| 1 |
module HBTM_m |
| 2 |
|
| 3 |
IMPLICIT none |
| 4 |
|
| 5 |
contains |
| 6 |
|
| 7 |
SUBROUTINE HBTM(paprs, pplay, t2m, q2m, ustar, flux_t, flux_q, u, v, t, q, & |
| 8 |
pblh, cape, EauLiq, ctei, pblT, therm, trmb1, trmb2, trmb3, plcl) |
| 9 |
|
| 10 |
! D'apr\'es Holstag et Boville et Troen et Mahrt |
| 11 |
! JAS 47 BLM |
| 12 |
|
| 13 |
! Algorithme th\'ese Anne Mathieu. Crit\'ere d'entra\^inement |
| 14 |
! Peter Duynkerke (JAS 50). Written by: Anne MATHIEU and Alain |
| 15 |
! LAHELLEC, 22nd November 1999. |
| 16 |
|
| 17 |
! Modifications : d\'ecembre 99 passage th \`a niveau plus bas. Voir fixer |
| 18 |
! la prise du th \`a z/Lambda = -.2 (max Ray) |
| 19 |
! Autre algorithme : entra\^inement ~ Theta + v =constante |
| 20 |
! mais comment ? The ? |
| 21 |
! On peut fixer q \`a 0.7 qsat (cf. non adiabatique) d'où T2 et The2. |
| 22 |
! Voir aussi KE pblh = niveau The_e ou l = env. |
| 23 |
|
| 24 |
! Adaptation \`a LMDZ version coupl\'ee. Pour le moment on fait |
| 25 |
! passer en argument les grandeurs de surface : flux, t, q2m. On |
| 26 |
! va utiliser syst\'ematiquement les grandeurs \`a 2 m mais on |
| 27 |
! garde la possibilit\'e de changer si besoin (jusqu'\`a pr\'esent |
| 28 |
! la forme de HB avec le premier niveau mod\`ele \'etait |
| 29 |
! conserv\'ee). |
| 30 |
|
| 31 |
USE dimphy, ONLY: klev, klon |
| 32 |
USE suphec_m, ONLY: rcpd, rd, retv, rg, rkappa, rtt |
| 33 |
USE yoethf_m, ONLY: r2es, rvtmp2 |
| 34 |
USE fcttre, ONLY: foeew |
| 35 |
|
| 36 |
! Arguments: |
| 37 |
|
| 38 |
! pression a inter-couche (Pa) |
| 39 |
REAL, intent(in):: paprs(klon, klev+1) |
| 40 |
! pression au milieu de couche (Pa) |
| 41 |
REAL, intent(in):: pplay(klon, klev) |
| 42 |
REAL, intent(in):: t2m(klon) ! temperature a 2 m |
| 43 |
! q a 2 et 10m |
| 44 |
REAL, intent(in):: q2m(klon) |
| 45 |
REAL, intent(in):: ustar(klon) |
| 46 |
REAL, intent(in):: flux_t(:), flux_q(:) ! (knon) flux à la surface |
| 47 |
|
| 48 |
! vitesse U (m/s) |
| 49 |
REAL, intent(in):: u(klon, klev) |
| 50 |
! vitesse V (m/s) |
| 51 |
REAL, intent(in):: v(klon, klev) |
| 52 |
! temperature (K) |
| 53 |
REAL, intent(in):: t(klon, klev) |
| 54 |
! vapeur d'eau (kg/kg) |
| 55 |
REAL, intent(in):: q(klon, klev) |
| 56 |
|
| 57 |
REAL, intent(out):: pblh(:) ! (knon) |
| 58 |
! Cape du thermique |
| 59 |
REAL Cape(klon) |
| 60 |
! Eau liqu integr du thermique |
| 61 |
REAL EauLiq(klon) |
| 62 |
! Critere d'instab d'entrainmt des nuages de |
| 63 |
REAL ctei(klon) |
