Restrict undefined grids from UPP computation Part 3 (#348)

* Modify the computation at defined points.

* The second part of changes from debug mode.

* The third part of changes from debug_mode.

* The fourth part of changes from debug_mode.

* The fifth part of changes from debug_mode.

* Update per comments from code reviewer.

* Remove legacy sigma level temperature from PRSLEV dataset.

* Fix MCONV computation on undefined grids.

* Fixes for global and regional inline post in debug mode.

* Update VERSION to 10.0.9.

* Remove duplicated line.

VERSION

@@ -1 +1 @@
-10.0.8
+10.0.9

parm/fv3lam.xml

@@ -1598,14 +1598,6 @@
         <scale>5.0</scale>
        </param>
 
-       <param>
-       <shortname>TMP_ON_SIGMA_LVL_HPC</shortname>
-       <pname>TMP</pname>
-       <scale_fact_fixed_sfc1>4</scale_fact_fixed_sfc1>
-       <level>9000. 8500. 8000. 7500. 7000.</level>
-        <scale>-4.0</scale>
-       </param>
-
        <param>
        <shortname>PBLREG_ON_SURFACE</shortname>
        <pname>PBLREG</pname>

parm/postxconfig-NT-fv3lam.txt

@@ -1,6 +1,6 @@
 2
 219
-259
+258
 PRSLEV
 32769
 ncep_nco
@@ -8416,43 +8416,6 @@ entire_atmos_single_lyr
 ?
 ?
 ?
-296
-TMP_ON_SIGMA_LVL_HPC
-?
-1
-tmpl4_0
-TMP
-?
-?
-sigma_lvl
-1
-4
-5
-9000. 8500. 8000. 7500. 7000.
-?
-0
-?
-0
-?
-?
-?
-0
-0.0
-0
-0.0
-?
-0
-0.0
-0
-0.0
-1
--4.0
-0
-0
-0
-?
-?
-?
 344
 PBLREG_ON_SURFACE
 ?

sorc/ncep_post.fd/ALLOCATE_ALL.f

@@ -46,7 +46,7 @@ SUBROUTINE ALLOCATE_ALL()
       integer ierr,jsx,jex
       integer i,j,l,k
 ! Allocate arrays
-      allocate(u(im,jsta_2l:jend_2u,lm))
+      allocate(u(im+1,jsta_2l:jend_2u,lm))
       allocate(v(im,jsta_2l:jvend_2u,lm))
       allocate(t(im,jsta_2l:jend_2u,lm))
 ! CHUANG ADD POTENTIAL TEMP BECAUSE WRF OUTPUT THETA
@@ -79,9 +79,23 @@ SUBROUTINE ALLOCATE_ALL()
 !$omp parallel do private(i,j,l)
       do l=1,lm
         do j=jsta_2l,jend_2u
-          do i=1,im
+          do i=1,im+1
             u(i,j,l)=0.
+          enddo
+        enddo
+      enddo
+!$omp parallel do private(i,j,l)
+      do l=1,lm
+        do j=jsta_2l,jvend_2u
+          do i=1,im
             v(i,j,l)=0.
+          enddo
+        enddo
+      enddo
+!$omp parallel do private(i,j,l)
+      do l=1,lm
+        do j=jsta_2l,jend_2u
+          do i=1,im
             t(i,j,l)=spval
             q(i,j,l)=spval
             uh(i,j,l)=spval
@@ -810,14 +824,14 @@ SUBROUTINE ALLOCATE_ALL()
           cldfra(i,j)=spval
           cprate(i,j)=spval
           cnvcfr(i,j)=spval
-          ivgtyp(i,j)=spval
-          isltyp(i,j)=spval
+          ivgtyp(i,j)=0
+          isltyp(i,j)=0
           hbotd(i,j)=spval
           htopd(i,j)=spval
           hbots(i,j)=spval
           htops(i,j)=spval
           cldefi(i,j)=spval
-          islope(i,j)=spval
+          islope(i,j)=0
           si(i,j)=spval
           lspa(i,j)=spval
           rswinc(i,j)=spval

