| 1 |
module coefkz_m |
| 2 |
|
| 3 |
IMPLICIT none |
| 4 |
|
| 5 |
contains |
| 6 |
|
| 7 |
SUBROUTINE coefkz(nsrf, knon, paprs, pplay, ksta, ksta_ter, ts, rugos, u, v, & |
| 8 |
t, q, qsurf, coefm, coefh) |
| 9 |
|
| 10 |
! Authors: F. Hourdin, M. Forichon, Z. X. Li (LMD/CNRS) |
| 11 |
! date: 1993/09/22 |
| 12 |
! Objet : calculer le coefficient de frottement du sol ("Cdrag") et les |
| 13 |
! coefficients d'échange turbulent dans l'atmosphère. |
| 14 |
|
| 15 |
USE indicesol, ONLY: is_oce |
| 16 |
USE dimphy, ONLY: klev, klon |
| 17 |
USE conf_gcm_m, ONLY: prt_level |
| 18 |
USE suphec_m, ONLY: rcpd, rd, retv, rg, rkappa, rlstt, rlvtt, rtt |
| 19 |
USE yoethf_m, ONLY: r2es, r5ies, r5les, rvtmp2 |
| 20 |
USE fcttre, ONLY: dqsatl, dqsats, foede, foeew, qsatl, qsats, thermcep |
| 21 |
USE conf_phys_m, ONLY: iflag_pbl |
| 22 |
use clcdrag_m, only: clcdrag |
| 23 |
|
| 24 |
! Arguments: |
| 25 |
|
| 26 |
integer, intent(in):: nsrf ! indicateur de la nature du sol |
| 27 |
INTEGER, intent(in):: knon ! nombre de points a traiter |
| 28 |
|
| 29 |
REAL, intent(in):: paprs(klon, klev+1) |
| 30 |
! pression a chaque intercouche (en Pa) |
| 31 |
|
| 32 |
real, intent(in):: pplay(klon, klev) |
| 33 |
! pression au milieu de chaque couche (en Pa) |
| 34 |
|
| 35 |
REAL, intent(in):: ksta, ksta_ter |
| 36 |
REAL, intent(in):: ts(klon) ! temperature du sol (en Kelvin) |
| 37 |
REAL, intent(in):: rugos(klon) ! longeur de rugosite (en m) |
| 38 |
REAL, intent(in):: u(klon, klev), v(klon, klev) ! wind |
| 39 |
REAL, intent(in):: t(klon, klev) ! temperature (K) |
| 40 |
real, intent(in):: q(klon, klev) ! vapeur d'eau (kg/kg) |
| 41 |
real, intent(in):: qsurf(klon) |
| 42 |
REAL, intent(out):: coefm(:, :) ! (knon, klev) coefficient, vitesse |
| 43 |
|
| 44 |
real, intent(out):: coefh(:, :) ! (knon, klev) |
| 45 |
! coefficient, chaleur et humidité |
| 46 |
|
| 47 |
! Local: |
| 48 |
|
| 49 |
INTEGER itop(knon) ! numero de couche du sommet de la couche limite |
| 50 |
|
| 51 |
! Quelques constantes et options: |
| 52 |
|
| 53 |
REAL, PARAMETER:: cepdu2 =0.1**2 |
| 54 |
REAL, PARAMETER:: CKAP = 0.4 |
| 55 |
REAL, PARAMETER:: cb = 5. |
| 56 |
REAL, PARAMETER:: cc = 5. |
| 57 |
REAL, PARAMETER:: cd = 5. |
| 58 |
REAL, PARAMETER:: clam = 160. |
