1 | SUBROUTINE prandtl_fluxes |
---|
2 | |
---|
3 | !------------------------------------------------------------------------------! |
---|
4 | ! Actual revisions: |
---|
5 | ! ----------------- |
---|
6 | ! |
---|
7 | ! |
---|
8 | ! Former revisions: |
---|
9 | ! ----------------- |
---|
10 | ! $Id: prandtl_fluxes.f90 4 2007-02-13 11:33:16Z raasch $ |
---|
11 | ! RCS Log replace by Id keyword, revision history cleaned up |
---|
12 | ! |
---|
13 | ! Revision 1.19 2006/04/26 12:24:35 raasch |
---|
14 | ! +OpenMP directives and optimization (array assignments replaced by DO loops) |
---|
15 | ! |
---|
16 | ! Revision 1.1 1998/01/23 10:06:06 raasch |
---|
17 | ! Initial revision |
---|
18 | ! |
---|
19 | ! |
---|
20 | ! Description: |
---|
21 | ! ------------ |
---|
22 | ! Diagnostic computation of vertical fluxes in the Prandtl layer from the |
---|
23 | ! values of the variables at grid point k=1 |
---|
24 | !------------------------------------------------------------------------------! |
---|
25 | |
---|
26 | USE arrays_3d |
---|
27 | USE control_parameters |
---|
28 | USE grid_variables |
---|
29 | USE indices |
---|
30 | |
---|
31 | IMPLICIT NONE |
---|
32 | |
---|
33 | INTEGER :: i, j, k |
---|
34 | REAL :: a, b, rifm, uv_total, z_p |
---|
35 | |
---|
36 | ! |
---|
37 | !-- Compute theta* |
---|
38 | IF ( constant_heatflux ) THEN |
---|
39 | ! |
---|
40 | !-- For a given heat flux in the Prandtl layer: |
---|
41 | !-- for u* use the value from the previous time step |
---|
42 | !$OMP PARALLEL DO |
---|
43 | DO i = nxl-1, nxr+1 |
---|
44 | DO j = nys-1, nyn+1 |
---|
45 | ts(j,i) = -shf(j,i) / ( us(j,i) + 1E-30 ) |
---|
46 | ! |
---|
47 | !-- ts must be limited, because otherwise overflow may occur in case of |
---|
48 | !-- us=0 when computing rif further below |
---|
49 | IF ( ts(j,i) < -1.05E5 ) ts = -1.0E5 |
---|
50 | IF ( ts(j,i) > 1.0E5 ) ts = 1.0E5 |
---|
51 | ENDDO |
---|
52 | ENDDO |
---|
53 | |
---|
54 | ELSE |
---|
55 | ! |
---|
56 | !-- For a given surface temperature: |
---|
57 | !-- (the Richardson number is still the one from the previous time step) |
---|
58 | !$OMP PARALLEL DO PRIVATE( a, b, k, z_p ) |
---|
59 | DO i = nxl-1, nxr+1 |
---|
60 | DO j = nys-1, nyn+1 |
---|
61 | |
---|
62 | k = nzb_s_inner(j,i) |
---|
63 | z_p = zu(k+1) - zw(k) |
---|
64 | |
---|
65 | IF ( rif(j,i) >= 0.0 ) THEN |
---|
66 | ! |
---|
67 | !-- Stable stratification |
---|
68 | ts(j,i) = kappa * ( pt(k+1,j,i) - pt(k,j,i) ) / ( & |
---|
69 | LOG( z_p / z0(j,i) ) + & |
---|
70 | 5.0 * rif(j,i) * ( z_p - z0(j,i) ) / z_p & |
---|
71 | ) |
---|
72 | ELSE |
---|
73 | ! |
---|
74 | !-- Unstable stratification |
---|
75 | a = SQRT( 1.0 - 16.0 * rif(j,i) ) |
---|
76 | b = SQRT( 1.0 - 16.0 * rif(j,i) / z_p * z0(j,i) ) |
---|
77 | ! |
---|
