/[lmdze]/trunk/dyn3d/dynetat0.f90
ViewVC logotype

Annotation of /trunk/dyn3d/dynetat0.f90

Parent Directory Parent Directory | Revision Log Revision Log

Revision 277 - (hide annotations)
Thu Jul 12 15:56:17 2018 UTC (5 years, 11 months ago) by guez
Original Path: trunk/dyn3d/dynetat0.f
File size: 25294 byte(s) 
Move fxhyp and fyhyp to module dynetat0_m to avoid side effect on
variables of module dynetat0_m. A downside is that we need to link
heavyside, coefpoly and tanh_cautious into the gcm and test_fxhyp
executables.

We must move invert_zoom_x and principal_cshift to module dynetat0_m
to avoid circular dependency.

Move definition of rlatu(1) and rlatu(jjm + 1) inside fyhyp to avoid
side effect on rlatu.
1 guez 3 module dynetat0_m
2    
3 guez 265 use dimensions, only: iim, jjm
4 guez 139
5 guez 3 IMPLICIT NONE
6    
7 guez 277 private iim, jjm, principal_cshift, invert_zoom_x, funcd
8 guez 139
9 guez 129 INTEGER day_ini
10 guez 130 ! day number at the beginning of the run, based at value 1 on
11     ! January 1st of annee_ref
12 guez 25
13 guez 139 integer:: day_ref = 1 ! jour de l'ann\'ee de l'\'etat initial
14     ! (= 350 si 20 d\'ecembre par exemple)
15 guez 129
16     integer:: annee_ref = 1998 ! Annee de l'etat initial (avec 4 chiffres)
17    
18 guez 276 REAL, protected:: clon ! longitude of the center of the zoom, in rad
19     real, protected:: clat ! latitude of the center of the zoom, in rad
20 guez 139
21 guez 276 real, protected:: grossismx, grossismy
22 guez 139 ! facteurs de grossissement du zoom, selon la longitude et la latitude
23     ! = 2 si 2 fois, = 3 si 3 fois, etc.
24    
25 guez 276 real, protected:: dzoomx, dzoomy
26 guez 139 ! extensions en longitude et latitude de la zone du zoom (fractions
27     ! de la zone totale)
28    
29 guez 276 real, protected:: taux, tauy
30 guez 151 ! raideur de la transition de l'int\'erieur \`a l'ext\'erieur du zoom
31 guez 139
32     real rlatu(jjm + 1)
33 guez 156 ! latitudes of points of the "scalar" and "u" grid, in rad
34 guez 139
35     real rlatv(jjm)
36 guez 156 ! latitudes of points of the "v" grid, in rad, in decreasing order
37 guez 139
38     real rlonu(iim + 1) ! longitudes of points of the "u" grid, in rad
39    
40     real rlonv(iim + 1)
41 guez 156 ! longitudes of points of the "scalar" and "v" grid, in rad
42 guez 139
43 guez 277 real, protected:: xprimu(iim + 1), xprimv(iim + 1)
44 guez 156 ! 2 pi / iim * (derivative of the longitudinal zoom function)(rlon[uv])
45 guez 139
46 guez 277 REAL, protected:: xprimm025(iim + 1), xprimp025(iim + 1)
47     REAL, protected:: rlatu1(jjm), rlatu2(jjm), yprimu1(jjm), yprimu2(jjm)
48 guez 259 REAL ang0, etot0, ptot0, ztot0, stot0
49 guez 277 INTEGER, PARAMETER, private:: nmax = 30000
50     DOUBLE PRECISION, private:: abs_y
51 guez 139
52     save
53    
54 guez 3 contains
55    
56 guez 128 SUBROUTINE dynetat0(vcov, ucov, teta, q, masse, ps, phis)
57 guez 3
58 guez 38 ! From dynetat0.F, version 1.2, 2004/06/22 11:45:30
59     ! Authors: P. Le Van, L. Fairhead
60 guez 39 ! This procedure reads the initial state of the atmosphere.