| 64 |
REAL pblT(klon) |
| 65 |
! thermal virtual temperature excess |
| 66 |
REAL therm(klon) |
| 67 |
REAL trmb1(klon), trmb2(klon), trmb3(klon) |
| 68 |
REAL plcl(klon) |
| 69 |
|
| 70 |
! Local: |
| 71 |
|
| 72 |
INTEGER knon ! nombre de points a calculer |
| 73 |
INTEGER isommet |
| 74 |
! limite max sommet pbl |
| 75 |
PARAMETER (isommet=klev) |
| 76 |
REAL vk |
| 77 |
! Von Karman => passer a .41 ! cf U.Olgstrom |
| 78 |
PARAMETER (vk=0.35) |
| 79 |
REAL ricr |
| 80 |
PARAMETER (ricr=0.4) |
| 81 |
! a |
| 82 |
REAL onet |
| 83 |
PARAMETER (onet=1.0/3.0) |
| 84 |
REAL zkmin |
| 85 |
PARAMETER (zkmin=0.01) |
| 86 |
REAL betam |
| 87 |
! pour Phim / h dans la S.L stable |
| 88 |
PARAMETER (betam=15.0) |
| 89 |
! z/OBL<>1 |
| 90 |
REAL sffrac |
| 91 |
! S.L. = z/h < .1 |
| 92 |
PARAMETER (sffrac=0.1) |
| 93 |
REAL binm |
| 94 |
PARAMETER (binm=betam*sffrac) |
| 95 |
|
| 96 |
REAL q_star, t_star |
| 97 |
! Lambert correlations T' q' avec T* q* |
| 98 |
REAL b1, b2, b212, b2sr |
| 99 |
PARAMETER (b1=70., b2=20.) |
| 100 |
|
| 101 |
REAL z(klon, klev) |
| 102 |
|
| 103 |
REAL zref |
| 104 |
! Niveau de ref a 2m peut eventuellement |
| 105 |
PARAMETER (zref=2.) |
| 106 |
! etre choisi a 10m |
| 107 |
|
| 108 |
INTEGER i, k |
| 109 |
REAL zxt |
| 110 |
! surface kinematic heat flux [mK/s] |
| 111 |
REAL khfs(klon) |
| 112 |
! sfc kinematic constituent flux [m/s] |
| 113 |
REAL kqfs(klon) |
| 114 |
! surface virtual heat flux |
| 115 |
REAL heatv(klon) |
| 116 |
! bulk Richardon no. mais en Theta_v |
| 117 |
REAL rhino(klon, klev) |
| 118 |
! pts w/unstbl pbl (positive virtual ht flx) |
| 119 |
LOGICAL unstbl(klon) |
| 120 |
LOGICAL check(klon) ! Richardson number > critical |
| 121 |
! flag de prolongerment cape pour pt Omega |
| 122 |
LOGICAL omegafl(klon) |
| 123 |
|
| 124 |
! Monin-Obukhov lengh |
| 125 |
REAL obklen(klon) |
| 126 |
|
| 127 |
REAL zdu2 |
| 128 |
! Algorithme thermique |
| 129 |
REAL s(klon, klev) ! [P/Po]^Kappa milieux couches |
| 130 |
! total water of thermal |
| 131 |
REAL qT_th(klon) |
| 132 |
! T thermique niveau precedent |
| 133 |
REAL qsatbef(klon) |
| 134 |
! le thermique est sature |
| 135 |
LOGICAL Zsat(klon) |
| 136 |
REAL zthvd, zthvu, qqsat |
| 137 |
REAL t2 |
| 138 |
|
| 139 |
! inverse phi function for momentum |
| 140 |
REAL phiminv(klon) |
| 141 |
! turbulent velocity scale for momentum |
| 142 |
REAL wm(klon) |
| 143 |
! current level height + one level up |
| 144 |
REAL zp(klon) |
| 145 |
