1 |
heimbach |
1.2 |
Example: Labrador Sea Region with Sea-Ice |
2 |
|
|
========================================= |
3 |
|
|
|
4 |
|
|
This example sets up a small (20x16x23) Labrador Sea experiment |
5 |
|
|
coupled to a dynamic thermodynamic sea-ice model. |
6 |
dimitri |
1.3 |
A brief description of the sea-ice model is in "seaice.ps". |
7 |
heimbach |
1.2 |
|
8 |
|
|
The domain of integration spans 280E to 320E and 46N to 78N. |
9 |
|
|
Horizontal grid spacing is 2 degrees. |
10 |
|
|
The 23 vertical levels and the bathymetry file |
11 |
|
|
bathyFile = 'bathy.labsea' |
12 |
|
|
are obtained from the the 2-degree ECCO configuration. |
13 |
|
|
|
14 |
|
|
Integration is initialized from annual-mean Levitus climatology |
15 |
|
|
hydrogThetaFile = 'LevCli_temp.labsea' |
16 |
|
|
hydrogSaltFile = 'LevCli_salt.labsea' |
17 |
|
|
|
18 |
|
|
Surface salinity relaxation is to the monthly mean Levitus climatology |
19 |
|
|
saltClimFile = 'SSS.labsea' |
20 |
|
|
|
21 |
|
|
Forcing files are a 1979-1999 monthly climatology computed from the |
22 |
|
|
NCEP reanalysis (see pkg/seaice/SEAICE_FFIELDS.h for units and signs) |
23 |
dimitri |
1.4 |
uwindFile = 'u10m.labsea79' # 10-m zonal wind |
24 |
|
|
vwindFile = 'v10m.labsea79' # 10-m meridional wind |
25 |
|
|
atempFile = 'tair.labsea1979' # 2-m air temperature |
26 |
|
|
aqhFile = 'qa.labsea1979' # 2-m specific humidity |
27 |
dimitri |
1.5 |
lwdownFile = 'flo.labsea1979' # downward longwave radiation |
28 |
|
|
swdownFile = 'fsh.labsea1979' # downward shortwave radiation |
29 |
dimitri |
1.4 |
precipFile = 'prate.labsea1979' # precipitation |
30 |
heimbach |
1.2 |
evapFile = 'evap.labsea1979' # evaporation |
31 |
|
|
|
32 |
|
|
|
33 |
dimitri |
1.3 |
Using testscript to test sea-ice code |
34 |
|
|
===================================== |
35 |
|
|
|
36 |
|
|
Running the testscript experiment: |
37 |
|
|
cd verification |
38 |
|
|
./testscript -force lab_sea |
39 |
dimitri |
1.4 |
The default experiment is Experiment 8, below. |
40 |
|
|
It uses pkg/gmredi, pkg/kpp, pkg/seaice, and pkg/exf. |
41 |
dimitri |
1.3 |
|
42 |
|
|
Note that fairly large differences in accuracy occur across different |
43 |
|
|
platforms. For example, testscript comparisons between g77 (Linux) |
44 |
|
|
and f77 (SGI) generated output gives: |
45 |
|
|
|
46 |
dimitri |
1.4 |
T S U V |
47 |
|
|
C D M c m s m s m s m s |
48 |
|
|
n p a R g m m e . m m e . m m e . m m e . |
49 |
|
|
f n k u 2 i a a d i a a d i a a d i a a d |
50 |
|
|
g d e n d n x n . n x n . n x n . n x n . |
51 |
dimitri |
1.3 |
|
52 |
dimitri |
1.4 |
Y Y Y Y 8 10 9 11 10 9 11 13 10 9 8 8 9 8 9 8 9 FAIL lab_sea |
53 |
dimitri |
1.3 |
|
54 |
|
|
|
55 |
|
|
Instructions for generating 1-CPU and 2-CPU executables |
56 |
|
|