61 guez 3
62 guez 79 use comconst, only: dtvr
63 guez 129 use conf_gcm_m, only: raz_date
64 guez 265 use dimensions, only: iim, jjm, llm, nqmx
65 guez 67 use disvert_m, only: pa
66 guez 18 use iniadvtrac_m, only: tname
67 guez 67 use netcdf, only: NF90_NOWRITE, NF90_NOERR
68 guez 44 use netcdf95, only: NF95_GET_VAR, nf95_open, nf95_inq_varid, NF95_CLOSE, &
69     NF95_Gw_VAR
70 guez 36 use nr_util, only: assert
71 guez 129 use temps, only: itau_dyn
72     use unit_nml_m, only: unit_nml
73 guez 3
74 guez 55 REAL, intent(out):: vcov(: , :, :) ! (iim + 1, jjm, llm)
75     REAL, intent(out):: ucov(:, :, :) ! (iim + 1, jjm + 1, llm)
76 guez 43 REAL, intent(out):: teta(:, :, :) ! (iim + 1, jjm + 1, llm)
77 guez 40 REAL, intent(out):: q(:, :, :, :) ! (iim + 1, jjm + 1, llm, nqmx)
78 guez 55 REAL, intent(out):: masse(:, :, :) ! (iim + 1, jjm + 1, llm)
79 guez 39 REAL, intent(out):: ps(:, :) ! (iim + 1, jjm + 1) in Pa
80     REAL, intent(out):: phis(:, :) ! (iim + 1, jjm + 1)
81 guez 3
82 guez 44 ! Local variables:
83 guez 38 INTEGER iq
84 guez 225 REAL, allocatable:: tab_cntrl(:) ! tableau des param\`etres du run
85 guez 38 INTEGER ierr, ncid, varid
86 guez 3
87 guez 129 namelist /dynetat0_nml/ day_ref, annee_ref
88    
89 guez 3 !-----------------------------------------------------------------------
90    
91     print *, "Call sequence information: dynetat0"
92    
93 guez 55 call assert((/size(ucov, 1), size(vcov, 1), size(masse, 1), size(ps, 1), &
94     size(phis, 1), size(q, 1), size(teta, 1)/) == iim + 1, "dynetat0 iim")
95     call assert((/size(ucov, 2), size(vcov, 2) + 1, size(masse, 2), &
96     size(ps, 2), size(phis, 2), size(q, 2), size(teta, 2)/) == jjm + 1, &
97     "dynetat0 jjm")
98     call assert((/size(vcov, 3), size(ucov, 3), size(teta, 3), size(q, 3), &
99     size(masse, 3)/) == llm, "dynetat0 llm")
100 guez 40 call assert(size(q, 4) == nqmx, "dynetat0 q nqmx")
101 guez 3
102 guez 139 ! Fichier \'etat initial :
103 guez 25 call nf95_open("start.nc", NF90_NOWRITE, ncid)
104 guez 3
105 guez 38 call nf95_inq_varid(ncid, "controle", varid)
106 guez 44 call NF95_Gw_VAR(ncid, varid, tab_cntrl)
107 guez 3
108 guez 79 call assert(int(tab_cntrl(1)) == iim, "dynetat0 tab_cntrl iim")
109     call assert(int(tab_cntrl(2)) == jjm, "dynetat0 tab_cntrl jjm")
110     call assert(int(tab_cntrl(3)) == llm, "dynetat0 tab_cntrl llm")
111 guez 3
112 guez 79 IF (dtvr /= tab_cntrl(12)) THEN
113     print *, 'Warning: the time steps from day_step and "start.nc" ' // &
114 guez 129 'are different.'
115 guez 79 print *, 'dtvr from day_step: ', dtvr
116     print *, 'dtvr from "start.nc": ', tab_cntrl(12)
117     print *, 'Using the value from day_step.'
118     ENDIF
119    
120 guez 38 etot0 = tab_cntrl(13)
121     ptot0 = tab_cntrl(14)
122     ztot0 = tab_cntrl(15)
123     stot0 = tab_cntrl(16)
124     ang0 = tab_cntrl(17)
125     pa = tab_cntrl(18)
126 guez 113
127 guez 38 clon = tab_cntrl(20)
128     clat = tab_cntrl(21)
129     grossismx = tab_cntrl(22)
130     grossismy = tab_cntrl(23)
131 guez 113 dzoomx = tab_cntrl(25)
132     dzoomy = tab_cntrl(26)
133     taux = tab_cntrl(28)
134     tauy = tab_cntrl(29)
135    
136 guez 129 print *, "Enter namelist 'dynetat0_nml'."
137     read(unit=*, nml=dynetat0_nml)
138     write(unit_nml, nml=dynetat0_nml)
139 guez 3
140 guez 129 if (raz_date) then
141 guez 139 print *, 'Resetting the date, using the namelist.'
