaluation of these two algorithms and coordinate with hrrr/rap developers Geoffrey Manikin and Ming Hu for unification. Please let me know if this proposal is working for you.

EdwardColon-NOAA · 2021-08-18T22:47:47Z

Hi Wen, For the sake of timeliness, I will go ahead and code the algorithm with the option of using the older method of calculating the effective layer top and bottom. We will need to revisit the issue with Geoff and Hui-ya later.

On Wed, Aug 18, 2021 at 6:43 PM WenMeng-NOAA ***@***.***> wrote: @EdwardColon-NOAA <https://github.com/EdwardColon-NOAA> The first way would be the one to get this PR merged quickly. We use separate algorithms for hrrr/rap and 3Drtma. For the long term solution, you might prepare the evaluation of these two algorithms and coordinate with hrrr/rap developers Geoffrey Manikin and Ming Hu for unification. Please let me know if this proposal is working for you. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#358 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACFXXQAQ3P7VANB5VPNFANDT5QZSDANCNFSM5BU5DEBQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email> .

-- ------------------------------------------------- Edward Colón IMSG at NOAA/NWS/NCEP/EMC 5830 University Research Ct. Rm #2025 Riverdale, MD 20737 Office: (301) 683-3815 Cell: (301) 213-3566

WenMeng-NOAA · 2021-08-18T22:51:03Z

@EdwardColon-NOAA Sounds good to me. Thanks!

… bottom correction for RTMA.

WenMeng-NOAA · 2021-08-20T18:21:13Z

@EdwardColon-NOAA I would like to know if the scheme 2 mentioned in Gang's comments is removed.

EdwardColon-NOAA · 2021-08-20T20:58:13Z

Hi Wen, I was out of the office this afternoon. Yes, Gang's scheme 2 coding blocks were removed from this pull request. Thanks, Edward

…

On Fri, Aug 20, 2021 at 2:21 PM WenMeng-NOAA ***@***.***> wrote: @EdwardColon-NOAA <https://github.com/EdwardColon-NOAA> I would like to know if the scheme 2 mentioned in Gang's comments is removed. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#358 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACFXXQCQHABOA4EAQIYBSADT52MKJANCNFSM5BU5DEBQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email> .

-- ------------------------------------------------- Edward Colón IMSG at NOAA/NWS/NCEP/EMC 5830 University Research Ct. Rm #2025 Riverdale, MD 20737 Office: (301) 683-3815 Cell: (301) 213-3566

WenMeng-NOAA · 2021-08-20T22:34:21Z

@EdwardColon-NOAA Thanks for clarifying. I conducted the regression tests. The hrrr/rap look good. But the 3drtma test failed as:
get_postfilename,post_fname=WRFTWO.GrbF00npset= 1 num_pset=
3 iSF_SURFACE_PHYSICS= 3
forrtl: severe (174): SIGSEGV, segmentation fault occurred
Image PC Routine Line Source
upp.x 00000000014C018D Unknown Unknown Unknown
libpthread-2.17.s 00002AAB9B936630 Unknown Unknown Unknown
upp.x 0000000000A1AAD9 miscln_ 3943 MISCLN.f
upp.x 0000000000706547 process_ 116 PROCESS.f
upp.x 000000000065FA79 MAIN__ 865 WRFPOST.f
upp.x 000000000040755E Unknown Unknown Unknown
libc-2.17.so 00002AAB9C220555 __libc_start_main Unknown Unknown
upp.x 0000000000407469 Unknown Unknown Unknown
953 42 1 0

Can you take a look at it?

… to address segmentation fault occurring during the rtma regression tests

WenMeng-NOAA · 2021-09-01T17:43:58Z

@EdwardColon-NOAA You had updated your PR yesterday. Will the current version be working in rtma regression tests, or your tweaking is not done yet?

EdwardColon-NOAA · 2021-09-01T17:50:37Z

Hi Wen, After the minor tweaks, the regression test did run to completion for the rtma case. They did not yield bitwise identical results which is expected due to the modified effective layer calculation but I tested it against the default effective layer calculation and it did yield bitwise results.

…

-Edward

On Wed, Sep 1, 2021 at 1:44 PM WenMeng-NOAA ***@***.***> wrote: @EdwardColon-NOAA <https://github.com/EdwardColon-NOAA> You had updated your PR yesterday. Will the current version be working in rtma regression tests, or your tweaking is not done yet? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#358 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACFXXQGJ7OCKFHSV63RDEE3T7ZQ6VANCNFSM5BU5DEBQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.

-- ------------------------------------------------- Edward Colón IMSG at NOAA/NWS/NCEP/EMC 5830 University Research Ct. Rm #2025 Riverdale, MD 20737 Office: (301) 683-3815 Cell: (301) 213-3566

WenMeng-NOAA · 2021-09-01T18:24:40Z

@EdwardColon-NOAA Thanks for updating and clarifying. I will resume my tests.

WenMeng-NOAA · 2021-09-01T21:50:24Z

@EdwardColon-NOAA Now my rtma tests can be run through. Can you sync your branch with upstream/develop and add change logs with your name in the modified fortran code?

WenMeng-NOAA · 2021-09-02T14:39:13Z

@EdwardColon-NOAA Can you verify the new rtma GRIB2 datasets from my test at /u/Wen.Meng/ptmp/rtma_2020113000 on Mars? Thanks!

EdwardColon-NOAA · 2021-09-02T15:06:17Z

Hi Wen, I verify the new rtma GRIB2 reference files. I am getting bitwise identical results with my regression test.

…

-Edward

On Thu, Sep 2, 2021 at 10:39 AM WenMeng-NOAA ***@***.***> wrote: @EdwardColon-NOAA <https://github.com/EdwardColon-NOAA> Can you verify the new rtma GRIB2 datasets from my test at /u/Wen.Meng/ptmp/rtma_2020113000 on Mars? Thanks! — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#358 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACFXXQHCW2O5N34IKJMVQF3T76EBZANCNFSM5BU5DEBQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>.

-- ------------------------------------------------- Edward Colón IMSG at NOAA/NWS/NCEP/EMC 5830 University Research Ct. Rm #2025 Riverdale, MD 20737 Office: (301) 683-3815 Cell: (301) 213-3566

WenMeng-NOAA · 2021-09-02T15:09:56Z

@EdwardColon-NOAA Thanks! I will recreate the baseline for rtma with this PR merged.

WenMeng-NOAA · 2021-09-02T16:47:14Z

The UPP regression tests were conducted on WCOSS-Dell, WCOSS-Cray and Hera. With this PR merged, the baseline for rtma will be recreated.

EdwardColon-NOAA added 5 commits March 2, 2021 10:48

Update params_grib2_tbl_new

e5bfb05

Update post_avblflds.xml

2dad12e

Merged changes from the EMC develop branch.

7d5511d

This commit includes changes to the computation of the effective laye…

8827d9c

…r top and bottom.

This commit includes changes to the computation of the effective laye…

de51871

…r top and bottom.

EdwardColon-NOAA changed the title ~~Feature/3 drtma8~~ Correction to the effective layer top and bottom calculation. The key modifications are constrained to the following subprograms Aug 6, 2021

EdwardColon-NOAA changed the title ~~Correction to the effective layer top and bottom calculation. The key modifications are constrained to the following subprograms~~ Correction to the effective layer top and bottom calculation. Aug 6, 2021

WenMeng-NOAA reviewed Aug 11, 2021

View reviewed changes

sorc/ncep_post.fd/ALLOCATE_ALL.f Outdated Show resolved Hide resolved