| 59 |
REAL, PARAMETER:: ratqs = 0.05 ! largeur de distribution de vapeur d'eau |
| 60 |
|
| 61 |
LOGICAL, PARAMETER:: richum = .TRUE. |
| 62 |
! utilise le nombre de Richardson humide |
| 63 |
|
| 64 |
REAL, PARAMETER:: ric = 0.4 ! nombre de Richardson critique |
| 65 |
REAL, PARAMETER:: prandtl = 0.4 |
| 66 |
|
| 67 |
REAL kstable ! diffusion minimale (situation stable) |
| 68 |
REAL, PARAMETER:: mixlen = 35. ! constante contrôlant longueur de mélange |
| 69 |
INTEGER, PARAMETER:: isommet = klev ! sommet de la couche limite |
| 70 |
|
| 71 |
LOGICAL, PARAMETER:: tvirtu = .TRUE. |
| 72 |
! calculer Ri d'une maniere plus performante |
| 73 |
|
| 74 |
LOGICAL, PARAMETER:: opt_ec = .FALSE. |
| 75 |
! formule du Centre Europeen dans l'atmosphere |
| 76 |
|
| 77 |
INTEGER i, k, kk |
| 78 |
REAL zgeop(klon, klev) |
| 79 |
REAL zmgeom(klon) |
| 80 |
REAL ri(klon) |
| 81 |
REAL l2(klon) |
| 82 |
|
| 83 |
REAL u1(klon), v1(klon), t1(klon), q1(klon), z1(klon) |
| 84 |
|
| 85 |
REAL zdphi, zdu2, ztvd, ztvu, cdn |
| 86 |
REAL scf |
| 87 |
REAL zt, zq, zdelta, zcvm5, zcor, zqs, zfr, zdqs |
| 88 |
REAL z2geomf, zalh2, alm2, zscfh, scfm |
| 89 |
REAL, PARAMETER:: t_coup = 273.15 |
| 90 |
REAL gamt(2:klev) ! contre-gradient pour la chaleur sensible: Kelvin/metre |
| 91 |
|
| 92 |
!-------------------------------------------------------------------- |
| 93 |
|
| 94 |
! Prescrire la valeur de contre-gradient |
| 95 |
if (iflag_pbl.eq.1) then |
| 96 |
DO k = 3, klev |
| 97 |
gamt(k) = -1.0E-03 |
| 98 |
ENDDO |
| 99 |
gamt(2) = -2.5E-03 |
| 100 |
else |
| 101 |
DO k = 2, klev |
| 102 |
gamt(k) = 0.0 |
| 103 |
ENDDO |
| 104 |
ENDIF |
| 105 |
|
| 106 |
IF ( nsrf .NE. is_oce ) THEN |
| 107 |
kstable = ksta_ter |
| 108 |
ELSE |
| 109 |
kstable = ksta |
| 110 |
ENDIF |
| 111 |
|
| 112 |
! Calculer les géopotentiels de chaque couche |
| 113 |
DO i = 1, knon |
| 114 |
zgeop(i, 1) = RD * t(i, 1) / (0.5 * (paprs(i, 1) + pplay(i, 1))) & |
| 115 |
* (paprs(i, 1) - pplay(i, 1)) |
| 116 |
ENDDO |
| 117 |
DO k = 2, klev |
| 118 |
DO i = 1, knon |
| 119 |
zgeop(i, k) = zgeop(i, k-1) & |
| 120 |
+ RD * 0.5*(t(i, k-1)+t(i, k)) / paprs(i, k) & |
| 121 |
* (pplay(i, k-1)-pplay(i, k)) |
| 122 |
ENDDO |
| 123 |
ENDDO |
| 124 |
|
| 125 |
