1 |
C $Header: /u/gcmpack/MITgcm/pkg/thsice/thsice_solve4temp.F,v 1.1 2004/04/07 23:40:34 jmc Exp $ |
2 |
C $Name: $ |
3 |
|
4 |
#include "THSICE_OPTIONS.h" |
5 |
|
6 |
CBOP |
7 |
C !ROUTINE: THSICE_SOLVE4TEMP |
8 |
C !INTERFACE: |
9 |
SUBROUTINE THSICE_SOLVE4TEMP( |
10 |
I useBlkFlx, flxExcSw, Tf, hi, hs, |
11 |
U flxSW, Tsf, qicen, |
12 |
O Tice, sHeating, flxCnB, |
13 |
O dTsf, flxAtm, evpAtm, |
14 |
I i,j,bi,bj, myThid) |
15 |
C !DESCRIPTION: \bv |
16 |
C *==========================================================* |
17 |
C | S/R THSICE_SOLVE4TEMP |
18 |
C *==========================================================* |
19 |
C | Solve (implicitly) for sea-ice and surface temperature |
20 |
C *==========================================================* |
21 |
C \ev |
22 |
|
23 |
C !USES: |
24 |
IMPLICIT NONE |
25 |
|
26 |
C == Global variables === |
27 |
#include "THSICE_SIZE.h" |
28 |
#include "THSICE_PARAMS.h" |
29 |
|
30 |
C !INPUT/OUTPUT PARAMETERS: |
31 |
C == Routine Arguments == |
32 |
C useBlkFlx :: use surf. fluxes from bulk-forcing external S/R |
33 |
C flxExcSw :: surf. heat flux (+=down) except SW, function of surf. temp Ts: |
34 |
C 0: Flx(Ts=0) ; 1: Flx(Ts=Ts^n) ; 2: d.Flx/dTs(Ts=Ts^n) |
35 |
C Tf :: freezing temperature (oC) of local sea-water |
36 |
C hi :: ice height |
37 |
C hs :: snow height |
38 |
C flxSW :: net Short-Wave flux (+=down) [W/m2]: input= at surface |
39 |
C :: output= at the sea-ice base to the ocean |
40 |
C Tsf :: surface (ice or snow) temperature |
41 |
C qicen :: ice enthalpy (J m-3) |
42 |
C Tice :: internal ice temperatures |
43 |
C sHeating :: surf heating left to melt snow or ice (= Atmos-conduction) |
44 |
C flxCnB :: heat flux conducted through the ice to bottom surface |
45 |
C dTsf :: surf. temp adjusment: Ts^n+1 - Ts^n |
46 |
C flxAtm :: net flux of energy from the atmosphere [W/m2] (+=down) |
47 |
C without snow precip. (energy=0 for liquid water at 0.oC) |
48 |
C evpAtm :: evaporation to the atmosphere (kg/m2/s) (>0 if evaporate) |
49 |
C i,j,bi,bj :: indices of current grid point |
50 |
C myThid :: Thread no. that called this routine. |
51 |
LOGICAL useBlkFlx |
52 |
_RL flxExcSw(0:2) |
53 |
_RL Tf |
54 |
_RL hi |
55 |
_RL hs |
56 |
|
57 |
_RL flxSW |
58 |
_RL Tsf |
59 |
_RL qicen(nlyr) |
60 |
|
61 |
_RL Tice (nlyr) |
62 |
_RL sHeating |
63 |
_RL flxCnB |
64 |
_RL dTsf |
65 |
_RL flxAtm |
66 |
_RL evpAtm |
67 |
INTEGER i,j, bi,bj |
68 |
INTEGER myThid |
69 |
CEOP |
70 |
|
71 |
#ifdef ALLOW_THSICE |
72 |
|
73 |
C ADAPTED FROM: |
74 |
C LANL CICE.v2.0.2 |
75 |
C----------------------------------------------------------------------- |
76 |
C.. thermodynamics (vertical physics) based on M. Winton 3-layer model |
77 |
C.. See Bitz, C. M. and W. H. Lipscomb, 1999: "An energy-conserving |
78 |
C.. thermodynamic sea ice model for climate study." J. Geophys. |
79 |
C.. Res., 104, 15669 - 15677. |