sorc/ncep_post.fd/BNDLYR.f

@@ -31,7 +31,7 @@
 !!   00-01-04  JIM TUCCILLO - MPI VERSION 
 !!   02-01-15  MIKE BALDWIN - WRF VERSION
 !!   20-11-10  JESSE MENG   - USE UPP_PHYSICS MODULE
-!!   21-04-01  JESSE MENG   - COMPUTATION ON DEFINED POINTS ONLY
+!!   21-08-20  Wen Meng     - Retrict computation fro undefined points.
 !!     
 !!     USAGE:    CALL BNDLYR(PBND,TBND,QBND,RHBND,UBND,VBND,
 !!                            WBND,OMGBND,PWTBND,QCNVBND)
@@ -208,6 +208,18 @@ SUBROUTINE BNDLYR(PBND,TBND,QBND,RHBND,UBND,VBND,       &
                 END IF 
                 QSBND(I,J,LBND) = QSBND(I,J,LBND) + QSAT*DP
               ENDIF
+              ELSE  !undeined grids
+                PBND(I,J,LBND)=SPVAL
+                TBND(I,J,LBND)=SPVAL
+                UBND(I,J,LBND)=SPVAL
+                VBND(I,J,LBND)=SPVAL
+                WBND(I,J,LBND)=SPVAL
+                OMGBND(I,J,LBND)=SPVAL
+                QCNVBND(I,J,LBND)=SPVAL
+                PWTBND(I,J,LBND)=SPVAL
+                QBND(I,J,LBND)=SPVAL
+                QSBND(I,J,LBND)=SPVAL
+                RHBND(I,J,LBND)=SPVAL
               ENDIF
             ENDDO
           ENDDO
@@ -273,9 +285,7 @@ SUBROUTINE BNDLYR(PBND,TBND,QBND,RHBND,UBND,VBND,       &
       DO LBND=1,NBND
         DO J=JSTA,JEND
           DO I=1,IM
-            IF(PSUM(I,J,LBND)/=0.)THEN
-            IF(T(I,J,LBND)<spval.and.Q(I,J,LBND)<spval.and.&
-               UH(I,J,LBND)<spval.and.VH(I,J,LBND)<spval) THEN
+            IF(PSUM(I,J,LBND)/=0..AND.TBND(I,J,LBND)<SPVAL)THEN
               RPSUM           = 1./PSUM(I,J,LBND)
               LVLBND(I,J,LBND)= LVLBND(I,J,LBND)/NSUM(I,J,LBND)
               PBND(I,J,LBND)  = (PBINT(I,J,LBND)+PBINT(I,J,LBND+1))*0.5
@@ -288,18 +298,8 @@ SUBROUTINE BNDLYR(PBND,TBND,QBND,RHBND,UBND,VBND,       &
                VBND(I,J,LBND)  = VBND(I,J,LBND)*RPSUM
               END IF 
               WBND(I,J,LBND)     = WBND(I,J,LBND)*RPSUM
+              IF(QCNVBND(I,J,LBND)<SPVAL) &
               QCNVBND(I,J,LBND)  = QCNVBND(I,J,LBND)*RPSUM
-            ELSE
-              LVLBND(I,J,LBND)= spval
-              PBND(I,J,LBND)  = spval
-              TBND(I,J,LBND)  = spval
-              QBND(I,J,LBND)  = spval
-              OMGBND(I,J,LBND)= spval
-              UBND(I,J,LBND)  = spval
-              VBND(I,J,LBND)  = spval
-              WBND(I,J,LBND)  = spval
-              QCNVBND(I,J,LBND)= spval
-            ENDIF
             ENDIF
           ENDDO
         ENDDO
@@ -308,14 +308,9 @@ SUBROUTINE BNDLYR(PBND,TBND,QBND,RHBND,UBND,VBND,       &
           DO J=JSTA_M,JEND_M
             DO I=2,IM-1
               IF(PVSUM(I,J,LBND)/=0.)THEN
-              IF(UBND(I,J,LBND)<spval.and.VBND(I,J,LBND)<spval.and.PVSUM(I,J,LBND)<spval)THEN
                 RPVSUM         = 1./PVSUM(I,J,LBND)
                 UBND(I,J,LBND) = UBND(I,J,LBND)*RPVSUM
                 VBND(I,J,LBND) = VBND(I,J,LBND)*RPVSUM
-              ELSE
-                UBND(I,J,LBND) = spval
-                VBND(I,J,LBND) = spval
-              ENDIF
               ENDIF
             ENDDO
           ENDDO
@@ -331,7 +326,7 @@ SUBROUTINE BNDLYR(PBND,TBND,QBND,RHBND,UBND,VBND,       &
       DO LBND=1,NBND
         DO J=JSTA,JEND
           DO I=1,IM
-            IF(PSUM(I,J,LBND)==0.)THEN
+            IF(PSUM(I,J,LBND)==0..AND.PBND(I,J,LBND)<SPVAL)THEN
               L    = LM
               PMIN = 9999999.
               PBND(I,J,LBND) = (PBINT(I,J,LBND)+PBINT(I,J,LBND+1))*0.5
@@ -355,7 +350,6 @@ SUBROUTINE BNDLYR(PBND,TBND,QBND,RHBND,UBND,VBND,       &
                 VBND(I,J,LBND) = VH(I,J,L)
               END IF 
               WBND(I,J,LBND)    = WH(I,J,L)
-            IF(T(I,J,LBND)<spval.and.Q(I,J,LBND)<spval)THEN
               QCNVBND(I,J,LBND) = QCNVG(I,J,L)
               IF(MODELNAME == 'GFS' .OR. MODELNAME == 'FV3R')THEN
                 ES   = FPVSNEW(T(I,J,L))
@@ -367,17 +361,11 @@ SUBROUTINE BNDLYR(PBND,TBND,QBND,RHBND,UBND,VBND,       &
               QSBND(I,J,LBND)  = QSAT
               OMGBND(I,J,LBND) = OMGA(I,J,L)
               PWTBND(I,J,LBND) = (Q(I,J,L)+CWM(I,J,L))*DP*GI
-            ELSE
-              QCNVBND(I,J,LBND)= spval
-              QSBND(I,J,LBND)  = spval
-              OMGBND(I,J,LBND) = spval
-              PWTBND(I,J,LBND) = spval
-            ENDIF
             ENDIF
 !
 !   RH, BOUNDS CHECK
 !
-          IF(T(I,J,LBND)<spval.and.Q(I,J,LBND)<spval)THEN            
+            IF(QSBND(I,J,LBND)/=0..AND.QBND(I,J,LBND)<SPVAL)THEN
             RHBND(I,J,LBND) = QBND(I,J,LBND)/QSBND(I,J,LBND)
             IF (RHBND(I,J,LBND)>1.0) THEN
               RHBND(I,J,LBND) = 1.0
@@ -387,10 +375,7 @@ SUBROUTINE BNDLYR(PBND,TBND,QBND,RHBND,UBND,VBND,       &
               RHBND(I,J,LBND) = 0.01
               QBND(I,J,LBND)  = RHBND(I,J,LBND)*QSBND(I,J,LBND)
             ENDIF
-          ELSE
-              RHBND(I,J,LBND) = spval
-              QBND(I,J,LBND)  = spval
-          ENDIF
+            ENDIF
           ENDDO
         ENDDO
 !