142 guez 129 day_ini = day_ref
143     itau_dyn = 0
144     else
145     day_ref = tab_cntrl(4)
146     annee_ref = tab_cntrl(5)
147     itau_dyn = tab_cntrl(31)
148     day_ini = tab_cntrl(30)
149     end if
150    
151     print *, "day_ini = ", day_ini
152    
153 guez 38 call NF95_INQ_VARID (ncid, "rlonu", varid)
154     call NF95_GET_VAR(ncid, varid, rlonu)
155 guez 3
156 guez 38 call NF95_INQ_VARID (ncid, "rlatu", varid)
157     call NF95_GET_VAR(ncid, varid, rlatu)
158 guez 3
159 guez 38 call NF95_INQ_VARID (ncid, "rlonv", varid)
160     call NF95_GET_VAR(ncid, varid, rlonv)
161 guez 3
162 guez 38 call NF95_INQ_VARID (ncid, "rlatv", varid)
163     call NF95_GET_VAR(ncid, varid, rlatv)
164 guez 3
165 guez 139 CALL nf95_inq_varid(ncid, 'xprimu', varid)
166     CALL nf95_get_var(ncid, varid, xprimu)
167 guez 3
168 guez 139 CALL nf95_inq_varid(ncid, 'xprimv', varid)
169     CALL nf95_get_var(ncid, varid, xprimv)
170 guez 3
171 guez 139 CALL nf95_inq_varid(ncid, 'xprimm025', varid)
172     CALL nf95_get_var(ncid, varid, xprimm025)
173 guez 3
174 guez 139 CALL nf95_inq_varid(ncid, 'xprimp025', varid)
175     CALL nf95_get_var(ncid, varid, xprimp025)
176    
177     call NF95_INQ_VARID (ncid, "rlatu1", varid)
178     call NF95_GET_VAR(ncid, varid, rlatu1)
179    
180     call NF95_INQ_VARID (ncid, "rlatu2", varid)
181     call NF95_GET_VAR(ncid, varid, rlatu2)
182    
183     CALL nf95_inq_varid(ncid, 'yprimu1', varid)
184     CALL nf95_get_var(ncid, varid, yprimu1)
185    
186     CALL nf95_inq_varid(ncid, 'yprimu2', varid)
187     CALL nf95_get_var(ncid, varid, yprimu2)
188    
189 guez 228 call NF95_INQ_VARID (ncid, "phis", varid)
190 guez 38 call NF95_GET_VAR(ncid, varid, phis)
191 guez 3
192 guez 38 call NF95_INQ_VARID (ncid, "ucov", varid)
193 guez 55 call NF95_GET_VAR(ncid, varid, ucov)
194 guez 3
195 guez 38 call NF95_INQ_VARID (ncid, "vcov", varid)
196 guez 55 call NF95_GET_VAR(ncid, varid, vcov)
197 guez 3
198 guez 38 call NF95_INQ_VARID (ncid, "teta", varid)
199 guez 43 call NF95_GET_VAR(ncid, varid, teta)
200 guez 3
201     DO iq = 1, nqmx
202 guez 38 call NF95_INQ_VARID(ncid, tname(iq), varid, ierr)
203 guez 157 IF (ierr == NF90_NOERR) THEN
204     call NF95_GET_VAR(ncid, varid, q(:, :, :, iq))
205     ELSE
206 guez 38 PRINT *, 'dynetat0: "' // tname(iq) // '" not found, ' // &
207     "setting it to zero..."
208 guez 40 q(:, :, :, iq) = 0.
209 guez 3 ENDIF
210     ENDDO
211    
212 guez 38 call NF95_INQ_VARID (ncid, "masse", varid)
213 guez 55 call NF95_GET_VAR(ncid, varid, masse)
214 guez 3
215 guez 38 call NF95_INQ_VARID (ncid, "ps", varid)
216 guez 39 call NF95_GET_VAR(ncid, varid, ps)
217 guez 85 ! Check that there is a single value at each pole:
218     call assert(ps(1, 1) == ps(2:, 1), "dynetat0 ps north pole")
219     call assert(ps(1, jjm + 1) == ps(2:, jjm + 1), "dynetat0 ps south pole")
220 guez 3
221 guez 25 call NF95_CLOSE(ncid)
222 guez 3
223     END SUBROUTINE dynetat0
224    
225 guez 276 !********************************************************************
226    
227     subroutine read_serre
228    
229     use unit_nml_m, only: unit_nml
230     use nr_util, only: assert, pi
231    
232     REAL:: clon_deg = 0. ! longitude of the center of the zoom, in degrees
233     real:: clat_deg = 0. ! latitude of the center of the zoom, in degrees
234    
235     namelist /serre_nml/ clon_deg, clat_deg, grossismx, grossismy, dzoomx, &
236     dzoomy, taux, tauy
237    
238     !-------------------------------------------------
239    
240     ! Default values:
241     grossismx = 1.
242     grossismy = 1.
243     dzoomx = 0.2
244     dzoomy = 0.2
245     taux = 3.
246     tauy = 3.
247    
248     print *, "Enter namelist 'serre_nml'."
249     read(unit=*, nml=serre_nml)
250     write(unit_nml, nml=serre_nml)
251    
252     call assert(grossismx >= 1. .and. grossismy >= 1., "read_serre grossism")
253     call assert(dzoomx > 0., dzoomx < 1., dzoomy < 1., &
254     "read_serre dzoomx dzoomy")
255     clon = clon_deg / 180. * pi
256     clat = clat_deg / 180. * pi
257    
258     end subroutine read_serre
259    
260 guez 277 !********************************************************************
261    
262     SUBROUTINE fyhyp
263    
264     ! From LMDZ4/libf/dyn3d/fyhyp.F, version 1.2, 2005/06/03 09:11:32
265    
266     ! Author: P. Le Van, from analysis by R. Sadourny
267    
268     ! Define rlatu, rlatv, rlatu2, yprimu2, rlatu1, yprimu1, using
269     ! clat, grossismy, dzoomy, tauy.