80 |
C.. Winton, M., 1999: "A reformulated three-layer sea ice model." |
81 |
C.. Submitted to J. Atmos. Ocean. Technol. |
82 |
C.. authors Elizabeth C. Hunke and William Lipscomb |
83 |
C.. Fluid Dynamics Group, Los Alamos National Laboratory |
84 |
C----------------------------------------------------------------------- |
85 |
Cc****subroutine thermo_winton(n,fice,fsnow,dqice,dTsfc) |
86 |
C.. Compute temperature change using Winton model with 2 ice layers, of |
87 |
C.. which only the top layer has a variable heat capacity. |
88 |
|
89 |
C == Local Variables == |
90 |
INTEGER k |
91 |
|
92 |
_RL frsnow ! fractional snow cover |
93 |
|
94 |
_RL fswpen ! SW penetrating beneath surface (W m-2) |
95 |
_RL fswdn ! SW absorbed at surface (W m-2) |
96 |
_RL fswint ! SW absorbed in ice (W m-2) |
97 |
_RL fswocn ! SW passed through ice to ocean (W m-2) |
98 |
|
99 |
_RL flxExceptSw ! net surface heat flux, except short-wave (W/m2) |
100 |
C evap :: evaporation over snow/ice [kg/m2/s] (>0 if evaporate) |
101 |
C dEvdT :: derivative of evap. with respect to Tsf [kg/m2/s/K] |
102 |
_RL evap, dEvdT |
103 |
_RL flx0 ! net surf heat flux, from Atmos. to sea-ice (W m-2) |
104 |
_RL fct ! heat conducted to top surface |
105 |
|
106 |
_RL df0dT ! deriv of flx0 wrt Tsf (W m-2 deg-1) |
107 |
|
108 |
_RL k12, k32 ! thermal conductivity terms |
109 |
_RL a10, b10 ! coefficients in quadratic eqn for T1 |
110 |
_RL a1, b1, c1 ! coefficients in quadratic eqn for T1 |
111 |
c _RL Tsf_start ! old value of Tsf |
112 |
|
113 |
_RL dt ! timestep |
114 |
|
115 |
INTEGER iceornot |
116 |
LOGICAL dBug |
117 |
|
118 |
1010 FORMAT(A,I3,3F8.3) |
119 |
1020 FORMAT(A,1P4E11.3) |
120 |
|
121 |
dt = thSIce_deltaT |
122 |
dBug = .FALSE. |
123 |
c dBug = ( bi.EQ.3 .AND. i.EQ.13 .AND. j.EQ.13 ) |
124 |
c dBug = ( bi.EQ.6 .AND. i.EQ.18 .AND. j.EQ.10 ) |
125 |
IF ( dBug ) WRITE(6,'(A,2I4,2I2)') 'ThSI_THERM: i,j=',i,j,bi,bj |
126 |
|
127 |
if ( hi.lt.himin ) then |
128 |
C If hi < himin, melt the ice. |
129 |
STOP 'THSICE_THERM: should not enter if hi<himin' |
130 |
endif |
131 |
|
132 |
C---+----1----+----2----+----3----+----4----+----5----+----6----+----7-|--+----| |
133 |
|
134 |
C fractional snow cover |
135 |
frsnow = 0. _d 0 |
136 |
if (hs .gt. 0. _d 0) frsnow = 1. _d 0 |
137 |
|
138 |
C Compute SW flux absorbed at surface and penetrating to layer 1. |
139 |
fswpen = flxSW * (1. _d 0 - frsnow) * i0 |
140 |
fswocn = fswpen * exp(-ksolar*hi) |
141 |
fswint = fswpen - fswocn |
142 |
|
143 |
fswdn = flxSW - fswpen |
144 |
|
145 |
C Compute conductivity terms at layer interfaces. |
146 |
|
147 |
k12 = 4. _d 0*kice*ksnow / (ksnow*hi + 4. _d 0*kice*hs) |
148 |
k32 = 2. _d 0*kice / hi |
149 |
|
150 |
C compute ice temperatures |
151 |
a1 = cpice |
152 |
b1 = qicen(1) + (cpwater-cpice )*Tmlt1 - Lfresh |
153 |
c1 = Lfresh * Tmlt1 |
154 |
Tice(1) = 0.5 _d 0 *(-b1 - sqrt(b1*b1-4. _d 0*a1*c1))/a1 |
