This shows you the differences between two versions of the page.
Both sides previous revision Previous revision Next revision | Previous revision Next revision Both sides next revision | ||
initial_conditions:oceanic [2016/01/08 14:37] eexarcho [[[how_to_generate_new_ocean_initial_conditions_|Generation of new ocean initial condition]]] |
initial_conditions:oceanic [2018/10/09 13:23] macosta [Brief description] |
||
---|---|---|---|
Line 1: | Line 1: | ||
- | ==== Brief description | + | ==== Shared in EC-Earth consortium |
+ | Details are here https:// | ||
+ | |||
+ | ==== Brief description ==== | ||
Please note that Ec-Earth2 includes NEMO2.2 and Ec-Earth3 includes NEMO3.3 | Please note that Ec-Earth2 includes NEMO2.2 and Ec-Earth3 includes NEMO3.3 | ||
^ ID^ OCEAN MODEL^ GRID^ FIRST^ LAST^ INTERVAL (MONTH)^ MEMBERS^ Comments^ References | | ^ ID^ OCEAN MODEL^ GRID^ FIRST^ LAST^ INTERVAL (MONTH)^ MEMBERS^ Comments^ References | | ||
- | | s4| NEMO2.2 / NEMO3.2 / NEMO3.3| ORCA1 and ORCAL46| 19580101| 20141201| 02, 05, 08, 11| fc0 - f4| ocean reanalysis - mask corrected for ORCA1 - vertically interpolated and extrapolated for ORCA1L46| Mogensen et al(2011), Balmaseda et al (2012) | | + | | s4| NEMO3.6 | ORCA1L46| 19600101| 20160701| all months | fc0 - f4| ocean reanalysis - mask corrected for ORCA1 - vertically interpolated and extrapolated for ORCA1L75| Mogensen et al(2011), Balmaseda et al (2012) | |
+ | | s4| NEMO2.2 / NEMO3.2 / NEMO3.3| ORCA1 and ORCAL46| 19580101| 20141201| 02, 05, 08, 11| fc0 - f4| ocean reanalysis - mask corrected for ORCA1 - vertically interpolated and extrapolated for ORCA1L46 | ||
| GLORYS2V1| NEMO3.3| ORCA1L46| 19921101| 20091101| 05,11| fc0| Interpolated initial conditions from an ocean reanalysis| Ferry et al (2010) | | | GLORYS2V1| NEMO3.3| ORCA1L46| 19921101| 20091101| 05,11| fc0| Interpolated initial conditions from an ocean reanalysis| Ferry et al (2010) | | ||
| GLORYS2V1| NEMO3.3| ORCA025L46| 19921101| 20091101| 05,11| fc0| Interpolated initial conditions from an ocean reanalysis| Ferry et al (2010) | | | GLORYS2V1| NEMO3.3| ORCA025L46| 19921101| 20091101| 05,11| fc0| Interpolated initial conditions from an ocean reanalysis| Ferry et al (2010) | | ||
Line 17: | Line 21: | ||
| m01w| NEMO3.3| ORCA025L46| 19921101| 20091101| 02, | | m01w| NEMO3.3| ORCA025L46| 19921101| 20091101| 02, | ||
| m01u| NEMO3.3| ORCA025L75| 19921101| 20091101| 02, | | m01u| NEMO3.3| ORCA025L75| 19921101| 20091101| 02, | ||
+ | | official_repro| NEMO3.6| ORCA01L75| 19500101 | 19500101 --- | fc00-fc04| Initial conditions for the official EC-Earth reproducibility test| https:// | ||
| | | | ||
Line 45: | Line 50: | ||
Hao Z., Balmaseda M. A., Mogensen K., 2015, The new eddy-permitting ORAP5 ocean reanalysis: description, | Hao Z., Balmaseda M. A., Mogensen K., 2015, The new eddy-permitting ORAP5 ocean reanalysis: description, | ||
- | ==== How_to_generate_new_ocean_initial_conditions_|Generation of new ocean initial condition | + | ==== How_to_generate_new_ocean_initial_conditions |
- | == ORAP5, ORCA025L75 | + | == ORAS4, ORCA1L42 and ORCA1L75 |
1. Update the ecaccess-certificate-create tool, which gives 7 days of access to ECMWF repositories. | 1. Update the ecaccess-certificate-create tool, which gives 7 days of access to ECMWF repositories. | ||
- | 2. Fetch the data locally by running the script.sh below (in moore or amdhahl with **sbatch script.sh**) | + | 2. clone the project https:// |
- | # | + | 3. run export PYTHONPATH=$PYTHONPATH:$path_to_the_Oce_IC |
- | #SBATCH -n 1 | + | |
- | #SBATCH -t 10-12:00:00 | + | |
- | #SBATCH -J fetchORAP5 | + | |
- | #SBATCH --mem=60000 | + | |
- | #SBATCH -o fetch-orap5-%j.log | + | |
- | + | ||
- | set -evx | + | 4. go into the path_to_the_Oce_IC and run fetch_ORAS4.sh after filling the start dates you want to download from EMCWF and interpolate. |
- | TMPDIR=/ | + | |
- | mkdir -p $TMPDIR | + | |
- | cd $TMPDIR | + | |
- | module load ECAC | + | == ORAP5, ORCA025L75 == |
- | DATASOURCE=" | + | 1. Update |
- | YEARI=2000 # First eyar to download | + | |
- | YEARF=2013 # Last year to download | + | |
- | MEMBER_LST=" | + | |
- | YEAR=$YEARI | + | |
