1 |
jscott |
1.1 |
This verifcation run is set up to run on two processers. The atmosphere and |
2 |
|
|
seaice run in the top directory; the ocean runs in a subdirectory rank_1. |
3 |
|
|
This rank_1 directory must be created - with appropriate input files - prior |
4 |
|
|
to running. |
5 |
|
|
|
6 |
|
|
A sample script to execute the mpirun command is run_cpl. |
7 |
|
|
|
8 |
|
|
It is also possible to run the full coupled model on a single processor, |
9 |
|
|
or to distribute the ocean to several nodes, with subdirectories rank_1 |
10 |
|
|
to rank_n, although the atmosphere/seaice cannot be parallelized. |
11 |
|
|
(additional documentation forthcoming) |
12 |
|
|
|
13 |
|
|
----------------------------------------------------------------- |
14 |
|
|
|
15 |
|
|
To build the atmospheric/seaice code: |
16 |
|
|
|
17 |
|
|
The atmospheric code is located at: |
18 |
|
|
http://mitgcm.org/cgi-bin/viewcvs.cgi/MITgcm_contrib/jscott/igsm/ |
19 |
|
|
|
20 |
|
|
Here, there is a source and include directory, and a makefile which builds |
21 |
|
|
an atmospheric library LIBATM.a |
22 |
|
|
|
23 |
|
|
To make the atmice executable, there is a script (which runs a makefile) |
24 |
|
|
in the verification exp. directory exe_atmice. This script first makes a |
25 |
|
|
ocean library LIB3DO.a from the build_atmice *.o files. The MITGCM |
26 |
|
|
options turn off pretty much everything except packages THSICE and ATM2D. |
27 |
|
|
|
28 |
|
|
The ocean code should be built in build_ocn, and the mitgcmuv executable |
29 |
|
|
which results is used directly. There are no special options needed (except |
30 |
|
|
the generic coupling option) in the ocean-only code. |
31 |
|
|
|
32 |
|
|
Pickups (to get you started) for the GISS atmosphere are located at: |
33 |
|
|
http://mitgcm.org/cgi-bin/viewcvs.cgi/MITgcm_contrib/jscott/igsm/pickup/ |
34 |
|
|
(these are fairly large and thus not included in input_atmice) |
35 |
|
|
|
36 |
|
|
Support files for the GISS atmosphere (referenced in name.dat) are |
37 |
|
|
in the tar ball: |
38 |
|
|
http://mitgcm.org/cgi-bin/viewcvs.cgi/MITgcm_contrib/jscott/ |
39 |
|
|
|
40 |
|
|
This verification experiment was run with CLM off, i.e. using a simple bucket |
41 |
|
|
model included in the GISS atmosphere. Support for using CLM with be |
42 |
|
|
forthcoming. |
43 |
|
|
|
44 |
|
|
name.dat notes: |
45 |
|
|
--------------- |
46 |
|
|
These are output files, and must be not present before running: |
47 |
|
|
ocndata4atm = 'ocnd4a0001.01' |
48 |
|
|
plotfl = 'plot0001.01' |
49 |
|
|
nwrfl = 'nwr0001.01' |
50 |
|
|
These are the atmos. pickups, which are continuously overwritten: |
51 |
|
|
file1 = 'pov10001.01.bin', |
52 |
|
|
file2 = 'pov20001.01.bin', |
53 |
|
|
These support files are needed (in addition to some others): |
54 |
|
|
bgrghg_data = '../ATM2D_DATA/INIT/forcing/ghgdata.GISS.2005.dat' |
55 |
|
|
fprratio = '../ATM2D_DATA/CLMDATA/pcp2lnd-data.001x046.txt' |
56 |
|
|
|
57 |
|
|
There are many other options to the 2D atmosphere; further support forthcoming. |
58 |
|
|
|
59 |
|
|
Questions: jscott@mit.edu |