sorc/ncep_post.fd/CALDWP.f

@@ -12,6 +12,7 @@
 !!                            AMBIENT TEMPERATURE.
 !!   98-06-08  T BLACK      - CONVERSION FROM 1-D TO 2-D
 !!   00-01-04  JIM TUCCILLO - MPI VERSION                
+!!   21-07-23  Wen Meng     - Retrict computation from undefined points
 !!     
 !! USAGE:    CALL CALDWP(P1D,Q1D,TDWP,T1D)
 !!   INPUT ARGUMENT LIST:
@@ -42,7 +43,7 @@ SUBROUTINE CALDWP(P1D,Q1D,TDWP,T1D)
 !
 !     SET PARAMETERS.
      use params_mod, only: eps, oneps, d001, h1m12
-     use ctlblk_mod, only: jsta, jend, im
+     use ctlblk_mod, only: jsta, jend, im, spval
 !- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
       implicit none
 !     
@@ -62,8 +63,12 @@ SUBROUTINE CALDWP(P1D,Q1D,TDWP,T1D)
 !$omp parallel do private(i,j)
       DO J=JSTA,JEND
         DO I=1,IM
+          IF(P1D(I,j)<spval .and. Q1D(I,J)<spval) THEN
           EVP(I,J) = P1D(I,J)*Q1D(I,J)/(EPS+ONEPS*Q1D(I,J))
           EVP(I,J) = MAX(H1M12,EVP(I,J)*D001)
+          ELSE
+          EVP(I,J) = spval
+          ENDIF
         ENDDO
       ENDDO
 !     