! Calculer le frottement au sol (Cdrag) |
| 126 |
|
| 127 |
DO i = 1, knon |
| 128 |
u1(i) = u(i, 1) |
| 129 |
v1(i) = v(i, 1) |
| 130 |
t1(i) = t(i, 1) |
| 131 |
q1(i) = q(i, 1) |
| 132 |
z1(i) = zgeop(i, 1) |
| 133 |
ENDDO |
| 134 |
|
| 135 |
CALL clcdrag(klon, knon, nsrf, .false., u1, v1, t1, q1, z1, ts, qsurf, & |
| 136 |
rugos, coefm(:, 1), coefh(:, 1)) |
| 137 |
|
| 138 |
! Calculer les coefficients turbulents dans l'atmosphere |
| 139 |
|
| 140 |
itop = isommet |
| 141 |
|
| 142 |
loop_vertical: DO k = 2, isommet |
| 143 |
loop_horiz: DO i = 1, knon |
| 144 |
zdu2 = MAX(cepdu2, (u(i, k)-u(i, k-1))**2 & |
| 145 |
+(v(i, k)-v(i, k-1))**2) |
| 146 |
zmgeom(i) = zgeop(i, k)-zgeop(i, k-1) |
| 147 |
zdphi =zmgeom(i) / 2.0 |
| 148 |
zt = (t(i, k)+t(i, k-1)) * 0.5 |
| 149 |
zq = (q(i, k)+q(i, k-1)) * 0.5 |
| 150 |
|
| 151 |
! calculer Qs et dQs/dT: |
| 152 |
|
| 153 |
IF (thermcep) THEN |
| 154 |
zdelta = MAX(0., SIGN(1., RTT-zt)) |
| 155 |
zcvm5 = R5LES*RLVTT/RCPD/(1.0+RVTMP2*zq)*(1.-zdelta) & |
| 156 |
+ R5IES*RLSTT/RCPD/(1.0+RVTMP2*zq)*zdelta |
| 157 |
zqs = R2ES * FOEEW(zt, zdelta) / pplay(i, k) |
| 158 |
zqs = MIN(0.5, zqs) |
| 159 |
zcor = 1./(1.-RETV*zqs) |
| 160 |
zqs = zqs*zcor |
| 161 |
zdqs = FOEDE(zt, zdelta, zcvm5, zqs, zcor) |
| 162 |
ELSE |
| 163 |
IF (zt < t_coup) THEN |
| 164 |
zqs = qsats(zt) / pplay(i, k) |
| 165 |
zdqs = dqsats(zt, zqs) |
| 166 |
ELSE |
| 167 |
zqs = qsatl(zt) / pplay(i, k) |
| 168 |
zdqs = dqsatl(zt, zqs) |
| 169 |
ENDIF |
| 170 |
ENDIF |
| 171 |
|
| 172 |
! calculer la fraction nuageuse (processus humide): |
| 173 |
|
| 174 |
zfr = (zq+ratqs*zq-zqs) / (2.0*ratqs*zq) |
| 175 |
zfr = MAX(0.0, MIN(1.0, zfr)) |
| 176 |
IF (.NOT.richum) zfr = 0.0 |
| 177 |
|
| 178 |
! calculer le nombre de Richardson: |
| 179 |
|
| 180 |
IF (tvirtu) THEN |
| 181 |
ztvd =( t(i, k) & |
| 182 |
+ zdphi/RCPD/(1.+RVTMP2*zq) & |
| 183 |
*( (1.-zfr) + zfr*(1.+RLVTT*zqs/RD/zt)/(1.+zdqs) ) & |
| 184 |
)*(1.+RETV*q(i, k)) |
| 185 |
ztvu =( t(i, k-1) & |
| 186 |
- zdphi/RCPD/(1.+RVTMP2*zq) & |
| 187 |
*( (1.-zfr) + zfr*(1.+RLVTT*zqs/RD/zt)/(1.+zdqs) ) & |
| 188 |
)*(1.+RETV*q(i, k-1)) |
| 189 |
ri(i) =zmgeom(i)*(ztvd-ztvu)/(zdu2*0.5*(ztvd+ztvu)) |
| 190 |
ri(i) = ri(i) & |
| 191 |
+ zmgeom(i)*zmgeom(i)/RG*gamt(k) & |
| 192 |
