This page shows how to set up an offline land analysis experiment based on CY48R1 by prepIFS (including snow depth analysis).


Start prepIFS

You can start prepIFS by the command below.

prepIFS -d /perm/dako/git/ifs-defaults -b dako_CY48R1.0_offline-snow-analysis.IFS-2462

'ifs-defaults' is set to the branch based on CY48R1.0 and related to the following Jira tickets.

  • IFS-2462 - Getting issue details... STATUS
  • IFS-2149 - Getting issue details... STATUS

Create new experiment 

At first, new experiment can be created as follows. Experiment type must be set to 'surface'.

Change executing host

You will change executing host to run your experiment on Atos.

Change each setting for your own experiment

Next, you can customize each setting. These are examples to set up an experiment including soil moisture and snow depth analysis at a resolution of TCo319. Checked boxes indicate that I modified them for my experiment. 

  • 'DESCRIPTION' must be filled as you like.

  • You can change branches including your own changes.

  • You can set your experiment period. 'LERA' can be used for open-loop (without data assimilation) only.

  • You can change settings for offline land surface model. In this case, horizontal resolution and time step are changed to the same as the seasonal prediction system.
  • In the default settings, atmospheric forcing is retrieved from ERA5.

  • 'FREQ_PP_OSM' must be set to 1 if you set up an offline land analysis experiment.

  • ’ACCOUNT’ must be filled depending on your account. In this case, it is set to 'ecrdasca' which corresponds to coupled assimilation team.

  • In case you don't need to build double precision modules, you can set 'LIFS_BUILD_DP' to Off.

  • If you run an experiment at higher resolution, you might have to change NPES_FC and THREADS_FC. NPES_FC=1 is enough for TCo319 experiment.
  • If you would like to change horizontal resolution, the following page will be useful: Old documentation of Offline ECLand simulations

  • You can change default precision settings at 'Numerical precision'. 

  • You can change settings for offline land analysis here.
    • LSEKF_SM: activate soil moisture analysis
    • LSEKF_SD: activate snow depth analysis
    • LMONTHLY_ARCH: In order to avoid high load on MARS by archiving output frequently, it is better to set it On. This is used for offline land analysis experiment only.
    • LSAVE_SEKF: To save the SEKF output in ECFS, including increments, innovations and Kalman gains, select LSAVE_SEKF to "ON"

Lastly, you will click 'Submit experiment type' and set up your own experiment. You can check the experiment status by ecFlow.