======================================================= |
57 |
|
|
|
58 |
|
|
Generating 1-CPU executable: |
59 |
dimitri |
1.4 |
cd verification/lab_sea/input |
60 |
dimitri |
1.3 |
ln -sf ../code/SIZE.h . |
61 |
dimitri |
1.4 |
ln -sf ../code/CPP_EEOPTIONS.h . |
62 |
|
|
ln -sf ../code/CPP_OPTIONS.NOEXF.h CPP_OPTIONS.h |
63 |
|
|
rm -f SEAICE_OPTIONS.h |
64 |
|
|
../../../tools/genmake -makefile |
65 |
|
|
==> on alhena use: |
66 |
|
|
../../../tools/genmake -platform=o2k_noopt -makefile |
67 |
|
|
make clean |
68 |
|
|
make depend |
69 |
|
|
make |
70 |
|
|
mv mitgcmuv mitgcmuv_1 |
71 |
|
|
|
72 |
|
|
Generating 1-CPU executable with SEAICE_EXTERNAL_FORCING: |
73 |
|
|
cd ../../../verification/lab_sea/input |
74 |
|
|
ln -sf ../code/SIZE.h . |
75 |
|
|
ln -sf ../code/CPP_EEOPTIONS.h . |
76 |
dimitri |
1.3 |
ln -sf ../code/CPP_OPTIONS.h . |
77 |
dimitri |
1.4 |
ln -sf ../code/ECCO_CPPOPTIONS.h . |
78 |
|
|
ln -sf ../code/SEAICE_OPTIONS.h . |
79 |
|
|
../../../tools/genmake -makefile |
80 |
|
|
==> on alhena use: |
81 |
|
|
../../../tools/genmake -platform=o2k_noopt -makefile |
82 |
|
|
make clean |
83 |
|
|
make depend |
84 |
|
|
make |
85 |
|
|
mv mitgcmuv mitgcmuv_exf_forcing |
86 |
|
|
|
87 |
dimitri |
1.5 |
1-CPU executable with SEAICE_EXTERNAL_FORCING, without READ_EVAP |
88 |
|
|
cd ../../../verification/lab_sea/input |
89 |
|
|
ln -sf ../code/SIZE.h . |
90 |
|
|
ln -sf ../code/CPP_EEOPTIONS.h . |
91 |
|
|
ln -sf ../code/CPP_OPTIONS.h . |
92 |
|
|
ln -sf ../code/ECCO_CPPOPTIONS.BULK.h ECCO_CPPOPTIONS.h |
93 |
|
|
ln -sf ../code/SEAICE_OPTIONS.h . |
94 |
|
|
../../../tools/genmake -makefile |
95 |
|
|
==> on alhena use: |
96 |
|
|
../../../tools/genmake -platform=o2k_noopt -makefile |
97 |
|
|
make clean |
98 |
|
|
make depend |
99 |
|
|
make |
100 |
|
|
mv mitgcmuv mitgcmuv_compute_evap |
101 |
|
|
|
102 |
dimitri |
1.4 |
Generating 1-CPU executable with SEAICE_EXTERNAL_FLUXES: |
103 |
|
|
cd ../../../verification/lab_sea/input |
104 |
|
|
ln -sf ../code/SIZE.h . |
105 |
dimitri |
1.3 |
ln -sf ../code/CPP_EEOPTIONS.h . |
106 |
dimitri |
1.4 |
ln -sf ../code/CPP_OPTIONS.h . |
107 |
|
|
ln -sf ../code/ECCO_CPPOPTIONS.BULK.h ECCO_CPPOPTIONS.h |
108 |
|
|
ln -sf ../code/SEAICE_OPTIONS.EXF_FLUXES.h SEAICE_OPTIONS.h |
109 |
dimitri |
1.3 |
../../../tools/genmake -makefile |
110 |
|
|
==> on alhena use: |
111 |
|
|
../../../tools/genmake -platform=o2k_noopt -makefile |
112 |
|
|
make clean |
113 |
|
|
make depend |
114 |
|
|
make |
115 |
dimitri |
1.4 |
mv mitgcmuv mitgcmuv_exf_fluxes |
116 |
dimitri |
1.3 |
|
117 |
|
|
Generating 2-CPU executable: |
118 |
|
|
cd ../../../verification/lab_sea/input |
119 |
dimitri |
1.4 |
ln -sf ../code/SIZE.2x1.h SIZE.h |
120 |
|
|
ln -sf ../code/CPP_EEOPTIONS.MPI.h CPP_EEOPTIONS.h |
121 |
|
|
ln -sf ../code/CPP_OPTIONS.NOEXF.h CPP_OPTIONS.h |