270    
271     ! Calcule les latitudes et dérivées dans la grille du GCM pour une
272     ! fonction f(y) à dérivée tangente hyperbolique.
273    
274     ! Il vaut mieux avoir : grossismy * dzoom < pi / 2
275    
276     use coefpoly_m, only: coefpoly, a0, a1, a2, a3
277     USE dimensions, only: jjm
278     use heavyside_m, only: heavyside
279    
280     ! Local:
281    
282     INTEGER, PARAMETER:: nmax=30000, nmax2=2*nmax
283     REAL dzoom ! distance totale de la zone du zoom (en radians)
284     DOUBLE PRECISION ylat(jjm + 1), yprim(jjm + 1)
285     DOUBLE PRECISION yuv
286     DOUBLE PRECISION, save:: yt(0:nmax2)
287     DOUBLE PRECISION fhyp(0:nmax2), beta
288     DOUBLE PRECISION, save:: ytprim(0:nmax2)
289     DOUBLE PRECISION fxm(0:nmax2)
290     DOUBLE PRECISION, save:: yf(0:nmax2)
291     DOUBLE PRECISION yypr(0:nmax2)
292     DOUBLE PRECISION yvrai(jjm + 1), yprimm(jjm + 1), ylatt(jjm + 1)
293     DOUBLE PRECISION pi, pis2, epsilon, pisjm
294     DOUBLE PRECISION yo1, yi, ylon2, ymoy, yprimin
295     DOUBLE PRECISION yfi, yf1, ffdy
296     DOUBLE PRECISION ypn
297     DOUBLE PRECISION, save::deply, y00
298    
299     INTEGER i, j, it, ik, iter, jlat, jjpn
300     INTEGER, save:: jpn
301     DOUBLE PRECISION yi2, heavyy0, heavyy0m
302     DOUBLE PRECISION fa(0:nmax2), fb(0:nmax2)
303     REAL y0min, y0max
304    
305     !-------------------------------------------------------------------
306    
307     print *, "Call sequence information: fyhyp"
308    
309     pi = 2.*asin(1.)
310     pis2 = pi/2.
311     pisjm = pi/real(jjm)
312     epsilon = 1e-3
313     dzoom = dzoomy*pi
314    
315     DO i = 0, nmax2
316     yt(i) = -pis2 + real(i)*pi/nmax2
317     END DO
318    
319     heavyy0m = heavyside(-clat)
320     heavyy0 = heavyside(clat)
321     y0min = 2.*clat*heavyy0m - pis2
322     y0max = 2.*clat*heavyy0 + pis2
323    
324     fa = 999.999
325     fb = 999.999
326    
327     DO i = 0, nmax2
328     IF (yt(i)<clat) THEN
329     fa(i) = tauy*(yt(i)-clat + dzoom/2.)
330     fb(i) = (yt(i)-2.*clat*heavyy0m + pis2)*(clat-yt(i))
331     ELSE IF (yt(i)>clat) THEN
332     fa(i) = tauy*(clat-yt(i) + dzoom/2.)
333     fb(i) = (2.*clat*heavyy0-yt(i) + pis2)*(yt(i)-clat)
334     END IF
335    
336     IF (200.*fb(i)<-fa(i)) THEN
337     fhyp(i) = -1.
338     ELSE IF (200.*fb(i)<fa(i)) THEN
339     fhyp(i) = 1.
340     ELSE
341     fhyp(i) = tanh(fa(i)/fb(i))
342     END IF
343    
344     IF (yt(i)==clat) fhyp(i) = 1.
345     IF (yt(i)==y0min .OR. yt(i)==y0max) fhyp(i) = -1.
346     END DO
347    
348     ! Calcul de beta
349    
350     ffdy = 0.
351    
352     DO i = 1, nmax2
353     ymoy = 0.5*(yt(i-1) + yt(i))
354     IF (ymoy<clat) THEN
355     fa(i) = tauy*(ymoy-clat + dzoom/2.)
356     fb(i) = (ymoy-2.*clat*heavyy0m + pis2)*(clat-ymoy)
357     ELSE IF (ymoy>clat) THEN
358     fa(i) = tauy*(clat-ymoy + dzoom/2.)
359     fb(i) = (2.*clat*heavyy0-ymoy + pis2)*(ymoy-clat)
360     END IF
361    
362     IF (200.*fb(i)<-fa(i)) THEN
363     fxm(i) = -1.
364     ELSE IF (200.*fb(i)<fa(i)) THEN
365     fxm(i) = 1.
