pyicon-quickplots

The quickplots are one frequently used application of pyicon to easily visualize data of ICON-O, ICON-A and HAMOCC. While the focus is on ocean plots, there are also many atmosphere plots.

An example for quickplots of the RUBY-0 configuration can be found here

Running the quickplos

Go to:

cd pyicon/tools

Look for a run_qp_driver* script which most closely resembles your configuration and modify it according to your needs.

Below you find a list of possible options which can to modify the quickplots:

run = 'slo1325'
runname = ''
path_data = '/work/mh0287/m211032/Icon/Git_Icon/icon.oes.20200506/experiments/slo1325/'

# --- path to quickplots
path_quickplots = '/mnt/lustre01/pf/zmaw/m300602/pyicon/all_qps/'

# --- set this to True if the simulation is still running
omit_last_file = True

# --- decide which set of plots to do
do_ocean_plots      = True
do_atmosphere_plots = True
do_hamocc_plots     = False

# --- grid information
gname     = 'r2b6_oce_r0004'
lev       = 'L64'
gname_atm = 'r2b4_atm_r0013'
lev_atm   = 'L84'

# --- path to interpolation files
path_grid        = '/mnt/lustre01/work/mh0033/m300602/icon/grids/r2b6_oce_r0004/'
path_grid_atm    = '/mnt/lustre01/work/mh0033/m300602/icon/grids/r2b4_atm_r0013/'
path_ckdtree     = '/mnt/lustre01/work/mh0033/m300602/icon/grids/r2b6_oce_r0004//ckdtree/'
path_ckdtree_atm = '/mnt/lustre01/work/mh0033/m300602/icon/grids/r2b4_atm_r0013//ckdtree/'

# --- grid files and reference data
fpath_tgrid         = '/pool/data/ICON/oes/input/r0004/OceanOnly_Global_IcosSymmetric_0039km_rotatedZ37d_BlackSea_Greenland_modified_srtm30_1min/R2B6_ocean-grid.nc'
fpath_tgrid_atm     = '/pool/data/ICON/grids/public/mpim/0013/icon_grid_0013_R02B04_G.nc'
fpath_ref_data_oce  = '/pool/data/ICON/oes/input/r0004/OceanOnly_Global_IcosSymmetric_0039km_rotatedZ37d_BlackSea_Greenland_modified_srtm30_1min/R2B6L64_initial_state.nc'
fpath_ref_data_atm  = '/mnt/lustre01/work/mh0033/m300602/icon/era/pyicon_prepare_era.nc'
fpath_fx            = '/pool/data/ICON/oes/input/r0004/OceanOnly_Global_IcosSymmetric_0039km_rotatedZ37d_BlackSea_Greenland_modified_srtm30_1min/R2B6L64_fx.nc'

# --- nc file prefixes
oce_def     = '_oce_def'
oce_moc     = '_oce_moc'
oce_mon     = '_oce_mon'
oce_ice     = '_oce_ice'
oce_monthly = '_oce_dbg'

atm_2d      = '_atm_2d_ml'
atm_3d      = '_atm_3d_ml'
atm_mon     = '_atm_mon'

# --- time average information (can be overwritten by qp_driver call)
tave_ints = [['4840-02-01', '4850-01-01']]
# --- decide which data files to take for time series plots
tstep     = '????????????????'
# --- set this to 12 for yearly averages in timeseries plots, set to 0 for no averaging
ave_freq = 12

# --- xarray usage
xr_chunks = None
load_xarray_dset = False
save_data = True
path_nc = '/home/mpim/m300602/work/tmp/test_pyicon_output/'

# --- information for re-gridding
sec_name_30w   = '30W_300pts'
rgrid_name     = 'global_0.3'
rgrid_name_atm = 'global_1.0_era'

verbose = False
do_write_final_config = False

# --- list containing figures which should not be plotted
red_list = []
# --- list containing figures which should be plotted
# (if empty all figures will be plotted)
green_list = []

After specifying your run_qp_driver_name.sh script, you have several options to run the script:

1. Either run it in the current session by

./run_qp_driver_name.sh

2. Or submit it so the queue

sbatch run_qp_driver_name.sh

3. Or use ipython. Therefore, save the above options with your modifications e.g. in pyicon/config_qp/conf_name.py, activate the pyicon environment by

source pyicon/tools/act_pyicon_py39.sh

go to:

cd pyicon/pyicon/quickplots/

then start ipython with:

ipython --pylab

and then run the script by:

%run qp_driver.py ../../config_qp/conf_name.py

In general, option 1 is the easiest, option 2 is best suited for high resolution simulations or long averaging periods, and option 3 is best for debugging.

Visualizing the quickplots

While all figures are saved as png files, these files are also collected on a web page which can either be opened on a local web browser or hosted on a web server.

In both cases, you can easiest copy the all_qps directory specified by path_quickplots in the options above to either your computer or to a web server. Finally, you have to open the index.html file.

Transferring quickplots to swiftbrowser

One time preparation

To transfer the quickplots to DKRZ’s swiftbrowser you have to first create a container. Therfore go to:

https://swiftbrowser.dkrz.de/login/

Log in with the project account under which you want to store the quickplots (e.g. mh0033) and you DKRZ credentials. Then add a new container (click on “+” symbol) with your name / username. Assuming you gave it the name test_container.

Make this container public: Click on “v”-symbol and “Share” Again “v” and “Make public”

Copy the public URL:

https://swiftbrowser.dkrz.de/public/dkrz_07387162e5cd4c81b1376bd7c648bb60/container_test/

Uploading data

To upload the results of the quickplot, please enter the all_qps directory and activate swift:

module unload py-python-swiftclient
module load py-python-swiftclient

If necessary, make a new swift token:

swift-token new
module switch py-python-swiftclient

Then upload the data:

swift upload container_test *

Share domain

Use public URL (see above):

https://swiftbrowser.dkrz.de/public/dkrz_07387162e5cd4c81b1376bd7c648bb60/container_test/index.html

If you click on this link, you the index.html will be opened in your webbrowser. If instead the webbrowser tries to download the index.html then there is something not working with the public access (or you entered the private and not the public domain).