78 | !-- If a borderline case occurs, the formula for stable |
---|
79 | !-- stratification must be used anyway, or else a zero division |
---|
80 | !-- would occur in the argument of the logarithm |
---|
81 | IF ( a == 1.0 .OR. b == 1.0 ) THEN |
---|
82 | ts(j,i) = kappa * ( pt(k+1,j,i) - pt(k,j,i) ) / ( & |
---|
83 | LOG( z_p / z0(j,i) ) + & |
---|
84 | 5.0 * rif(j,i) * ( z_p - z0(j,i) ) / z_p & |
---|
85 | ) |
---|
86 | ELSE |
---|
87 | ts(j,i) = kappa * ( pt(k+1,j,i) - pt(k,j,i) ) / ( & |
---|
88 | LOG( (1.0+b) / (1.0-b) * (1.0-a) / (1.0+a) ) & |
---|
89 | ) |
---|
90 | ENDIF |
---|
91 | ENDIF |
---|
92 | |
---|
93 | ENDDO |
---|
94 | ENDDO |
---|
95 | ENDIF |
---|
96 | |
---|
97 | ! |
---|
98 | !-- Compute z_p/L (corresponds to the Richardson-flux number) |
---|
99 | IF ( .NOT. moisture ) THEN |
---|
100 | !$OMP PARALLEL DO PRIVATE( k, z_p ) |
---|
101 | DO i = nxl-1, nxr+1 |
---|
102 | DO j = nys-1, nyn+1 |
---|
103 | k = nzb_s_inner(j,i) |
---|
104 | z_p = zu(k+1) - zw(k) |
---|
105 | rif(j,i) = z_p * kappa * g * ts(j,i) / & |
---|
106 | ( pt(k+1,j,i) * ( us(j,i)**2 + 1E-30 ) ) |
---|
107 | ! |
---|
108 | !-- Limit the value range of the Richardson numbers. |
---|
109 | !-- This is necessary for very small velocities (u,v --> 0), because |
---|
110 | !-- the absolute value of rif can then become very large, which in |
---|
111 | !-- consequence would result in very large shear stresses and very |
---|
112 | !-- small momentum fluxes (both are generally unrealistic). |
---|
113 | IF ( rif(j,i) < rif_min ) rif(j,i) = rif_min |
---|
114 | IF ( rif(j,i) > rif_max ) rif(j,i) = rif_max |
---|
115 | ENDDO |
---|
116 | ENDDO |
---|
117 | ELSE |
---|
118 | !$OMP PARALLEL DO PRIVATE( k, z_p ) |
---|
119 | DO i = nxl-1, nxr+1 |
---|
120 | DO j = nys-1, nyn+1 |
---|
121 | k = nzb_s_inner(j,i) |
---|
122 | z_p = zu(k+1) - zw(k) |
---|
123 | rif(j,i) = z_p * kappa * g * & |
---|
124 | ( ts(j,i) + 0.61 * pt(k+1,j,i) * qs(j,i) ) / & |
---|
125 | ( vpt(k+1,j,i) * ( us(j,i)**2 + 1E-30 ) ) |
---|
126 | ! |
---|
127 | !-- Limit the value range of the Richardson numbers. |
---|
128 | !-- This is necessary for very small velocities (u,v --> 0), because |
---|
129 | !-- the absolute value of rif can then become very large, which in |
---|
130 | !-- consequence would result in very large shear stresses and very |
---|
131 | !-- small momentum fluxes (both are generally unrealistic). |
---|
132 | IF ( rif(j,i) < rif_min ) rif(j,i) = rif_min |
---|
133 | IF ( rif(j,i) > rif_max ) rif(j,i) = rif_max |
---|
134 | ENDDO |
---|
135 | ENDDO |
---|
136 | ENDIF |
---|
137 | |
---|
138 | ! |
---|
139 | !-- Compute u* at the scalars' grid points |
---|
140 | !$OMP PARALLEL DO PRIVATE( a, b, k, uv_total, z_p ) |
---|
141 | DO i = nxl, nxr |
---|
142 | DO j = nys, nyn |
---|
143 | |
---|
144 | k = nzb_s_inner(j,i) |