155 |
Tice(2) = (Lfresh-qicen(2)) / cpice |
156 |
|
157 |
if (Tice(1).gt.0. _d 0 .or. Tice(2).gt.0. _d 0) then |
158 |
write (6,*) 'BBerr Tice(1) > 0 = ',Tice(1) |
159 |
write (6,*) 'BBerr Tice(2) > 0 = ',Tice(2) |
160 |
endif |
161 |
IF ( dBug ) WRITE(6,1010) 'ThSI_THERM: k,Tice=',0,Tsf,Tice |
162 |
C---+----1----+----2----+----3----+----4----+----5----+----6----+----7-|--+----| |
163 |
|
164 |
C Compute coefficients used in quadratic formula. |
165 |
|
166 |
a10 = rhoi*cpice *hi/(2. _d 0*dt) + |
167 |
& k32 * (4. _d 0*dt*k32 + rhoi*cpice *hi) |
168 |
& / (6. _d 0*dt*k32 + rhoi*cpice *hi) |
169 |
b10 = -hi* |
170 |
& (rhoi*cpice*Tice(1)+rhoi*Lfresh*Tmlt1/Tice(1)) |
171 |
& /(2. _d 0*dt) |
172 |
& - k32 * (4. _d 0*dt*k32*Tf+rhoi*cpice *hi*Tice(2)) |
173 |
& / (6. _d 0*dt*k32 + rhoi*cpice *hi) - fswint |
174 |
c1 = rhoi*Lfresh*hi*Tmlt1 / (2. _d 0*dt) |
175 |
|
176 |
C Compute new surface and internal temperatures; iterate until |
177 |
C Tsfc converges. |
178 |
|
179 |
C ----- begin iteration ----- |
180 |
do 100 k = 1,nitMaxTsf |
181 |
|
182 |
C Save temperatures at start of iteration. |
183 |
c Tsf_start = Tsf |
184 |
|
185 |
IF ( useBlkFlx ) THEN |
186 |
C Compute top surface flux. |
187 |
if (hs.gt.3. _d -1) then |
188 |
iceornot=2 |
189 |
else |
190 |
iceornot=1 |
191 |
endif |
192 |
CALL THSICE_GET_BULKF( |
193 |
I iceornot, Tsf, |
194 |
O flxExceptSw, df0dT, evap, dEvdT, |
195 |
I i,j,bi,bj,myThid ) |
196 |
flx0 = fswdn + flxExceptSw |
197 |
ELSE |
198 |
flx0 = fswdn + flxExcSw(1) |
199 |
df0dT = flxExcSw(2) |
200 |
ENDIF |
201 |
|
202 |
C Compute new top layer and surface temperatures. |
203 |
C If Tsfc is computed to be > 0 C, fix Tsfc = 0 and recompute T1 |
204 |
C with different coefficients. |
205 |
|
206 |
a1 = a10 - k12*df0dT / (k12-df0dT) |
207 |
b1 = b10 - k12*(flx0-df0dT*Tsf) / (k12-df0dT) |
208 |
Tice(1) = -(b1 + sqrt(b1*b1-4. _d 0*a1*c1))/(2. _d 0*a1) |
209 |
dTsf = (flx0 + k12*(Tice(1)-Tsf)) / (k12-df0dT) |
210 |
Tsf = Tsf + dTsf |
211 |
if (Tsf .gt. 0. _d 0) then |
212 |
IF(dBug) WRITE(6,1010) 'ThSI_THERM: k,tice=', |
213 |
& k,Tsf,Tice(1),dTsf |
214 |
a1 = a10 + k12 |
215 |
b1 = b10 ! note b1 = b10 - k12*Tf0 |
216 |
Tice(1) = (-b1 - sqrt(b1*b1-4. _d 0*a1*c1))/(2. _d 0*a1) |
217 |
Tsf = 0. _d 0 |
218 |
IF ( useBlkFlx ) THEN |
219 |
if (hs.gt.3. _d -1) then |
220 |
iceornot=2 |
221 |
else |
222 |
iceornot=1 |
223 |
endif |
224 |
CALL THSICE_GET_BULKF( |
225 |
I iceornot, Tsf, |
226 |
O flxExceptSw, df0dT, evap, dEvdT, |
227 |
I i,j,bi,bj,myThid ) |
228 |
flx0 = fswdn + flxExceptSw |
229 |
dTsf = 0. _d 0 |
230 |
ELSE |
231 |
flx0 = fswdn + flxExcSw(0) |
232 |
dTsf = 1000. |
233 |
df0dT = 0. |
234 |
ENDIF |
235 |
Tsf = 0. _d 0 |
236 |
endif |
237 |
|
238 |
C Check for convergence. If no convergence, then repeat. |
239 |
C |
240 |
C Convergence test: Make sure Tsfc has converged, within prescribed error. |