REAL zcor |
| 146 |
|
| 147 |
REAL pblmin |
| 148 |
|
| 149 |
!----------------------------------------------------------------- |
| 150 |
|
| 151 |
knon = size(pblh) |
| 152 |
|
| 153 |
! initialisations |
| 154 |
q_star = 0 |
| 155 |
t_star = 0 |
| 156 |
|
| 157 |
b212=sqrt(b1*b2) |
| 158 |
b2sr=sqrt(b2) |
| 159 |
|
| 160 |
! Calculer les hauteurs de chaque couche |
| 161 |
! (geopotentielle Int_dp/ro = Int_[Rd.T.dp/p] z = geop/g) |
| 162 |
! pourquoi ne pas utiliser Phi/RG ? |
| 163 |
DO i = 1, knon |
| 164 |
z(i, 1) = RD * t(i, 1) / (0.5*(paprs(i, 1)+pplay(i, 1))) & |
| 165 |
* (paprs(i, 1)-pplay(i, 1)) / RG |
| 166 |
s(i, 1) = (pplay(i, 1)/paprs(i, 1))**RKappa |
| 167 |
ENDDO |
| 168 |
! s(k) = [pplay(k)/ps]^kappa |
| 169 |
! + + + + + + + + + pplay <-> s(k) t dp=pplay(k-1)-pplay(k) |
| 170 |
! ----------------- paprs <-> sig(k) |
| 171 |
! + + + + + + + + + pplay <-> s(k-1) |
| 172 |
! + + + + + + + + + pplay <-> s(1) t dp=paprs-pplay z(1) |
| 173 |
! ----------------- paprs <-> sig(1) |
| 174 |
|
| 175 |
DO k = 2, klev |
| 176 |
DO i = 1, knon |
| 177 |
z(i, k) = z(i, k-1) & |
| 178 |
+ RD * 0.5*(t(i, k-1)+t(i, k)) / paprs(i, k) & |
| 179 |
* (pplay(i, k-1)-pplay(i, k)) / RG |
| 180 |
s(i, k) = (pplay(i, k) / paprs(i, 1))**RKappa |
| 181 |
ENDDO |
| 182 |
ENDDO |
| 183 |
|
| 184 |
! Determination des grandeurs de surface |
| 185 |
DO i = 1, knon |
| 186 |
! Niveau de ref choisi a 2m |
| 187 |
zxt = t2m(i) |
| 188 |
|
| 189 |
! convention >0 vers le bas ds lmdz |
| 190 |
khfs(i) = - flux_t(i)*zxt*Rd / (RCPD*paprs(i, 1)) |
| 191 |
kqfs(i) = - flux_q(i)*zxt*Rd / paprs(i, 1) |
| 192 |
! verifier que khfs et kqfs sont bien de la forme w'l' |
| 193 |
heatv(i) = khfs(i) + 0.608*zxt*kqfs(i) |
| 194 |
! a comparer aussi aux sorties de clqh : flux_T/RoCp et flux_q/RoLv |
| 195 |
! Theta et qT du thermique sans exces (interpolin vers surf) |
| 196 |
! chgt de niveau du thermique (jeudi 30/12/1999) |
| 197 |
! (interpolation lineaire avant integration phi_h) |
| 198 |
qT_th(i) = q2m(i) |
| 199 |
ENDDO |
| 200 |
|
| 201 |
DO i = 1, knon |
| 202 |
! Global Richardson |
| 203 |
rhino(i, 1) = 0.0 |
| 204 |
check(i) = .TRUE. |
| 205 |
! on initialise pblh a l'altitude du 1er niv |
| 206 |
pblh(i) = z(i, 1) |
| 207 |
plcl(i) = 6000. |
| 208 |
! Lambda = -u*^3 / (alpha.g.kvon.<w'Theta'v> |
| 209 |
obklen(i) = -t(i, 1)*ustar(i)**3/(RG*vk*heatv(i)) |
| 210 |
trmb1(i) = 0. |
| 211 |
trmb2(i) = 0. |
| 212 |
trmb3(i) = 0. |
| 213 |
ENDDO |
| 214 |
|
| 215 |
