/[MITgcm]/MITgcm/model/src/calc_phi_hyd.F
ViewVC logotype

Contents of /MITgcm/model/src/calc_phi_hyd.F

Parent Directory Parent Directory | Revision Log Revision Log | View Revision Graph Revision Graph


Revision 1.13 - (show annotations) (download)
Mon May 14 21:51:24 2001 UTC (23 years ago) by heimbach
Branch: MAIN
CVS Tags: checkpoint40pre1, checkpoint39
Changes since 1.12: +33 -3 lines
Modifications/fixes to support TAMC differentiability
(mostly missing or wrong directives).

1 C $Header: /u/gcmpack/models/MITgcmUV/model/src/calc_phi_hyd.F,v 1.12 2001/03/25 22:33:52 heimbach Exp $
2 C $Name: checkpoint38 $
3
4 #include "CPP_OPTIONS.h"
5
6 SUBROUTINE CALC_PHI_HYD(
7 I bi, bj, iMin, iMax, jMin, jMax, K,
8 I theta, salt,
9 U phiHyd,
10 I myThid)
11 C /==========================================================\
12 C | SUBROUTINE CALC_PHI_HYD |
13 C | o Integrate the hydrostatic relation to find the Hydros. |
14 C | Potential (ocean: Pressure/rho ; atmos = geopotential)|
15 C | On entry: |
16 C | theta,salt are the current thermodynamics quantities|
17 C | (unchanged on exit) |
18 C | phiHyd(i,j,1:k-1) is the hydrostatic Potential |
19 C | at cell centers (tracer points) |
20 C | - 1:k-1 layers are valid |
21 C | - k:Nr layers are invalid |
22 C | phiHyd(i,j,k) is the hydrostatic Potential |
23 C | at cell the interface k (w point above) |
24 C | On exit: |
25 C | phiHyd(i,j,1:k) is the hydrostatic Potential |
26 C | at cell centers (tracer points) |
27 C | - 1:k layers are valid |
28 C | - k+1:Nr layers are invalid |
29 C | phiHyd(i,j,k+1) is the hydrostatic Potential (P/rho) |
30 C | at cell the interface k+1 (w point below)|
31 C | |
32 C \==========================================================/
33 IMPLICIT NONE
34 C == Global variables ==
35 #include "SIZE.h"
36 #include "GRID.h"
37 #include "EEPARAMS.h"
38 #include "PARAMS.h"
39 #ifdef ALLOW_AUTODIFF_TAMC
40 #include "tamc.h"
41 #include "tamc_keys.h"
42 #endif /* ALLOW_AUTODIFF_TAMC */
43
44 C == Routine arguments ==
45 INTEGER bi,bj,iMin,iMax,jMin,jMax,K
46 _RL theta(1-OLx:sNx+OLx,1-OLy:sNy+OLy,Nr,nSx,nSy)
47 _RL salt(1-OLx:sNx+OLx,1-OLy:sNy+OLy,Nr,nSx,nSy)
48 _RL phiHyd(1-OLx:sNx+OLx,1-OLy:sNy+OLy,Nr)
49 INTEGER myThid
50
51 #ifdef INCLUDE_PHIHYD_CALCULATION_CODE
52
53 C == Local variables ==
54 INTEGER i,j
55 _RL alphaRho(1-OLx:sNx+OLx,1-OLy:sNy+OLy)
56 _RL dRloc,dRlocKp1
57 _RL ddRm1, ddRp1, ddRm, ddRp
58 _RL atm_cp, atm_kappa, atm_po
59
60 #ifdef ALLOW_AUTODIFF_TAMC
61 act1 = bi - myBxLo(myThid)
62 max1 = myBxHi(myThid) - myBxLo(myThid) + 1
63
64 act2 = bj - myByLo(myThid)
65 max2 = myByHi(myThid) - myByLo(myThid) + 1
66
67 act3 = myThid - 1
68 max3 = nTx*nTy
69
70 act4 = ikey_dynamics - 1
71
72 ikey = (act1 + 1) + act2*max1
73 & + act3*max1*max2
74 & + act4*max1*max2*max3
75 #endif /* ALLOW_AUTODIFF_TAMC */
76
77 IF ( buoyancyRelation .eq. 'OCEANIC' ) THEN
78 C This is the hydrostatic pressure calculation for the Ocean
79 C which uses the FIND_RHO() routine to calculate density
80 C before integrating g*rho over the current layer/interface
81
82 dRloc=drC(k)
83 IF (k.EQ.1) dRloc=drF(1)
84 IF (k.EQ.Nr) THEN
85 dRlocKp1=0.
86 ELSE
87 dRlocKp1=drC(k+1)
88 ENDIF
89
90 C-- If this is the top layer we impose the boundary condition
91 C P(z=eta) = P(atmospheric_loading)
92 IF (k.EQ.1) THEN
93 DO j=jMin,jMax
94 DO i=iMin,iMax
95 C *NOTE* The loading should go here but has not been implemented yet
96 phiHyd(i,j,k)=0.
97 ENDDO
98 ENDDO
99 ENDIF
100
101 C Calculate density
102 #ifdef ALLOW_AUTODIFF_TAMC
103 kkey = (ikey-1)*Nr + k
104 CADJ STORE theta(:,:,k,bi,bj) = comlev1_bibj_k, key=kkey, byte=isbyte
105 CADJ STORE salt (:,:,k,bi,bj) = comlev1_bibj_k, key=kkey, byte=isbyte
106 #endif /* ALLOW_AUTODIFF_TAMC */
107 CALL FIND_RHO( bi, bj, iMin, iMax, jMin, jMax, k, k, eosType,
108 & theta, salt,