366     ELSE
367     fxm(i) = tanh(fa(i)/fb(i))
368     END IF
369     IF (ymoy==clat) fxm(i) = 1.
370     IF (ymoy==y0min .OR. yt(i)==y0max) fxm(i) = -1.
371     ffdy = ffdy + fxm(i)*(yt(i)-yt(i-1))
372     END DO
373    
374     beta = (grossismy*ffdy-pi)/(ffdy-pi)
375    
376     IF (2. * beta - grossismy <= 0.) THEN
377     print *, 'Attention ! La valeur beta calculee dans la routine fyhyp ' &
378     // 'est mauvaise. Modifier les valeurs de grossismy, tauy ou ' &
379     // 'dzoomy et relancer.'
380     STOP 1
381     END IF
382    
383     ! calcul de Ytprim
384    
385     DO i = 0, nmax2
386     ytprim(i) = beta + (grossismy-beta)*fhyp(i)
387     END DO
388    
389     ! Calcul de Yf
390    
391     yf(0) = -pis2
392     DO i = 1, nmax2
393     yypr(i) = beta + (grossismy-beta)*fxm(i)
394     END DO
395    
396     DO i = 1, nmax2
397     yf(i) = yf(i-1) + yypr(i)*(yt(i)-yt(i-1))
398     END DO
399    
400     ! yuv = 0. si calcul des latitudes aux pts. U
401     ! yuv = 0.5 si calcul des latitudes aux pts. V
402    
403     loop_ik: DO ik = 1, 4
404     IF (ik==1) THEN
405     yuv = 0.
406     jlat = jjm + 1
407     ELSE IF (ik==2) THEN
408     yuv = 0.5
409     jlat = jjm
410     ELSE IF (ik==3) THEN
411     yuv = 0.25
412     jlat = jjm
413     ELSE IF (ik==4) THEN
414     yuv = 0.75
415     jlat = jjm
416     END IF
417    
418     yo1 = 0.
419     DO j = 1, jlat
420     yo1 = 0.
421     ylon2 = -pis2 + pisjm*(real(j) + yuv-1.)
422     yfi = ylon2
423    
424     it = nmax2
425     DO while (it >= 1 .and. yfi < yf(it))
426     it = it - 1
427     END DO
428    
429     yi = yt(it)
430     IF (it==nmax2) THEN
431     it = nmax2 - 1
432     yf(it + 1) = pis2
433     END IF
434    
435     ! Interpolation entre yi(it) et yi(it + 1) pour avoir Y(yi)
436     ! et Y'(yi)
437    
438     CALL coefpoly(yf(it), yf(it + 1), ytprim(it), ytprim(it + 1), &
439     yt(it), yt(it + 1))
440    
441     yf1 = yf(it)
442     yprimin = a1 + 2.*a2*yi + 3.*a3*yi*yi
443    
444     iter = 1
445     DO
446     yi = yi - (yf1-yfi)/yprimin
447     IF (abs(yi-yo1)<=epsilon .or. iter == 300) exit
448     yo1 = yi
449     yi2 = yi*yi
450     yf1 = a0 + a1*yi + a2*yi2 + a3*yi2*yi
451     yprimin = a1 + 2.*a2*yi + 3.*a3*yi2
452     END DO
453     if (abs(yi-yo1) > epsilon) then
454     print *, 'Pas de solution.', j, ylon2
455     STOP 1
456     end if
457    
458     yprimin = a1 + 2.*a2*yi + 3.*a3*yi*yi
459     yprim(j) = pi/(jjm*yprimin)
460     yvrai(j) = yi
461     END DO
462    
463     DO j = 1, jlat - 1
464     IF (yvrai(j + 1)<yvrai(j)) THEN
465     print *, 'Problème avec rlat(', j + 1, ') plus petit que rlat(', &
466     j, ')'
467     STOP 1
468     END IF
469     END DO
470    
471     print *, 'Reorganisation des latitudes pour avoir entre - pi/2 et pi/2'
472    
473     IF (ik==1) THEN
474     ypn = pis2
475     DO j = jjm + 1, 1, -1
476     IF (yvrai(j)<=ypn) exit
477     END DO
478    
479     jpn = j
480     y00 = yvrai(jpn)
481     deply = pis2 - y00
482     END IF
483    
484     DO j = 1, jjm + 1 - jpn
485     ylatt(j) = -pis2 - y00 + yvrai(jpn + j-1)
486     yprimm(j) = yprim(jpn + j-1)
487     END DO
488    
489     jjpn = jpn
490     IF (jlat==jjm) jjpn = jpn - 1
491    
492     DO j = 1, jjpn
493     ylatt(j + jjm + 1-jpn) = yvrai(j) + deply
494     yprimm(j + jjm + 1-jpn) = yprim(j)
495     END DO
496    
497     ! Fin de la reorganisation
498    
499     DO j = 1, jlat
500     ylat(j) = ylatt(jlat + 1-j)
501     yprim(j) = yprimm(jlat + 1-j)
502     END DO
503    
504     DO j = 1, jlat
505     yvrai(j) = ylat(j)*180./pi
506     END DO
507    
508     IF (ik==1) THEN
509     DO j = 1, jjm + 1
510     rlatu(j) = ylat(j)
511     END DO
512     ELSE IF (ik==2) THEN
513     DO j = 1, jjm
514     rlatv(j) = ylat(j)
515     END DO
516     ELSE IF (ik==3) THEN
517     DO j = 1, jjm
518     rlatu2(j) = ylat(j)
519     yprimu2(j) = yprim(j)
520     END DO
521     ELSE IF (ik==4) THEN
522     DO j = 1, jjm
523     rlatu1(j) = ylat(j)
524     yprimu1(j) = yprim(j)
525     END DO
526     END IF
527     END DO loop_ik
528    
