| 32 |
|
|
| 33 |
CONTENTS |
CONTENTS |
| 34 |
|
|
|
|
|
|
|
|
| 35 |
code: modifications to MITgcm to run in cyclindrical coordinates. At |
code: modifications to MITgcm to run in cyclindrical coordinates. At |
| 36 |
the time of this writing, it requires checkpoint48 to compile and run. |
the time of this writing, it requires checkpoint48 to compile and run. |
| 37 |
|
|
| 38 |
|
input: initialization and input files for MITgcm |
| 39 |
|
|
| 40 |
filter: source and data files to run the data assimilation and Kalman |
filter: source and data files to run the data assimilation and Kalman |
| 41 |
filter. |
filter. |
| 42 |
|
|
| 43 |
|
utils: assorted Matlab m-files |
| 44 |
|
|
| 45 |
|
EnKF: empty |
| 46 |
|
|
| 47 |
|
run: empty |
| 48 |
|
|
| 49 |
|
|
| 50 |
|
COMPILING AND RUNNING THE MODEL |
| 51 |
|
|
| 52 |
|
The following series of commands creates the MITgcm executable. |
| 53 |
|
|
| 54 |
|
From the MITgcm root: |
| 55 |
|
|
| 56 |
|
cd verification/osse |
| 57 |
|
mkdir build # directory name is arbitrary |
| 58 |
|
cd build |
| 59 |
|
../../../tools/genmake -mods=../code # creates makefile |
| 60 |
|
make depend # contructs Fortran code from model codebase |
| 61 |
|
make # makes executable mitgcmuv |
| 62 |
|
|
| 63 |
|
Try a test run, while redirecting the standard output to a file. The |
| 64 |
|
input directory contains files that should be present in the working |
| 65 |
|
directory when you run mitgcmuv. It is recommended that you copy the |
| 66 |
|
input contents to a different directory for running -- mitgcmuv does |
| 67 |
|
not change them, but they can be lost in the multitudes of output |
| 68 |
|
files. |
| 69 |
|
|
| 70 |
|
The data assimilation code is made by typing ``make'' in the filter |
| 71 |
|
directory. |
| 72 |
|
|
| 73 |
|
|