122 |
|
|
rm -f SEAICE_OPTIONS.h |
123 |
dimitri |
1.3 |
../../../tools/genmake -mpi -makefile |
124 |
|
|
==> on alhena for comparison purposes use: |
125 |
|
|
../../../tools/genmake -mpi -platform=o2k_noopt -makefile |
126 |
|
|
==> on alhena for fast execution use: |
127 |
|
|
../../../tools/genmake -mpi -platform=o2k -makefile |
128 |
|
|
make clean |
129 |
|
|
make depend |
130 |
|
|
make |
131 |
|
|
mv mitgcmuv mitgcmuv_2x1 |
132 |
|
|
|
133 |
|
|
|
134 |
heimbach |
1.2 |
Instructions for running Experiment 1 |
135 |
|
|
===================================== |
136 |
|
|
|
137 |
dimitri |
1.4 |
This is a 1-cpu, 10-hour integration used to make sure that all |
138 |
|
|
the files are available and that the sea-ice model compiles and |
139 |
|
|
runs. The reading of atmospheric forcing files and the computation |
140 |
|
|
of open-water bulk fluxes are carried out within pkg/seaice. |
141 |
heimbach |
1.2 |
|
142 |
dimitri |
1.3 |
To run Experiment 1: |
143 |
|
|
cd ../../../verification/lab_sea/input |
144 |
dimitri |
1.4 |
ln -sf DATA.10hours data |
145 |
|
|
ln -sf DATA.SEAICE.adi data.seaice |
146 |
dimitri |
1.3 |
mitgcmuv_1 >&! output.txt |
147 |
heimbach |
1.2 |
mkdir exp1 |
148 |
|
|
mv *tave.0000000010.data exp1 |
149 |
|
|
|
150 |
dimitri |
1.3 |
Use matlab script lookat_exp1.m to compare the output |
151 |
dimitri |
1.4 |
of exp1 with that from release1_p10 sea-ice code: |
152 |
dimitri |
1.3 |
cd ../../../verification/lab_sea/matlab |
153 |
heimbach |
1.2 |
matlab |
154 |
|
|
lookat_exp1 |
155 |
|
|
|
156 |
dimitri |
1.3 |
|
157 |
heimbach |
1.2 |
Instructions for running Experiment 2 |
158 |
|
|
===================================== |
159 |
|
|
|
160 |
|
|
This is a 1-cpu test of the LSR solver. The solution is |
161 |
|
|
compared to that of experiment 1, which used the ADI solver. |
162 |
|
|
|
163 |
dimitri |
1.3 |
To run Experiment 2: |
164 |
|
|
cd ../../../verification/lab_sea/input |
165 |
dimitri |
1.4 |
ln -sf DATA.10hours data |
166 |
|
|
ln -sf DATA.SEAICE.lsr data.seaice |
167 |
dimitri |
1.3 |
mitgcmuv_1 >&! output.txt |
168 |
heimbach |
1.2 |
mkdir exp2 |
169 |
|
|
mv *tave.0000000010.data exp2 |
170 |
|
|
|
171 |
dimitri |
1.3 |
Use matlab script lookat_exp2.m to compare |
172 |
heimbach |
1.2 |
the output of exp2 to that of exp1: |
173 |
dimitri |
1.3 |
cd ../../../verification/lab_sea/matlab |
174 |
heimbach |
1.2 |
matlab |
175 |
|
|
lookat_exp2 |
176 |
|
|
|
177 |
dimitri |
1.3 |
|
178 |
heimbach |
1.2 |
Instructions for running Experiment 3 |
179 |
|
|
===================================== |
180 |
|
|
|
181 |
|
|
This is a test of periodic boundary conditions for LSR |
182 |
|
|
and ADI solvers. The domain has a flat bottom and |
183 |
|
|
is periodic both in the x and the y directions. |