---|
145 | z_p = zu(k+1) - zw(k) |
---|
146 | |
---|
147 | ! |
---|
148 | !-- Compute the absolute value of the horizontal velocity |
---|
149 | uv_total = SQRT( ( 0.5 * ( u(k+1,j,i) + u(k+1,j,i+1) ) )**2 + & |
---|
150 | ( 0.5 * ( v(k+1,j,i) + v(k+1,j+1,i) ) )**2 & |
---|
151 | ) |
---|
152 | |
---|
153 | IF ( rif(j,i) >= 0.0 ) THEN |
---|
154 | ! |
---|
155 | !-- Stable stratification |
---|
156 | us(j,i) = kappa * uv_total / ( & |
---|
157 | LOG( z_p / z0(j,i) ) + & |
---|
158 | 5.0 * rif(j,i) * ( z_p - z0(j,i) ) / z_p & |
---|
159 | ) |
---|
160 | ELSE |
---|
161 | ! |
---|
162 | !-- Unstable stratification |
---|
163 | a = 1.0 / SQRT( SQRT( 1.0 - 16.0 * rif(j,i) ) ) |
---|
164 | b = 1.0 / SQRT( SQRT( 1.0 - 16.0 * rif(j,i) / z_p * z0(j,i) ) ) |
---|
165 | ! |
---|
166 | !-- If a borderline case occurs, the formula for stable stratification |
---|
167 | !-- must be used anyway, or else a zero division would occur in the |
---|
168 | !-- argument of the logarithm. |
---|
169 | IF ( a == 1.0 .OR. b == 1.0 ) THEN |
---|
170 | us(j,i) = kappa * uv_total / ( & |
---|
171 | LOG( z_p / z0(j,i) ) + & |
---|
172 | 5.0 * rif(j,i) * ( z_p - z0(j,i) ) / z_p & |
---|
173 | ) |
---|
174 | ELSE |
---|
175 | us(j,i) = kappa * uv_total / ( & |
---|
176 | LOG( (1.0+b) / (1.0-b) * (1.0-a) / (1.0+a) ) + & |
---|
177 | 2.0 * ( ATAN( b ) - ATAN( a ) ) & |
---|
178 | ) |
---|
179 | ENDIF |
---|
180 | ENDIF |
---|
181 | ENDDO |
---|
182 | ENDDO |
---|
183 | |
---|
184 | ! |
---|
185 | !-- Compute u'w' for the total model domain. |
---|
186 | !-- First compute the corresponding component of u* and square it. |
---|
187 | !$OMP PARALLEL DO PRIVATE( a, b, k, rifm, z_p ) |
---|
188 | DO i = nxl, nxr |
---|
189 | DO j = nys, nyn |
---|
190 | |
---|
191 | k = nzb_u_inner(j,i) |
---|
192 | z_p = zu(k+1) - zw(k) |
---|
193 | |
---|
194 | ! |
---|
195 | !-- Compute Richardson-flux number for this point |
---|
196 | rifm = 0.5 * ( rif(j,i-1) + rif(j,i) ) |
---|
197 | IF ( rifm >= 0.0 ) THEN |
---|
198 | ! |
---|
199 | !-- Stable stratification |
---|
200 | usws(j,i) = kappa * u(k+1,j,i) / ( & |
---|
201 | LOG( z_p / z0(j,i) ) + & |
---|
202 | 5.0 * rifm * ( z_p - z0(j,i) ) / z_p & |
---|
203 | ) |
---|
204 | ELSE |
---|
205 | ! |
---|
206 | !-- Unstable stratification |
---|
207 | a = 1.0 / SQRT( SQRT( 1.0 - 16.0 * rifm ) ) |
---|
208 | b = 1.0 / SQRT( SQRT( 1.0 - 16.0 * rifm / z_p * z0(j,i) ) ) |
---|
209 | ! |
---|
210 | !-- If a borderline case occurs, the formula for stable stratification |
---|
211 | !-- must be used anyway, or else a zero division would occur in the |
---|
212 | !-- argument of the logarithm. |
---|
213 | IF ( a == 1.0 .OR. B == 1.0 ) THEN |
---|
214 | usws(j,i) = kappa * u(k+1,j,i) / ( & |
---|
215 | LOG( z_p / z0(j,i) ) + & |
---|