241 |
C (Energy conservation is guaranteed within machine roundoff, even |
242 |
C if Tsfc has not converged.) |
243 |
C If no convergence, then repeat. |
244 |
|
245 |
IF ( dBug ) WRITE(6,1010) 'ThSI_THERM: k,tice=', |
246 |
& k,Tsf,Tice(1),dTsf |
247 |
IF ( useBlkFlx ) THEN |
248 |
if (abs(dTsf).lt.Terrmax) then |
249 |
goto 150 |
250 |
elseif (k.eq.nitMaxTsf) then |
251 |
write (6,*) 'BB: thermw conv err ',i,j,bi,bj,dTsf |
252 |
write (6,*) 'BB: thermw conv err, iceheight ', hi |
253 |
write (6,*) 'BB: thermw conv err: Tsf, flx0', Tsf,flx0 |
254 |
if (Tsf.lt.-70. _d 0) stop !QQQQ fails |
255 |
endif |
256 |
ELSE |
257 |
goto 150 |
258 |
ENDIF |
259 |
|
260 |
100 continue ! surface temperature iteration |
261 |
150 continue |
262 |
C ------ end iteration ------------ |
263 |
|
264 |
c if (Tsf.lt.-70. _d 0) then |
265 |
cQQ print*,'QQQQQ Tsf =',Tsf |
266 |
cQQ stop !QQQQ fails |
267 |
c endif |
268 |
|
269 |
C Compute new bottom layer temperature. |
270 |
|
271 |
Tice(2) = (2. _d 0*dt*k32*(Tice(1)+2. _d 0*Tf) |
272 |
& + rhoi*cpice *hi*Tice(2)) |
273 |
& /(6. _d 0*dt*k32 + rhoi*cpice *hi) |
274 |
IF ( dBug ) WRITE(6,1010) 'ThSI_THERM: k,Tice=',k,Tsf,Tice |
275 |
|
276 |
|
277 |
C Compute final flux values at surfaces. |
278 |
|
279 |
fct = k12*(Tsf-Tice(1)) |
280 |
flxCnB = 4. _d 0*kice *(Tice(2)-Tf)/hi |
281 |
flx0 = flx0 + df0dT*dTsf |
282 |
IF ( useBlkFlx ) THEN |
283 |
evpAtm = evap |
284 |
C-- needs to update also Evap (Tsf changes) since Latent heat has been updated |
285 |
evpAtm = evap + dEvdT*dTsf |
286 |
C- energy flux to Atmos: use net short-wave flux at surf. and |
287 |
C use latent heat = Lvap (energy=0 for liq. water at 0.oC) |
288 |
flxAtm = flxSW + flxExceptSw + df0dT*dTsf |
289 |
& + evpAtm*Lfresh |
290 |
ELSE |
291 |
flxAtm = 0. |
292 |
evpAtm = 0. |
293 |
ENDIF |
294 |
sHeating = flx0 - fct |
295 |
|
296 |
C- SW flux at sea-ice base left to the ocean |
297 |
flxSW = fswocn |
298 |
|
299 |
IF (dBug) WRITE(6,1020) 'ThSI_THERM: flx0,fct,Dif,flxCnB=', |
300 |
& flx0,fct,flx0-fct,flxCnB |
301 |
|
302 |
C Compute new enthalpy for each layer. |
303 |
|
304 |
qicen(1) = -cpwater*Tmlt1 + cpice *(Tmlt1-Tice(1)) + |
305 |
& Lfresh*(1. _d 0-Tmlt1/Tice(1)) |
306 |
qicen(2) = -cpice *Tice(2) + Lfresh |
307 |
|
308 |
C Make sure internal ice temperatures do not exceed Tmlt. |
309 |
C (This should not happen for reasonable values of i0.) |
310 |
|
311 |
if (Tice(1) .ge. Tmlt1) then |
312 |
write (6,'(A,2I4,2I3,1P2E14.6)') |
313 |
& 'BBerr - Bug: IceT(1) > Tmlt',i,j,bi,bj,Tice(1),Tmlt1 |
314 |
endif |
315 |
if (Tice(2) .ge. 0. _d 0) then |
316 |
write (6,'(A,2I4,2I3,1P2E14.6)') |
317 |
& 'BBerr - Bug: IceT(2) > 0',i,j,bi,bj,Tice(2) |
318 |
endif |
319 |
|
320 |
#endif /* ALLOW_THSICE */ |
321 |
|
322 |
C---+----1----+----2----+----3----+----4----+----5----+----6----+----7-|--+----| |
323 |
|
324 |
RETURN |
325 |
END |