*(paprs(i, k)/101325.0)**RKAPPA & |
| 193 |
/(zdu2*0.5*(ztvd+ztvu)) |
| 194 |
ELSE |
| 195 |
! calcul de Ridchardson compatible LMD5 |
| 196 |
ri(i) =(RCPD*(t(i, k)-t(i, k-1)) & |
| 197 |
-RD*0.5*(t(i, k)+t(i, k-1))/paprs(i, k) & |
| 198 |
*(pplay(i, k)-pplay(i, k-1)) & |
| 199 |
)*zmgeom(i)/(zdu2*0.5*RCPD*(t(i, k-1)+t(i, k))) |
| 200 |
ri(i) = ri(i) + & |
| 201 |
zmgeom(i)*zmgeom(i)*gamt(k)/RG & |
| 202 |
*(paprs(i, k)/101325.0)**RKAPPA & |
| 203 |
/(zdu2*0.5*(t(i, k-1)+t(i, k))) |
| 204 |
ENDIF |
| 205 |
|
| 206 |
! finalement, les coefficients d'echange sont obtenus: |
| 207 |
|
| 208 |
cdn = SQRT(zdu2) / zmgeom(i) * RG |
| 209 |
|
| 210 |
IF (opt_ec) THEN |
| 211 |
z2geomf = zgeop(i, k-1)+zgeop(i, k) |
| 212 |
alm2 = (0.5*ckap/RG*z2geomf & |
| 213 |
/(1.+0.5*ckap/rg/clam*z2geomf))**2 |
| 214 |
zalh2 = (0.5*ckap/rg*z2geomf & |
| 215 |
/(1.+0.5*ckap/RG/(clam*SQRT(1.5*cd))*z2geomf))**2 |
| 216 |
IF (ri(i) < 0.) THEN |
| 217 |
! situation instable |
| 218 |
scf = ((zgeop(i, k)/zgeop(i, k-1))**(1./3.)-1.)**3 & |
| 219 |
/ (zmgeom(i)/RG)**3 / (zgeop(i, k-1)/RG) |
| 220 |
scf = SQRT(-ri(i)*scf) |
| 221 |
scfm = 1.0 / (1.0+3.0*cb*cc*alm2*scf) |
| 222 |
zscfh = 1.0 / (1.0+3.0*cb*cc*zalh2*scf) |
| 223 |
coefm(i, k) = cdn * alm2 * (1. - 2. * cb * ri(i) * scfm) |
| 224 |
coefh(i, k) = cdn*zalh2*(1.-3.0*cb*ri(i)*zscfh) |
| 225 |
ELSE |
| 226 |
! situation stable |
| 227 |
scf = SQRT(1.+cd*ri(i)) |
| 228 |
coefm(i, k) = cdn * alm2 / (1. + 2. * cb * ri(i) / scf) |
| 229 |
coefh(i, k) = cdn*zalh2/(1.+3.0*cb*ri(i)*scf) |
| 230 |
ENDIF |
| 231 |
ELSE |
| 232 |
l2(i) = (mixlen*MAX(0.0, (paprs(i, k)-paprs(i, itop(i)+1)) & |
| 233 |
/(paprs(i, 2)-paprs(i, itop(i)+1)) ))**2 |
| 234 |
coefm(i, k) = sqrt(max(cdn**2 * (ric - ri(i)) / ric, kstable)) |
| 235 |
coefm(i, k)= l2(i) * coefm(i, k) |
| 236 |
coefh(i, k) = coefm(i, k) / prandtl ! h et m different |
| 237 |
ENDIF |
| 238 |
ENDDO loop_horiz |
| 239 |
ENDDO loop_vertical |
| 240 |
|
| 241 |
! Au-delà du sommet, pas de diffusion turbulente : |
| 242 |
forall (i = 1: knon) |
| 243 |
coefh(i, itop(i) + 1:) = 0. |
| 244 |
coefm(i, itop(i) + 1:) = 0. |
| 245 |
END forall |
| 246 |
|
| 247 |
END SUBROUTINE coefkz |
| 248 |
|
| 249 |
end module coefkz_m |
Parent Directory
| 
Revision Log