109 & alphaRho, myThid)
110
111 C Hydrostatic pressure at cell centers
112 DO j=jMin,jMax
113 DO i=iMin,iMax
114 #ifdef ALLOW_AUTODIFF_TAMC
115 c Patrick, is this directive correct or even necessary in
116 c this new code?
117 c Yes, because of phiHyd(i,j,k+1)=phiHyd(i,j,k)+...
118 c within the k-loop.
119 CADJ GENERAL
120 #endif /* ALLOW_AUTODIFF_TAMC */
121
122 C---------- This discretization is the "finite volume" form
123 C which has not been used to date since it does not
124 C conserve KE+PE exactly even though it is more natural
125 C
126 c IF (k.LT.Nr) phiHyd(i,j,k+1)=phiHyd(i,j,k)+
127 c & drF(K)*gravity*alphaRho(i,j)*recip_rhoConst
128 c phiHyd(i,j,k)=phiHyd(i,j,k)+
129 c & 0.5*drF(K)*gravity*alphaRho(i,j)*recip_rhoConst
130 C-----------------------------------------------------------------------
131
132 C---------- This discretization is the "energy conserving" form
133 C which has been used since at least Adcroft et al., MWR 1997
134 C
135 phiHyd(i,j,k)=phiHyd(i,j,k)+
136 & 0.5*dRloc*gravity*alphaRho(i,j)*recip_rhoConst
137 IF (k.LT.Nr) phiHyd(i,j,k+1)=phiHyd(i,j,k)+
138 & 0.5*dRlocKp1*gravity*alphaRho(i,j)*recip_rhoConst
139 C-----------------------------------------------------------------------
140 ENDDO
141 ENDDO
142
143
144
145 ELSEIF ( buoyancyRelation .eq. 'ATMOSPHERIC' ) THEN
146 C This is the hydrostatic geopotential calculation for the Atmosphere
147 C The ideal gas law is used implicitly here rather than calculating
148 C the specific volume, analogous to the oceanic case.
149
150 C Integrate d Phi / d pi
151
152 C *NOTE* These constants should be in the data file and PARAMS.h
153 atm_cp=1004. _d 0
154 atm_kappa=2. _d 0/7. _d 0
155 atm_po=1. _d 5
156 IF (K.EQ.1) THEN
157 ddRp1=atm_cp*( ((rC(K)/atm_po)**atm_kappa)
158 & -((rF(K)/atm_po)**atm_kappa) )
159 DO j=jMin,jMax
160 DO i=iMin,iMax
161 ddRp=ddRp1
162 IF (hFacC(I,J, K ,bi,bj).EQ.0.) ddRp=0.
163 C------------ The integration for the first level phi(k=1) is the
164 C same for both the "finite volume" and energy conserving
165 C methods.
166 C *NOTE* The geopotential boundary condition should go
167 C here but has not been implemented yet
168 phiHyd(i,j,K)=0.
169 & -ddRp*(theta(I,J,K,bi,bj)-tRef(K))
170 C-----------------------------------------------------------------------
171 ENDDO
172 ENDDO
173 ELSE
174
175 C-------- This discretization is the "finite volume" form which
176 C integrates the hydrostatic equation of each half/sub-layer.
177 C This seems most natural and could easily allow for lopped cells
178 C by replacing rF(K) with the height of the surface (not implemented).
179 C in the lower layers (e.g. at k=1).
180 C
181 c ddRm1=atm_cp*( ((rF( K )/atm_po)**atm_kappa)
182 c & -((rC(K-1)/atm_po)**atm_kappa) )
183 c ddRp1=atm_cp*( ((rC( K )/atm_po)**atm_kappa)
184 c & -((rF( K )/atm_po)**atm_kappa) )
185 C-----------------------------------------------------------------------
186
187
188 C-------- This discretization is the energy conserving form
189 ddRp1=atm_cp*( ((rC( K )/atm_po)**atm_kappa)
190 & -((rC(K-1)/atm_po)**atm_kappa) )*0.5
191 ddRm1=ddRp1
192 C-----------------------------------------------------------------------
193
194 DO j=jMin,jMax
195 DO i=iMin,iMax
196 ddRp=ddRp1
197 ddRm=ddRm1
198 IF (hFacC(I,J, K ,bi,bj).EQ.0.) ddRp=0.
199 IF (hFacC(I,J,K-1,bi,bj).EQ.0.) ddRm=0.
200 phiHyd(i,j,K)=phiHyd(i,j,K-1)
201 & -( ddRm*(theta(I,J,K-1,bi,bj)-tRef(K-1))
202 & +ddRp*(theta(I,J, K ,bi,bj)-tRef( K )) )
203 C Old code (atmos-exact) looked like this
204 Cold phiHyd(i,j,K)=phiHyd(i,j,K-1) - ddRm1*
205 Cold & (theta(I,J,K-1,bi,bj)+theta(I,J,K,bi,bj)-2.*tRef(K))
206 ENDDO
207 ENDDO
208 ENDIF
209
210
211 ELSE
212 STOP 'CALC_PHI_HYD: We should never reach this point!'
213 ENDIF
214
215 #endif
216
217 RETURN
218 END

  ViewVC Help
Powered by ViewVC 1.1.22