216 | 5.0 * rifm * ( z_p - z0(j,i) ) / z_p & |
---|
217 | ) |
---|
218 | ELSE |
---|
219 | usws(j,i) = kappa * u(k+1,j,i) / ( & |
---|
220 | LOG( (1.0+b) / (1.0-b) * (1.0-a) / (1.0+a) ) + & |
---|
221 | 2.0 * ( ATAN( b ) - ATAN( a ) ) & |
---|
222 | ) |
---|
223 | ENDIF |
---|
224 | ENDIF |
---|
225 | usws(j,i) = -usws(j,i) * ABS( usws(j,i) ) |
---|
226 | ENDDO |
---|
227 | ENDDO |
---|
228 | |
---|
229 | ! |
---|
230 | !-- Compute v'w' for the total model domain. |
---|
231 | !-- First compute the corresponding component of u* and square it. |
---|
232 | !$OMP PARALLEL DO PRIVATE( a, b, k, rifm, z_p ) |
---|
233 | DO i = nxl, nxr |
---|
234 | DO j = nys, nyn |
---|
235 | |
---|
236 | k = nzb_v_inner(j,i) |
---|
237 | z_p = zu(k+1) - zw(k) |
---|
238 | |
---|
239 | ! |
---|
240 | !-- Compute Richardson-flux number for this point |
---|
241 | rifm = 0.5 * ( rif(j-1,i) + rif(j,i) ) |
---|
242 | IF ( rifm >= 0.0 ) THEN |
---|
243 | ! |
---|
244 | !-- Stable stratification |
---|
245 | vsws(j,i) = kappa * v(k+1,j,i) / ( & |
---|
246 | LOG( z_p / z0(j,i) ) + & |
---|
247 | 5.0 * rifm * ( z_p - z0(j,i) ) / z_p & |
---|
248 | ) |
---|
249 | ELSE |
---|
250 | ! |
---|
251 | !-- Unstable stratification |
---|
252 | a = 1.0 / SQRT( SQRT( 1.0 - 16.0 * rifm ) ) |
---|
253 | b = 1.0 / SQRT( SQRT( 1.0 - 16.0 * rifm / z_p * z0(j,i) ) ) |
---|
254 | ! |
---|
255 | !-- If a borderline case occurs, the formula for stable stratification |
---|
256 | !-- must be used anyway, or else a zero division would occur in the |
---|
257 | !-- argument of the logarithm. |
---|
258 | IF ( a == 1.0 .OR. b == 1.0 ) THEN |
---|
259 | vsws(j,i) = kappa * v(k+1,j,i) / ( & |
---|
260 | LOG( z_p / z0(j,i) ) + & |
---|
261 | 5.0 * rifm * ( z_p - z0(j,i) ) / z_p & |
---|
262 | ) |
---|
263 | ELSE |
---|
264 | vsws(j,i) = kappa * v(k+1,j,i) / ( & |
---|
265 | LOG( (1.0+b) / (1.0-b) * (1.0-a) / (1.0+a) ) + & |
---|
266 | 2.0 * ( ATAN( b ) - ATAN( a ) ) & |
---|
267 | ) |
---|
268 | ENDIF |
---|
269 | ENDIF |
---|
270 | vsws(j,i) = -vsws(j,i) * ABS( vsws(j,i) ) |
---|
271 | ENDDO |
---|
272 | ENDDO |
---|
273 | |
---|
274 | ! |
---|
275 | !-- If required compute q* |
---|
276 | IF ( moisture .OR. passive_scalar ) THEN |
---|
277 | IF ( constant_waterflux ) THEN |
---|
278 | ! |
---|
279 | !-- For a given water flux in the Prandtl layer: |
---|
280 | !$OMP PARALLEL DO |
---|
281 | DO i = nxl-1, nxr+1 |
---|
282 | DO j = nys-1, nyn+1 |
---|
283 | qs(j,i) = -qsws(j,i) / ( us(j,i) + 1E-30 ) |
---|
284 | ENDDO |
---|
285 | ENDDO |
---|
286 | |
---|
287 | ELSE |
---|
288 | !$OMP PARALLEL DO PRIVATE( a, b, k, z_p ) |
---|
289 | DO i = nxl-1, nxr+1 |
---|
290 | DO j = nys-1, nyn+1 |
---|
291 | |
---|
292 | k = nzb_s_inner(j,i) |
---|
293 | z_p = zu(k+1) - zw(k) |
---|
294 | |
---|
295 | IF ( rif(j,i) >= 0.0 ) THEN |