sorc/ncep_post.fd/CALLCL.f

@@ -19,6 +19,7 @@
 !!   00-01-04  JIM TUCCILLO - MPI VERSION            
 !!   02-04-24  MIKE BALDWIN - WRF VERSION            
 !!   19-10-30  Bo CUI - REMOVE "GOTO" STATEMENT
+!!   21-07-28  W Meng - Restriction compuatation from undefined grids
 !!     
 !! USAGE:    CALL CALLCL(P1D,T1D,Q1D,PLCL,ZLCL)
 !!   INPUT ARGUMENT LIST:
@@ -88,6 +89,7 @@ SUBROUTINE CALLCL(P1D,T1D,Q1D,PLCL,ZLCL)
       DO 30 J=JSTA_M,JEND_M
       DO 30 I=2,IM-1
 !     DO 30 I=1,IM
+      IF(P1D(I,J)<spval.and.Q1D(I,J)<spval)THEN
       EVP       = P1D(I,J)*Q1D(I,J)/(EPS+ONEPS*Q1D(I,J))
       RMX       = EPS*EVP/(P1D(I,J)-EVP)
       RKAPA     = 1.0 / (D2845*(1.0-D28*RMX))
@@ -107,6 +109,7 @@ SUBROUTINE CALLCL(P1D,T1D,Q1D,PLCL,ZLCL)
         EXIT
       ENDIF
  20   CONTINUE
+      ENDIF
  30   CONTINUE
 !     
 !     END OF ROUTINE.

sorc/ncep_post.fd/CALMCVG.f

@@ -81,7 +81,11 @@ SUBROUTINE CALMCVG(Q1D,U1D,V1D,QCNVG)
 !$omp  parallel do private(i,j)
       DO J=JSTA_2L,JEND_2U
         DO I=1,IM
+          IF(U1D(I,J)<SPVAL)THEN
           QCNVG(I,J) = 0.
+          ELSE
+          QCNVG(I,J) = SPVAL
+          ENDIF
           UWND(I,J)  = U1D(I,J)
           VWND(I,J)  = V1D(I,J)
           IF (UWND(I,J) == SPVAL) UWND(I,J) = D00
@@ -97,7 +101,8 @@ SUBROUTINE CALMCVG(Q1D,U1D,V1D,QCNVG)
        DO J=JSTA_M,JEND_M
          DO I=2,IM-1
            IF(Q1D(I,J+1)<SPVAL.AND.Q1D(I,J-1)<SPVAL.AND.          &
-              Q1D(I+1,J)<SPVAL.AND.Q1D(I-1,J)<SPVAL) THEN
+              Q1D(I+1,J)<SPVAL.AND.Q1D(I-1,J)<SPVAL.AND. &
+              Q1D(I,J)<SPVAL) THEN
              R2DX   = 1./(2.*DX(I,J))   !MEB DX?
              R2DY   = 1./(2.*DY(I,J))   !MEB DY?  
              QUDX   = (Q1D(I+1,J)*UWND(I+1,J)-Q1D(I-1,J)*UWND(I-1,J))*R2DX
@@ -141,7 +146,8 @@ SUBROUTINE CALMCVG(Q1D,U1D,V1D,QCNVG)
          IEND = IM-1-MOD(J,2)
          DO I=2,IEND
           IF(QV(I+IHE(J),J)<SPVAL.AND.UWND(I+IHE(J),J)<SPVAL.AND.&
-             QV(I+IHW(J),J)<SPVAL.AND.UWND(I+IHW(J),J)<SPVAL) THEN
+             QV(I+IHW(J),J)<SPVAL.AND.UWND(I+IHW(J),J)<SPVAL.AND.&
+             QV(I,J)<SPVAL.AND.QV(I,J-1)<SPVAL.AND.QV(I,J+1)<SPVAL) THEN
            R2DX   = 1./(2.*DX(I,J))
            R2DY   = 1./(2.*DY(I,J))
            QUDX   = (QV(I+IHE(J),J)*UWND(I+IHE(J),J)                   &