! PBL height calculation: Search for level of pbl. Scan upward |
| 216 |
! until the Richardson number between the first level and the |
| 217 |
! current level exceeds the "critical" value. (bonne idee Nu de |
| 218 |
! separer le Ric et l'exces de temp du thermique) |
| 219 |
DO k = 2, isommet |
| 220 |
DO i = 1, knon |
| 221 |
IF (check(i)) THEN |
| 222 |
! pourquoi / niveau 1 (au lieu du sol) et le terme en u*^2 ? |
| 223 |
zdu2 = u(i, k)**2+v(i, k)**2 |
| 224 |
zdu2 = max(zdu2, 1.0e-20) |
| 225 |
! Theta_v environnement |
| 226 |
zthvd=t(i, k)/s(i, k)*(1.+RETV*q(i, k)) |
| 227 |
|
| 228 |
! therm Theta_v sans exces (avec hypothese fausse de H&B, sinon, |
| 229 |
! passer par Theta_e et virpot) |
| 230 |
zthvu = T2m(i)*(1.+RETV*qT_th(i)) |
| 231 |
! Le Ri par Theta_v |
| 232 |
! On a nveau de ref a 2m ??? |
| 233 |
rhino(i, k) = (z(i, k)-zref)*RG*(zthvd-zthvu) & |
| 234 |
/(zdu2*0.5*(zthvd+zthvu)) |
| 235 |
|
| 236 |
IF (rhino(i, k).GE.ricr) THEN |
| 237 |
pblh(i) = z(i, k-1) + (z(i, k-1)-z(i, k)) * & |
| 238 |
(ricr-rhino(i, k-1))/(rhino(i, k-1)-rhino(i, k)) |
| 239 |
! test04 |
| 240 |
pblh(i) = pblh(i) + 100. |
| 241 |
pblT(i) = t(i, k-1) + (t(i, k)-t(i, k-1)) * & |
| 242 |
(pblh(i)-z(i, k-1))/(z(i, k)-z(i, k-1)) |
| 243 |
check(i) = .FALSE. |
| 244 |
ENDIF |
| 245 |
ENDIF |
| 246 |
ENDDO |
| 247 |
ENDDO |
| 248 |
|
| 249 |
! Set pbl height to maximum value where computation exceeds number of |
| 250 |
! layers allowed |
| 251 |
DO i = 1, knon |
| 252 |
if (check(i)) pblh(i) = z(i, isommet) |
| 253 |
ENDDO |
| 254 |
|
| 255 |
! Improve estimate of pbl height for the unstable points. |
| 256 |
! Find unstable points (sensible heat flux is upward): |
| 257 |
DO i = 1, knon |
| 258 |
IF (heatv(i) > 0.) THEN |
| 259 |
unstbl(i) = .TRUE. |
| 260 |
check(i) = .TRUE. |
| 261 |
ELSE |
| 262 |
unstbl(i) = .FALSE. |
| 263 |
check(i) = .FALSE. |
| 264 |
ENDIF |
| 265 |
ENDDO |
| 266 |
|
| 267 |
! For the unstable case, compute velocity scale and the |
| 268 |
! convective temperature excess: |
| 269 |
DO i = 1, knon |
| 270 |
IF (check(i)) THEN |
| 271 |
phiminv(i) = (1.-binm*pblh(i)/obklen(i))**onet |
| 272 |
|
| 273 |
! CALCUL DE wm |
| 274 |
! Ici on considerera que l'on est dans la couche de surf jusqu'a 100 |
| 275 |
! On prend svt couche de surface=0.1*h mais on ne connait pas h |
| 276 |
! Dans la couche de surface |
| 277 |
wm(i)= ustar(i)*phiminv(i) |
| 278 |
|
| 279 |
! forme Mathieu : |
| 280 |
q_star = kqfs(i)/wm(i) |
| 281 |