529     DO j = 1, jjm
530     ylat(j) = rlatu(j) - rlatu(j + 1)
531     END DO
532    
533     DO j = 1, jjm
534     IF (rlatu1(j) <= rlatu2(j)) THEN
535     print *, 'Attention ! rlatu1 < rlatu2 ', rlatu1(j), rlatu2(j), j
536     STOP 13
537     ENDIF
538    
539     IF (rlatu2(j) <= rlatu(j+1)) THEN
540     print *, 'Attention ! rlatu2 < rlatup1 ', rlatu2(j), rlatu(j+1), j
541     STOP 14
542     ENDIF
543    
544     IF (rlatu(j) <= rlatu1(j)) THEN
545     print *, ' Attention ! rlatu < rlatu1 ', rlatu(j), rlatu1(j), j
546     STOP 15
547     ENDIF
548    
549     IF (rlatv(j) <= rlatu2(j)) THEN
550     print *, ' Attention ! rlatv < rlatu2 ', rlatv(j), rlatu2(j), j
551     STOP 16
552     ENDIF
553    
554     IF (rlatv(j) >= rlatu1(j)) THEN
555     print *, ' Attention ! rlatv > rlatu1 ', rlatv(j), rlatu1(j), j
556     STOP 17
557     ENDIF
558    
559     IF (rlatv(j) >= rlatu(j)) THEN
560     print *, ' Attention ! rlatv > rlatu ', rlatv(j), rlatu(j), j
561     STOP 18
562     ENDIF
563     ENDDO
564    
565     print *, 'Latitudes'
566     print 3, minval(ylat(:jjm)) *180d0/pi, maxval(ylat(:jjm))*180d0/pi
567    
568     3 Format(1x, ' Au centre du zoom, la longueur de la maille est', &
569     ' d environ ', f0.2, ' degres ', /, &
570     ' alors que la maille en dehors de la zone du zoom est ', &
571     "d'environ ", f0.2, ' degres ')
572    
573     rlatu(1) = pi / 2.
574     rlatu(jjm + 1) = -rlatu(1)
575    
576     END SUBROUTINE fyhyp
577    
578     !********************************************************************
579    
580     SUBROUTINE fxhyp
581    
582     ! From LMDZ4/libf/dyn3d/fxhyp.F, version 1.2, 2005/06/03 09:11:32
583     ! Author: P. Le Van, from formulas by R. Sadourny
584    
585     ! Compute xprimm025, rlonv, xprimv, rlonu, xprimu, xprimp025,
586     ! using clon, grossismx, dzoomx, taux.
587    
588     ! Calcule les longitudes et dérivées dans la grille du GCM pour
589     ! une fonction x_f(\tilde x) à dérivée tangente hyperbolique.
590    
591     ! Il vaut mieux avoir : grossismx \times delta < pi
592    
593     ! Le premier point scalaire pour une grille regulière (grossismx =
594     ! 1) avec clon = 0 est à - 180 degrés.
595    
596     USE dimensions, ONLY: iim
597     use nr_util, only: pi, pi_d, twopi, twopi_d, arth
598     use tanh_cautious_m, only: tanh_cautious
599    
600     ! Local:
601     real rlonm025(iim + 1), rlonp025(iim + 1), d_rlonv(iim)
602     REAL delta, h
603     DOUBLE PRECISION, dimension(0:nmax):: xtild, fhyp, G, Xf, ffdx
604     DOUBLE PRECISION beta
605     INTEGER i, is2
606     DOUBLE PRECISION xmoy(nmax), fxm(nmax)
607    
608     !----------------------------------------------------------------------
609    
610     print *, "Call sequence information: fxhyp"
611    
612     if (grossismx == 1.) then
613     h = twopi / iim
614    
615     xprimm025(:iim) = h
616     xprimp025(:iim) = h
617     xprimv(:iim) = h
618     xprimu(:iim) = h
619    
620     rlonv(:iim) = arth(- pi + clon, h, iim)
621     rlonm025(:iim) = rlonv(:iim) - 0.25 * h
622     rlonp025(:iim) = rlonv(:iim) + 0.25 * h
623     rlonu(:iim) = rlonv(:iim) + 0.5 * h
624     else
625     delta = dzoomx * twopi_d
626     xtild = arth(0d0, pi_d / nmax, nmax + 1)
627     forall (i = 1:nmax) xmoy(i) = 0.5d0 * (xtild(i-1) + xtild(i))
628    
629     ! Compute fhyp:
630     fhyp(1:nmax - 1) = tanh_cautious(taux * (delta / 2d0 &
631     - xtild(1:nmax - 1)), xtild(1:nmax - 1) &
632     * (pi_d - xtild(1:nmax - 1)))
633     fhyp(0) = 1d0
634     fhyp(nmax) = -1d0
635    
636     fxm = tanh_cautious(taux * (delta / 2d0 - xmoy), xmoy * (pi_d - xmoy))
637    
638     ! Compute \int_0 ^{\tilde x} F:
639    
640     ffdx(0) = 0d0
641    
642     DO i = 1, nmax
643     ffdx(i) = ffdx(i - 1) + fxm(i) * (xtild(i) - xtild(i-1))
644     END DO
645    
646     print *, "ffdx(nmax) = ", ffdx(nmax)
647     beta = (pi_d - grossismx * ffdx(nmax)) / (pi_d - ffdx(nmax))
648     print *, "beta = ", beta
649    
650     IF (2d0 * beta - grossismx <= 0d0) THEN
651     print *, 'Bad choice of grossismx, taux, dzoomx.'