184 |
dimitri |
1.3 |
All forcing files are null or constant (u10m = v10m = 5 m/s). |
185 |
heimbach |
1.2 |
|
186 |
dimitri |
1.3 |
To run Experiment 3: |
187 |
|
|
cd ../../../verification/lab_sea/input |
188 |
dimitri |
1.4 |
ln -sf DATA.1hour data |
189 |
|
|
ln -sf DATA.SEAICE.testadi data.seaice |
190 |
dimitri |
1.3 |
mitgcmuv_1 >&! output.txt |
191 |
heimbach |
1.2 |
mkdir exp3a |
192 |
|
|
mv *tave.0000000001.data exp3a |
193 |
dimitri |
1.4 |
ln -sf DATA.SEAICE.testlsr data.seaice |
194 |
dimitri |
1.3 |
mitgcmuv_1 >&! output.txt |
195 |
heimbach |
1.2 |
mkdir exp3b |
196 |
|
|
mv *tave.0000000001.data exp3b |
197 |
|
|
|
198 |
dimitri |
1.3 |
Use matlab script lookat_exp3.m to compare |
199 |
heimbach |
1.2 |
the output of exp2 to that of exp1: |
200 |
dimitri |
1.3 |
cd ../../../verification/lab_sea/matlab |
201 |
heimbach |
1.2 |
matlab |
202 |
|
|
lookat_exp3 |
203 |
|
|
|
204 |
|
|
If the dynamic ice solvers are correct, they should converge to a |
205 |
|
|
constant solution. The tests above show that both the ADI and the LSR |
206 |
|
|
solver are unable to handle periodic domains and therefore that they |
207 |
|
|
cannot be "correctly" parallelized. |
208 |
|
|
|
209 |
dimitri |
1.3 |
|
210 |
heimbach |
1.2 |
Instructions for running Experiment 4 |
211 |
|
|
===================================== |
212 |
|
|
|
213 |
|
|
This is a 2-cpu, 10-hour integration used to test |
214 |
|
|
tile edges for sea-ice dynamic solvers. |
215 |
|
|
|
216 |
dimitri |
1.3 |
To run Experiment 4: |
217 |
|
|
cd ../../../verification/lab_sea/input |
218 |
dimitri |
1.4 |
ln -sf DATA.10hours data |
219 |
|
|
ln -sf DATA.SEAICE.adi data.seaice |
220 |
heimbach |
1.2 |
mpirun -np 2 mitgcmuv_2x1 |
221 |
|
|
mkdir exp4 |
222 |
|
|
mv *tave.0000000010.data exp4 |
223 |
|
|
|
224 |
|
|
Use the matlab script lookat_exp4.m to compare |
225 |
|
|
the 2-cpu output to that of exp1: |
226 |
dimitri |
1.3 |
cd ../../../verification/lab_sea/matlab |
227 |
heimbach |
1.2 |
matlab |
228 |
|
|
lookat_exp4 |
229 |
|
|
|
230 |
|
|
The inaccuracy at the tile boundary can be decreased by increasing |
231 |
|
|
NPSEUDO in data.seaice at the expense of computation time. |
232 |
|
|
Also as sea-ice ages (thickens) the difference at the tiles becomes |
233 |
|
|
increasingly small. For solver pkg/seaice/adi.F a value of NPSEUDO=10 |
234 |
|
|
appears adequate for forward integrations but cannot be used for sea-ice |
235 |
|
|
adjoint model. Work is underway to fix this problem. |
236 |
|
|
|
237 |
|
|
|
238 |
|
|
Instructions for running Experiment 5 |
239 |
|
|
===================================== |
240 |
|
|
|
241 |
|
|
This is a 2-cpu, 2-year, test integration. It illustrates |
242 |
dimitri |
1.3 |
convention used by seaice_get_forcing.F for multi-year forcing. |