t_star = khfs(i)/wm(i) |
| 282 |
|
| 283 |
therm(i) = sqrt( b1*(1.+2.*RETV*qT_th(i))*t_star**2 & |
| 284 |
+ (RETV*T2m(i))**2*b2*q_star**2 & |
| 285 |
+ max(0., 2.*RETV*T2m(i)*b212*q_star*t_star)) |
| 286 |
|
| 287 |
! Theta et qT du thermique (forme H&B) avec exces |
| 288 |
! (attention, on ajoute therm(i) qui est virtuelle ...) |
| 289 |
! pourquoi pas sqrt(b1)*t_star ? |
| 290 |
qT_th(i) = qT_th(i) + b2sr*q_star |
| 291 |
! new on differre le calcul de Theta_e |
| 292 |
rhino(i, 1) = 0. |
| 293 |
ENDIF |
| 294 |
ENDDO |
| 295 |
|
| 296 |
! Improve pblh estimate for unstable conditions using the |
| 297 |
! convective temperature excess : |
| 298 |
DO k = 2, isommet |
| 299 |
DO i = 1, knon |
| 300 |
IF (check(i)) THEN |
| 301 |
zdu2 = u(i, k)**2 + v(i, k)**2 |
| 302 |
zdu2 = max(zdu2, 1e-20) |
| 303 |
! Theta_v environnement |
| 304 |
zthvd=t(i, k)/s(i, k)*(1.+RETV*q(i, k)) |
| 305 |
|
| 306 |
! et therm Theta_v (avec hypothese de constance de H&B, |
| 307 |
zthvu = T2m(i)*(1.+RETV*qT_th(i)) + therm(i) |
| 308 |
|
| 309 |
! Le Ri par Theta_v |
| 310 |
! Niveau de ref 2m |
| 311 |
rhino(i, k) = (z(i, k)-zref)*RG*(zthvd-zthvu) & |
| 312 |
/(zdu2*0.5*(zthvd+zthvu)) |
| 313 |
|
| 314 |
IF (rhino(i, k).GE.ricr) THEN |
| 315 |
pblh(i) = z(i, k-1) + (z(i, k-1)-z(i, k)) * & |
| 316 |
(ricr-rhino(i, k-1))/(rhino(i, k-1)-rhino(i, k)) |
| 317 |
! test04 |
| 318 |
pblh(i) = pblh(i) + 100. |
| 319 |
pblT(i) = t(i, k-1) + (t(i, k)-t(i, k-1)) * & |
| 320 |
(pblh(i)-z(i, k-1))/(z(i, k)-z(i, k-1)) |
| 321 |
check(i) = .FALSE. |
| 322 |
ENDIF |
| 323 |
ENDIF |
| 324 |
ENDDO |
| 325 |
ENDDO |
| 326 |
|
| 327 |
! Set pbl height to maximum value where computation exceeds number of |
| 328 |
! layers allowed |
| 329 |
DO i = 1, knon |
| 330 |
if (check(i)) pblh(i) = z(i, isommet) |
| 331 |
ENDDO |
| 332 |
|
| 333 |
! PBL height must be greater than some minimum mechanical mixing depth |
| 334 |
! Several investigators have proposed minimum mechanical mixing depth |
| 335 |
! relationships as a function of the local friction velocity, u*. We |
| 336 |
! make use of a linear relationship of the form h = c u* where c=700. |
| 337 |
! The scaling arguments that give rise to this relationship most often |
| 338 |
! represent the coefficient c as some constant over the local coriolis |
| 339 |
! parameter. Here we make use of the experimental results of Koracin |
| 340 |
! and Berkowicz (1988) [BLM, Vol 43] for wich they recommend 0.07/f |
| 341 |