652     print *, 'Decrease dzoomx or grossismx.'
653     STOP 1
654     END IF
655    
656     G = beta + (grossismx - beta) * fhyp
657    
658     Xf(:nmax - 1) = beta * xtild(:nmax - 1) + (grossismx - beta) &
659     * ffdx(:nmax - 1)
660     Xf(nmax) = pi_d
661    
662     call invert_zoom_x(beta, xf, xtild, G, rlonm025(:iim), xprimm025(:iim), &
663     xuv = - 0.25d0)
664     call invert_zoom_x(beta, xf, xtild, G, rlonv(:iim), xprimv(:iim), &
665     xuv = 0d0)
666     call invert_zoom_x(beta, xf, xtild, G, rlonu(:iim), xprimu(:iim), &
667     xuv = 0.5d0)
668     call invert_zoom_x(beta, xf, xtild, G, rlonp025(:iim), xprimp025(:iim), &
669     xuv = 0.25d0)
670     end if
671    
672     is2 = 0
673    
674     IF (MINval(rlonm025(:iim)) < - pi - 0.1 &
675     .or. MAXval(rlonm025(:iim)) > pi + 0.1) THEN
676     IF (clon <= 0.) THEN
677     is2 = 1
678    
679     do while (rlonm025(is2) < - pi .and. is2 < iim)
680     is2 = is2 + 1
681     end do
682    
683     if (rlonm025(is2) < - pi) then
684     print *, 'Rlonm025 plus petit que - pi !'
685     STOP 1
686     end if
687     ELSE
688     is2 = iim
689    
690     do while (rlonm025(is2) > pi .and. is2 > 1)
691     is2 = is2 - 1
692     end do
693    
694     if (rlonm025(is2) > pi) then
695     print *, 'Rlonm025 plus grand que pi !'
696     STOP 1
697     end if
698     END IF
699     END IF
700    
701     call principal_cshift(is2, rlonm025, xprimm025)
702     call principal_cshift(is2, rlonv, xprimv)
703     call principal_cshift(is2, rlonu, xprimu)
704     call principal_cshift(is2, rlonp025, xprimp025)
705    
706     forall (i = 1: iim) d_rlonv(i) = rlonv(i + 1) - rlonv(i)
707     print *, "Minimum longitude step:", MINval(d_rlonv) * 180. / pi, "degrees"
708     print *, "Maximum longitude step:", MAXval(d_rlonv) * 180. / pi, "degrees"
709    
710     ! Check that rlonm025 <= rlonv <= rlonp025 <= rlonu:
711     DO i = 1, iim + 1
712     IF (rlonp025(i) < rlonv(i)) THEN
713     print *, 'rlonp025(', i, ') = ', rlonp025(i)
714     print *, "< rlonv(", i, ") = ", rlonv(i)
715     STOP 1
716     END IF
717    
718     IF (rlonv(i) < rlonm025(i)) THEN
719     print *, 'rlonv(', i, ') = ', rlonv(i)
720     print *, "< rlonm025(", i, ") = ", rlonm025(i)
721     STOP 1
722     END IF
723    
724     IF (rlonp025(i) > rlonu(i)) THEN
725     print *, 'rlonp025(', i, ') = ', rlonp025(i)
726     print *, "> rlonu(", i, ") = ", rlonu(i)
727     STOP 1
728     END IF
729     END DO
730    
731     END SUBROUTINE fxhyp
732    
733     !********************************************************************
734    
735     subroutine principal_cshift(is2, xlon, xprimm)
736    
737     ! Add or subtract 2 pi so that xlon is near [-pi, pi], then cshift
738     ! so that xlon is in ascending order. Make the same cshift on
739     ! xprimm. Use clon.