- | workdir=`pwd` | + | 2. git clone https:// |
- | | + | |
- | echo $YEAR | + | 2. Fetch the data locally (for example when in fat nodes) |
- | | + | |
- | mkdir -p / | + | |
- | for mem in $MEMBER_LST ; do | + | |
- | for file in $( ecaccess-file-dir ${DATASOURCE}/ | + | |
- | do | + | |
- | if [[ " | + | |
- | echo $file | + | |
- | if [[ ! -f / | + | |
- | while [[ ! -f / | + | |
- | do | + | |
- | ecaccess-file-get -progress ${DATASOURCE}/ | + | |
- | mv $file / | + | |
- | done | + | |
- | fi | + | |
- | fi | + | |
- | done | + | |
- | done | + | |
- | YEAR=$((YEAR+1)) | + | |
- | done | + | |
+ | cd cpg_tools/ | ||
+ | sbatch dl_restarts.sh | ||
+ | | ||
+ | 3. Copy the data using the correct naming, in the repository of the initial conditions. | ||
+ | | ||
+ | - Pay attention that for initial conditions for 1st May and 1st November we use the **closest available dates** in the original data, which **are not always the same for each year**. | ||
- | 3. Copy the data using the correct naming, in the repository of the initial conditions. **Pay attention that for initial conditions for 1st May and 1st November we use the closest available dates in the original data, which are not always the same for each year. Also, we check the size in the original data, because some dates have half size, so they do not have all the variables we need for interpolation, for example the restart | + | - Also, we check the size in the original data, because some dates **have half size**, which means that they do not have all the variables we need for interpolation. For example, the restart |
ll / | ll / | ||
| | ||
-rw-r--r-- 1 pbretonn Earth 9184196770 Oct 7 04:31 / | -rw-r--r-- 1 pbretonn Earth 9184196770 Oct 7 04:31 / | ||
- | -rw-r--r-- 1 pbretonn Earth 9181022277 Oct 7 04:54 / | + | |
- | -rw-r--r-- 1 pbretonn Earth 9183757715 Oct 7 05:17 / | + | -rw-r--r-- 1 pbretonn Earth 9183757715 Oct 7 05:17 / |
- | -rw-r--r-- 1 pbretonn Earth 9181268805 Oct 7 05:45 / | + | -rw-r--r-- 1 pbretonn Earth 9181268805 Oct 7 05:45 / |
- | -rw-r--r-- 1 pbretonn Earth 9178049931 Oct 7 06:13 / | + | -rw-r--r-- 1 pbretonn Earth 9178049931 Oct 7 06:13 / |
- | -rw-r--r-- 1 pbretonn Earth 9179559914 Oct 7 06:35 / | + | -rw-r--r-- 1 pbretonn Earth 9179559914 Oct 7 06:35 / |
- | -rw-r--r-- 1 pbretonn Earth 4927336766 Oct 7 06:50 / | + | -rw-r--r-- 1 pbretonn Earth 4927336766 Oct 7 06:50 / |
- | After the manual checking, we run the following script: | + | After the manual checking, we run the following script: |
+ | cd cpg_tools/ | ||
+ | sbatch copy_orap5.sh | ||
+ | |||
+ | |||
+ | |||
+ | == ORAP5, ORCA1L46 == | ||
+ | We use a software called **SOSIE** [[http:// | ||
+ | cd cpg_tools/ | ||
+ | sbatch launchscript.sh | ||
- | dir=/ | + | |
- | outdir=/ | + | == GLOSEA5, |
- | curdir=`pwd` | + | |
- | + | ||
- | for year in {2005..2007} ; do | + | |
- | ##May start dates | + | |
- | echo $year | + | |
- | dirin=${dir}/ | + | |
- | if [ -e $dirin | + | |
- | cd $dirin | + | |
- | for file in $( ls | + | |
- | inname=$(echo $file | cut -f1 -d" | + | |
- | ofile=ORAP5_fc0_${year}0430_restart.nc.gz | + | |
- | if [ ! -e ${outdir}/ | + | |
- | cd $curdir | + | |
- | gunzip -c | + | |
- | tar xfv ${inname}.tar | + | |
- | gzip -c ${inname}.nc > ${outdir}/ | + | |
- | rm -rf ${curdir}/ | + | |
- | fi | + | |
- | done | + | |
- | fi | + | |
+ | We need to extrapolate the files to the mask we use locally. We do the following separately per year (here as example for 19940501, in the fat nodes) | ||
+ | cd cpg_tools/ | ||
+ | | ||
- | # November start dates | + | == GLOSEA5, ORCA1L46 == |
- | echo $year | + | |
- | dirin=${dir}/ | + | |
- | if [ -e $dirin | + | |
- | cd $dirin | + | |
- | for file in $( ls | + | |
- | inname=$(echo $file | cut -f1 -d" | + | |
- | ofile=ORAP5_fc0_${year}1031_restart.nc.gz | + | |
- | if [ ! -e ${outdir}/ | + | |
- | cd $curdir | + | |
- | gunzip -c | + | |
- | tar xfv ${inname}.tar | + | |
- | gzip -c ${inname}.nc > ${outdir}/ | + | |
- | rm -rf ${curdir}/ | + | |
- | fi | + | |
- | done | + | |
- | fi | + | |
- | done | + | |
+ | Similarly, as in ORAP5 ORCA1L46, we need to interpolate by using a software called **SOSIE** [[http:// | ||
+ | |