! where f was evaluated at 39.5 N and 52 N. Thus we use a typical mid |
| 342 |
! latitude value for f so that c = 0.07/f = 700. |
| 343 |
DO i = 1, knon |
| 344 |
pblmin = 700. * ustar(i) |
| 345 |
pblh(i) = MAX(pblh(i), pblmin) |
| 346 |
! par exemple : |
| 347 |
pblT(i) = t(i, 2) + (t(i, 3)-t(i, 2)) * & |
| 348 |
(pblh(i)-z(i, 2))/(z(i, 3)-z(i, 2)) |
| 349 |
ENDDO |
| 350 |
|
| 351 |
! pblh is now available; do preparation for diffusivity calculation: |
| 352 |
DO i = 1, knon |
| 353 |
check(i) = .TRUE. |
| 354 |
Zsat(i) = .FALSE. |
| 355 |
! omegafl utilise pour prolongement CAPE |
| 356 |
omegafl(i) = .FALSE. |
| 357 |
Cape(i) = 0. |
| 358 |
EauLiq(i) = 0. |
| 359 |
CTEI(i) = 0. |
| 360 |
|
| 361 |
! Do additional preparation for unstable cases only, set temperature |
| 362 |
! and moisture perturbations depending on stability. |
| 363 |
! Remarque : les formule sont prises dans leur forme CS |
| 364 |
IF (unstbl(i)) THEN |
| 365 |
! Niveau de ref du thermique |
| 366 |
zxt=(T2m(i)-zref*0.5*RG/RCPD/(1.+RVTMP2*qT_th(i))) & |
| 367 |
*(1.+RETV*qT_th(i)) |
| 368 |
phiminv(i) = (1. - binm*pblh(i)/obklen(i))**onet |
| 369 |
wm(i) = ustar(i)*phiminv(i) |
| 370 |
ENDIF |
| 371 |
ENDDO |
| 372 |
|
| 373 |
! Main level loop to compute the diffusivities and |
| 374 |
! counter-gradient terms: |
| 375 |
loop_level: DO k = 2, isommet |
| 376 |
! Find levels within boundary layer: |
| 377 |
DO i = 1, knon |
| 378 |
zp(i) = z(i, k) |
| 379 |
IF (zkmin == 0. .AND. zp(i) > pblh(i)) zp(i) = pblh(i) |
| 380 |
ENDDO |
| 381 |
|
| 382 |
! For all layers, compute integral info and CTEI |
| 383 |
DO i = 1, knon |
| 384 |
if (check(i) .or. omegafl(i)) then |
| 385 |
if (.not. Zsat(i)) then |
| 386 |
T2 = T2m(i) * s(i, k) |
| 387 |
! thermodyn functions |
| 388 |
qqsat= r2es * FOEEW(T2, RTT >= T2) / pplay(i, k) |
| 389 |
qqsat=MIN(0.5, qqsat) |
| 390 |
zcor=1./(1.-retv*qqsat) |
| 391 |
qqsat=qqsat*zcor |
| 392 |
|
| 393 |
if (qqsat < qT_th(i)) then |
| 394 |
! on calcule lcl |
| 395 |
if (k == 2) then |
| 396 |
plcl(i) = z(i, k) |
| 397 |
else |
| 398 |
plcl(i) = z(i, k-1) + (z(i, k-1)-z(i, k)) & |
| 399 |
* (qT_th(i)-qsatbef(i)) / (qsatbef(i)-qqsat) |
| 400 |
endif |
| 401 |
Zsat(i) = .true. |
| 402 |
endif |
| 403 |
endif |
| 404 |
qsatbef(i) = qqsat |
| 405 |
! cette ligne a deja ete faite normalement ? |
| 406 |
endif |
| 407 |
ENDDO |
| 408 |
end DO loop_level |
| 409 |
|
| 410 |
END SUBROUTINE HBTM |
| 411 |
|
| 412 |
end module HBTM_m |
Parent Directory
| 
Revision Log