243 |
heimbach |
1.2 |
|
244 |
dimitri |
1.3 |
To run Experiment 5: |
245 |
|
|
cd ../../../verification/lab_sea/input |
246 |
dimitri |
1.4 |
ln -sf DATA.2years data |
247 |
|
|
ln -sf DATA.SEAICE.adi data.seaice |
248 |
dimitri |
1.3 |
ln -sf evap.labsea1979 evap.labsea1980 |
249 |
|
|
ln -sf flo.labsea1979 flo.labsea1980 |
250 |
|
|
ln -sf fsh.labsea1979 fsh.labsea1980 |
251 |
heimbach |
1.2 |
ln -sf prate.labsea1979 prate.labsea1980 |
252 |
dimitri |
1.3 |
ln -sf qa.labsea1979 qa.labsea1980 |
253 |
|
|
ln -sf tair.labsea1979 tair.labsea1980 |
254 |
|
|
ln -sf u10m.labsea79 u10m.labsea80 |
255 |
|
|
ln -sf v10m.labsea79 v10m.labsea80 |
256 |
heimbach |
1.2 |
mpirun -np 2 mitgcmuv_2x1 |
257 |
|
|
|
258 |
dimitri |
1.3 |
Use the matlab script lookat_exp5.m to compare |
259 |
heimbach |
1.2 |
the 2-cpu output to SMMR-SSM/I data: |
260 |
dimitri |
1.3 |
cd ../../../verification/lab_sea/matlab |
261 |
heimbach |
1.2 |
matlab |
262 |
|
|
lookat_exp5 |
263 |
|
|
|
264 |
|
|
Disclaimer: |
265 |
|
|
The comparison here is just for fun, not really supposed |
266 |
|
|
to look anything like the data. |
267 |
|
|
Otherwise it would put a lot of people out of business :-) |
268 |
|
|
|
269 |
|
|
Instructions for running Experiment 6 |
270 |
|
|
===================================== |
271 |
|
|
|
272 |
|
|
This is a 1-cpu test of sea-ice thermodynamics (no dynamics). |
273 |
|
|
The solution is compared to that of experiment 1. |
274 |
|
|
|
275 |
dimitri |
1.3 |
To run Experiment 6: |
276 |
|
|
cd ../../../verification/lab_sea/input |
277 |
dimitri |
1.4 |
ln -sf DATA.10hours data |
278 |
|
|
ln -sf DATA.SEAICE.nodynamics data.seaice |
279 |
dimitri |
1.3 |
mitgcmuv_1 >&! output.txt |
280 |
heimbach |
1.2 |
mkdir exp6 |
281 |
|
|
mv *tave.0000000010.data exp6 |
282 |
|
|
|
283 |
|
|
Use the matlab script lookat_exp6.m to compare |
284 |
dimitri |
1.3 |
the output of exp6 to that of exp1: |
285 |
heimbach |
1.2 |
cd ../verification/lab_sea/matlab |
286 |
|
|
matlab |
287 |
|
|
lookat_exp6 |
288 |
dimitri |
1.3 |
|
289 |
|
|
Instructions for running Experiment 7 |
290 |
|
|
===================================== |
291 |
|
|
|
292 |
|
|
This is a 2-cpu, 10-hour integration used to test |
293 |
|
|
tile edges for sea-ice thermodynamics (no dynamics). |
294 |
|
|
2-CPU executable from experiment 4 is required. |
295 |
|
|
|
296 |
|
|
To run Experiment 7: |
297 |
|
|
cd ../../../verification/lab_sea/input |
298 |
dimitri |
1.4 |
ln -sf DATA.10hours data |
299 |
|
|
ln -sf DATA.SEAICE.nodynamics data.seaice |
300 |
dimitri |
1.3 |
mpirun -np 2 mitgcmuv_2x1 |
301 |
|
|
mkdir exp7 |
302 |
|
|
mv *tave.0000000010.data exp7 |
303 |
|
|
|
304 |
|
|
Use the matlab script lookat_exp7.m to compare |
305 |
|
|
the output of exp7 to that of exp6: |