---|
296 | ! |
---|
297 | !-- Stable stratification |
---|
298 | qs(j,i) = kappa * ( q(k+1,j,i) - q(k,j,i) ) / ( & |
---|
299 | LOG( z_p / z0(j,i) ) + & |
---|
300 | 5.0 * rif(j,i) * ( z_p - z0(j,i) ) / z_p & |
---|
301 | ) |
---|
302 | ELSE |
---|
303 | ! |
---|
304 | !-- Unstable stratification |
---|
305 | a = SQRT( 1.0 - 16.0 * rif(j,i) ) |
---|
306 | b = SQRT( 1.0 - 16.0 * rif(j,i) / z_p * z0(j,i) ) |
---|
307 | ! |
---|
308 | !-- If a borderline case occurs, the formula for stable |
---|
309 | !-- stratification must be used anyway, or else a zero division |
---|
310 | !-- would occur in the argument of the logarithm. |
---|
311 | IF ( a == 1.0 .OR. b == 1.0 ) THEN |
---|
312 | qs(j,i) = kappa * ( q(k+1,j,i) - q(k,j,i) ) / ( & |
---|
313 | LOG( z_p / z0(j,i) ) + & |
---|
314 | 5.0 * rif(j,i) * ( z_p - z0(j,i) ) / z_p & |
---|
315 | ) |
---|
316 | ELSE |
---|
317 | qs(j,i) = kappa * ( q(k+1,j,i) - q(k,j,i) ) / ( & |
---|
318 | LOG( (1.0+b) / (1.0-b) * (1.0-a) / (1.0+a) ) & |
---|
319 | ) |
---|
320 | ENDIF |
---|
321 | ENDIF |
---|
322 | |
---|
323 | ENDDO |
---|
324 | ENDDO |
---|
325 | ENDIF |
---|
326 | ENDIF |
---|
327 | |
---|
328 | ! |
---|
329 | !-- Exchange the boundaries for u* and the momentum fluxes (fluxes only for |
---|
330 | !-- completeness's sake). |
---|
331 | CALL exchange_horiz_2d( us ) |
---|
332 | CALL exchange_horiz_2d( usws ) |
---|
333 | CALL exchange_horiz_2d( vsws ) |
---|
334 | IF ( moisture .OR. passive_scalar ) CALL exchange_horiz_2d( qsws ) |
---|
335 | |
---|
336 | ! |
---|
337 | !-- Compute the vertical kinematic heat flux |
---|
338 | IF ( .NOT. constant_heatflux ) THEN |
---|
339 | !$OMP PARALLEL DO |
---|
340 | DO i = nxl-1, nxr+1 |
---|
341 | DO j = nys-1, nyn+1 |
---|
342 | shf(j,i) = -ts(j,i) * us(j,i) |
---|
343 | ENDDO |
---|
344 | ENDDO |
---|
345 | ENDIF |
---|
346 | |
---|
347 | ! |
---|
348 | !-- Compute the vertical water/scalar flux |
---|
349 | IF ( .NOT. constant_heatflux .AND. ( moisture .OR. passive_scalar ) ) THEN |
---|
350 | !$OMP PARALLEL DO |
---|
351 | DO i = nxl-1, nxr+1 |
---|
352 | DO j = nys-1, nyn+1 |
---|
353 | qsws(j,i) = -qs(j,i) * us(j,i) |
---|
354 | ENDDO |
---|
355 | ENDDO |
---|
356 | ENDIF |
---|
357 | |
---|
358 | ! |
---|
359 | !-- Bottom boundary condition for the TKE |
---|
360 | IF ( ibc_e_b == 2 ) THEN |
---|
361 | !$OMP PARALLEL DO |
---|
362 | DO i = nxl-1, nxr+1 |
---|
363 | DO j = nys-1, nyn+1 |
---|
364 | e(nzb_s_inner(j,i)+1,j,i) = ( us(j,i) / 0.1 )**2 |
---|
365 | ! |
---|
366 | !-- As a test: cm = 0.4 |
---|
367 | ! e(nzb_s_inner(j,i)+1,j,i) = ( us(j,i) / 0.4 )**2 |
---|
368 | e(nzb_s_inner(j,i),j,i) = e(nzb_s_inner(j,i)+1,j,i) |
---|
369 | ENDDO |
---|
370 | ENDDO |
---|
371 | ENDIF |
---|
372 | |
---|
373 | |
---|
374 | END SUBROUTINE prandtl_fluxes |
---|