Difference between revisions of "8. Retrieve historical ECMWF fields (ERA40) for FLEXPART"
| Line 1: | Line 1: | ||
'''''Retrieving ECMWF historcal grid data (example for ERA_40):''''' | '''''Retrieving ECMWF historcal grid data (example for ERA_40):''''' | ||
| − | |||
| − | When ssh'ing onto username@eceaccess.ecmwf.int | + | |
| − | Choose: ecgate | + | 1) Make sure your ECtrans setup settings on: https://ecaccess.ecmwf.int/ is valid. |
| + | |||
| + | 2) When ssh'ing onto username@eceaccess.ecmwf.int Choose: ecgate | ||
% Choose c1a (not ecgate) upon login to access the supercomputer facility (HPCE) | % Choose c1a (not ecgate) upon login to access the supercomputer facility (HPCE) | ||
| − | |||
| − | The files that need editing is: | + | 3) The files you need to extract ERA_40can be found e.g. '''/home/ms/no/sb1d/flex_extract_ecgate_V4''', copy this directory. |
| + | |||
| + | |||
| + | 4) The files that need editing is: | ||
| + | '''seq_control.V4''' -executes seq_extract.V4 for the dates you put in. | ||
| + | ''./seq_extract.V4 1990 1990 06 06 20 31 1'' | ||
| + | - start_year end_year start_month end_month start_day end_day 1=complete day(s) | ||
| + | |||
| + | |||
| + | |||
'''flex_ecmwf_91_global_ecgate_ERA-40''' | '''flex_ecmwf_91_global_ecgate_ERA-40''' | ||
- here paths, destination and username needs to be edited | - here paths, destination and username needs to be edited | ||
| − | |||
| − | |||
| − | |||
| − | + | ||
List of files: | List of files: | ||
Revision as of 14:06, 26 October 2012
Retrieving ECMWF historcal grid data (example for ERA_40):
1) Make sure your ECtrans setup settings on: https://ecaccess.ecmwf.int/ is valid.
2) When ssh'ing onto username@eceaccess.ecmwf.int Choose: ecgate % Choose c1a (not ecgate) upon login to access the supercomputer facility (HPCE)
3) The files you need to extract ERA_40can be found e.g. /home/ms/no/sb1d/flex_extract_ecgate_V4, copy this directory.
4) The files that need editing is:
seq_control.V4 -executes seq_extract.V4 for the dates you put in.
./seq_extract.V4 1990 1990 06 06 20 31 1
- start_year end_year start_month end_month start_day end_day 1=complete day(s)
flex_ecmwf_91_global_ecgate_ERA-40 - here paths, destination and username needs to be edited
List of files:
flex_ecmwf_91_finegrid_*
extract_ecmwf_ftpbase
seq_extract.V4*
seq_control.V4*
submit*
1. Edit the times in the file: seq_control.V4* ./seq_extract.V4 1990 1990 06 06 29 29 1
2. In file flex_ecmwf_91_finegrid_* modify the following parameters:
M_GRID 20 (resolution 0.2 deg, for resolution of e.g. 10 deg use 1000)
M_LEFT -3600 (36 W)
M_LOWER 4000 (40 N)
M_UPPER 7600 (76 N)
M_RIGHT 3600 (36 E)
M_FORMAT GRIB2 (GRIB1 or GRIB2)
DESTINATION nina (or another MS association set up on the web interface, e.g flexpart)
ECFSDIR ectmp:/${USER}/GRIM/ (directory name)
SOURCECODE /home/ms/no/sb9/flex_extract_ecgate_V4 (your username)
3. Change all paths to refer to your own user grep sb9 * to list all files containing the path with user sb9 which need to be changed.
4. Copy the template file flex_ecmwf_91_finegrid_* to extract_ecmwf_ftpbase
5. Run the job:
./submit
6. Check your job: llq to see your job in the queue eoj <jobid> provides details on the performance of a running job. (also automatically provided at end of job)
7. Get your files: The retrieved files are automatically transferred to your ecgate account (via ECTRANS) and then to NILU by ftp. (NB: cannot ftp directly from HPCE, but must go via ecgate). If you can’t find your files in your defined NILU-directory, check the directory you specified in flex_ecmwf_91_finegrid_* (ECFSDIR) for where it is temporarily stored on the ECWMF system. Then try ftp it over again using the ftp commands specified previously.