740    
741     USE dimensions, ONLY: iim
742     use nr_util, only: twopi
743    
744     integer, intent(in):: is2
745     real, intent(inout):: xlon(:), xprimm(:) ! (iim + 1)
746    
747     !-----------------------------------------------------
748    
749     if (is2 /= 0) then
750     IF (clon <= 0.) THEN
751     IF (is2 /= 1) THEN
752     xlon(:is2 - 1) = xlon(:is2 - 1) + twopi
753     xlon(:iim) = cshift(xlon(:iim), shift = is2 - 1)
754     xprimm(:iim) = cshift(xprimm(:iim), shift = is2 - 1)
755     END IF
756     else
757     xlon(is2 + 1:iim) = xlon(is2 + 1:iim) - twopi
758     xlon(:iim) = cshift(xlon(:iim), shift = is2)
759     xprimm(:iim) = cshift(xprimm(:iim), shift = is2)
760     end IF
761     end if
762    
763     xlon(iim + 1) = xlon(1) + twopi
764     xprimm(iim + 1) = xprimm(1)
765    
766     end subroutine principal_cshift
767    
768     !**********************************************************************
769    
770     subroutine invert_zoom_x(beta, xf, xtild, G, xlon, xprim, xuv)
771    
772     ! Using clon and grossismx.
773    
774     use coefpoly_m, only: coefpoly, a1, a2, a3
775     USE dimensions, ONLY: iim
776     use nr_util, only: pi_d, twopi_d
777     use numer_rec_95, only: hunt, rtsafe
778    
779     DOUBLE PRECISION, intent(in):: beta, Xf(0:), xtild(0:), G(0:) ! (0:nmax)
780    
781     real, intent(out):: xlon(:), xprim(:) ! (iim)
782    
783     DOUBLE PRECISION, intent(in):: xuv
784     ! between - 0.25 and 0.5
785     ! 0. si calcul aux points scalaires
786     ! 0.5 si calcul aux points U
787    
788     ! Local:
789     DOUBLE PRECISION Y
790     DOUBLE PRECISION h ! step of the uniform grid
791     integer i, it
792    
793     DOUBLE PRECISION xvrai(iim), Gvrai(iim)
794     ! intermediary variables because xlon and xprim are single precision
795    
796     !------------------------------------------------------------------
797    
798     print *, "Call sequence information: invert_zoom_x"
799     it = 0 ! initial guess
800     h = twopi_d / iim
801    
802     DO i = 1, iim
803     Y = - pi_d + (i + xuv - 0.75d0) * h
804     ! - pi <= y < pi
805     abs_y = abs(y)
806    
807     ! Distinguish boundaries in order to avoid roundoff error.
808     ! funcd should be exactly equal to 0 at xtild(it) or xtild(it +
809     ! 1) and could be very small with the wrong sign so rtsafe
810     ! would fail.
811     if (abs_y == 0d0) then
812     xvrai(i) = 0d0
813     gvrai(i) = grossismx
814     else if (abs_y == pi_d) then
815     xvrai(i) = pi_d
816     gvrai(i) = 2d0 * beta - grossismx
817     else
818     call hunt(xf, abs_y, it, my_lbound = 0)
819     ! {0 <= it <= nmax - 1}
820    
821     ! Calcul de xvrai(i) et Gvrai(i)
822     CALL coefpoly(Xf(it), Xf(it + 1), G(it), G(it + 1), xtild(it), &
823     xtild(it + 1))
824     xvrai(i) = rtsafe(funcd, xtild(it), xtild(it + 1), xacc = 1d-6)
825     Gvrai(i) = a1 + xvrai(i) * (2d0 * a2 + xvrai(i) * 3d0 * a3)
826     end if
827    
828     if (y < 0d0) xvrai(i) = - xvrai(i)
829     end DO
830    
831     DO i = 1, iim -1
832     IF (xvrai(i + 1) < xvrai(i)) THEN
833     print *, 'xvrai(', i + 1, ') < xvrai(', i, ')'
834     STOP 1
835     END IF
836     END DO
837    
838     xlon = xvrai + clon
839     xprim = h / Gvrai
840    
841     end subroutine invert_zoom_x
842    
843     !**********************************************************************
844    
845     SUBROUTINE funcd(x, fval, fderiv)
846    
847     use coefpoly_m, only: a0, a1, a2, a3
848    
849     DOUBLE PRECISION, INTENT(IN):: x
850     DOUBLE PRECISION, INTENT(OUT):: fval, fderiv
851    
852     fval = a0 + x * (a1 + x * (a2 + x * a3)) - abs_y
853     fderiv = a1 + x * (2d0 * a2 + x * 3d0 * a3)
854    
855     END SUBROUTINE funcd
856    
857 guez 3 end module dynetat0_m
  ViewVC Help
Powered by ViewVC 1.1.21