sorc/ncep_post.fd/CALTAU.f

@@ -17,6 +17,7 @@
 !!   02-01-15  MIKE BALDWIN - WRF VERSION, OUTPUT IS ON MASS-POINTS
 !!   05-02-23  H CHUANG - COMPUTE STRESS FOR NMM ON WIND POINTS
 !!   05-07-07  BINBIN ZHOU - ADD RSM STRESS for A GRID     
+!!   21-07-26  W Meng - Restrict computation from undefined grids
 !! USAGE:    CALL CALTAU(TAUX,TAUY)
 !!   INPUT ARGUMENT LIST:
 !!     NONE     
@@ -99,6 +100,8 @@ SUBROUTINE CALTAU(TAUX,TAUY)
        DO I=1,IM
 !
         LMHK = NINT(LMH(I,J))
+        IF(EL(I,J,LMHK-1)<spval.and.Z0(I,J)<spval.and. &
+           UZ0(I,J)<spval.and.VZ0(I,J)<spval)THEN
 !
 !       COMPUTE THICKNESS OF LAYER AT MASS POINT.
 !
@@ -128,6 +131,10 @@ SUBROUTINE CALTAU(TAUX,TAUY)
         ELSQR     = EL(I,J,LMHK-1)*EL(I,J,LMHK-1)
         TAUX(I,J) = RHO*ELSQR*DELUDZ*DELUDZ
         TAUY(I,J) = RHO*ELSQR*DELVDZ*DELVDZ
+        ELSE
+        TAUX(I,J) = spval
+        TAUY(I,J) = spval
+        ENDIF
 
 !
        END DO

sorc/ncep_post.fd/CALTHTE.f

@@ -12,6 +12,7 @@
 !!   93-06-18  RUSS TREADON
 !!   98-06-16  T BLACK - CONVERSION FROM 1-D TO 2-D
 !!   00-01-04  JIM TUCCILLO - MPI VERSION
+!!   21-07-28  W Meng - Restrict computation from undefined grids
 !!     
 !! USAGE:    CALL CALTHTE(P1D,T1D,Q1D,THTE)
 !!   INPUT ARGUMENT LIST:
@@ -40,7 +41,7 @@ SUBROUTINE CALTHTE(P1D,T1D,Q1D,THTE)
 !
 !     
       use params_mod, only: d00, eps, oneps, d01, h1m12, p1000, h1
-      use ctlblk_mod, only: jsta, jend, im
+      use ctlblk_mod, only: jsta, jend, im, spval
 !- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
       implicit none
 !
@@ -77,6 +78,7 @@ SUBROUTINE CALTHTE(P1D,T1D,Q1D,THTE)
 !$omp parallel do private(i,j,p,t,q,evp,rmx,ckapa,rkapa,arg,denom,tlcl,plcl,fac,eterm,thetae)
       DO J=JSTA,JEND
         DO I=1,IM
+          IF(P1D(I,J)<spval.and.T1D(I,J)<spval.and.Q1D(I,J)<spval)THEN
           P        = P1D(I,J)
           T        = T1D(I,J)
           Q        = Q1D(I,J)
@@ -92,6 +94,7 @@ SUBROUTINE CALTHTE(P1D,T1D,Q1D,THTE)
           ETERM    = (D3376/TLCL-D00254)*(RMX*KG2G*(H1+D81*RMX))
           THETAE   = T*FAC*EXP(ETERM)
           THTE(I,J)= THETAE
+          ENDIF
         ENDDO
       ENDDO
 !