306 |
|
|
cd ../verification/lab_sea/matlab |
307 |
|
|
matlab |
308 |
|
|
lookat_exp7 |
309 |
dimitri |
1.4 |
|
310 |
|
|
|
311 |
|
|
Instructions for running Experiment 8 |
312 |
|
|
===================================== |
313 |
|
|
|
314 |
|
|
This is the default experiment which is executed by |
315 |
|
|
verification/testscript. It is a 1-cpu, 10-hour integration |
316 |
|
|
used to test CPP option SEAICE_EXTERNAL_FORCING. Atmospheric |
317 |
|
|
state is read-in using pkg/exf, rather than pkg/seaice routines. |
318 |
|
|
This experiment uses pkg/gmredi, pkg/kpp, pkg/seaice, and pkg/exf. |
319 |
|
|
|
320 |
|
|
To run Experiment 8: |
321 |
|
|
cd ../../../verification/lab_sea/input |
322 |
|
|
ln -sf DATA.10hours_exf data |
323 |
|
|
ln -sf DATA.SEAICE.adi_exf data.seaice |
324 |
|
|
mitgcmuv_exf_forcing >&! output.txt |
325 |
|
|
mkdir exp8 |
326 |
|
|
mv *tave.0000000010.data exp8 |
327 |
|
|
|
328 |
|
|
Use matlab script lookat_exp8.m to compare |
329 |
|
|
the output of exp8 to that of exp1: |
330 |
|
|
cd ../../../verification/lab_sea/matlab |
331 |
|
|
matlab |
332 |
|
|
lookat_exp8 |
333 |
|
|
|
334 |
|
|
|
335 |
|
|
Instructions for running Experiment 9 |
336 |
|
|
===================================== |
337 |
|
|
|
338 |
|
|
This is a 1-cpu, 10-hour integration used to test CPP option |
339 |
|
|
SEAICE_EXTERNAL_FLUXES. Both the atmospheric state and the |
340 |
|
|
open-water surface fluxes are provided by pkg/exf. |
341 |
|
|
|
342 |
|
|
To run Experiment 9: |
343 |
|
|
cd ../../../verification/lab_sea/input |
344 |
|
|
ln -sf DATA.10hours_exf data |
345 |
|
|
ln -sf DATA.SEAICE.adi_exf data.seaice |
346 |
|
|
mitgcmuv_exf_fluxes >&! output.txt |
347 |
|
|
mkdir exp9 |
348 |
|
|
mv *tave.0000000010.data exp9 |
349 |
|
|
|
350 |
|
|
Use matlab script lookat_exp9.m to compare |
351 |
|
|
the output of exp9 to that of exp8: |
352 |
|
|
cd ../../../verification/lab_sea/matlab |
353 |
|
|
matlab |
354 |
|
|
lookat_exp9 |
355 |
dimitri |
1.5 |
|
356 |
|
|
|
357 |
|
|
Instructions for running Experiment 10 |
358 |
|
|
====================================== |
359 |
|
|
|
360 |
|
|
This is a 1-cpu, 10-hour integration similar to exp8, but |
361 |
|
|
with bulk formula computation of evaporation fields. |
362 |
|
|
|
363 |
|
|
To run Experiment 10: |
364 |
|
|
cd ../../../verification/lab_sea/input |
365 |
|
|
ln -sf DATA.10hours_exf data |
366 |
|
|
ln -sf DATA.SEAICE.adi_exf data.seaice |
367 |
|
|
mitgcmuv_compute_evap >&! output.txt |
368 |
|
|
mkdir exp10 |
369 |
|
|
mv *tave.0000000010.data exp10 |
370 |
|
|
|
371 |
|
|
Use matlab script lookat_exp10.m to compare |
372 |
|
|
the output of exp10 to that of exp8: |
373 |
|
|
cd ../../../verification/lab_sea/matlab |
374 |
|
|
matlab |
375 |
|
|
lookat_exp10 |