Coupled surface-hydrology application with data assimilation.
NASA Land Coupler Driver and Mediator: The NASA Land Coupler driver and mediator have been designed in collaboration by NOAA, NASA, and NCAR. The custom driver couples LIS and WRF-Hydro within the NUOPC framework. Coupled system configuration is controlled via settings in the nlc.runconfig file. The custom mediator couples an in memory ensemble to multiple instances of a component, which allows users to run a coupled ensemble entirely in memory. Each component is individually configured based on the needs of the component (see documentation for each component).
Land Information System (LIS):
LIS is a land surface modeling framework devloped by NASA.
Documentation: https://lis.gsfc.nasa.gov/
Repository: https://github.com/NASA-LIS/LISF
WRF-Hydro:
WRF-Hydro is a hydrometeorological and hydrologic modeling system
developed by the research applications labratory at NCAR.
Documentation: https://ral.ucar.edu/projects/wrf_hydro/overview
Repository: https://github.com/NCAR/wrf_hydro_nwm_public
We are using ssh URLs for cloning submodules, so you will need to ensure that you have added your public SSH key from your platform (Discover and/or Cheyenne) to your GitHub account. Help is available here: https://help.github.com/articles/connecting-to-github-with-ssh/
These have been tested on Discover (NASA) and Cheyenne (NCAR).
Clone the repository including the submodules.
$ git clone --recursive [email protected]:NESII/lishydro.git <directory>
LIS and WRF-Hydro are included as git submodules. The following commands provide an alternative to using the --recursive git clone option.
$ cd <directory> # go into the cloned repository
$ git submodule init
$ git submodule update
System requires ESMF v8.0.1 or later.
Configure Submodules
Configure LIS and WRFHYDRO using the configure.sh script. For more information
run ./configure.sh --help.
$ ./configure.sh
Build Application
Build NASA Land Coupler (NLC) using the build.sh script. For more information
run ./build.sh --help. NLC.exe will be installed into src/driver.
$ ./build.sh
NOTE: Runs are currently only supported on Discover and Cheyenne.
Use cases are preconfigured using settings in usecases directory and preprocessed data. Input data, including configuration, domain, and parameter files, is stored on Discover and Cheyenne
Current supported use cases:
| Use Cases | Description |
|---|---|
| coupled_irene.noahmp.nldas2 | Coupled LIS-WRFHYDRO Hurricane Irene NoahMP v3.6 |
| coupled_irene.noah.nldas2 | Coupled LIS-WRFHYDRO Hurricane Irene Noah v3.6 |
| coupled_irene_noMed.noah.nldas2 | Directly Coupled LIS-WRFHYDRO Hurricane Irene (No MED) |
| coupled_tar.noahmp_v4.0.1_cold.nldas2 | Coupled LIS-WRFHYDRO Tar River NoahMP v4.0.1 |
| coupled_tuolumne.ens002.noahmp.nldas2 | Coupled LIS-WRFHYDRO Tuolumne NoahMP v4.0.1 Ensemble |
| coupled_tuolumne.ens020.noahmp_v4.0.1_da.nldas2 | Coupled LIS-WRFHYDRO Tuolumne NoahMP v4.0.1 Ensemble with DA |
| coupled_tuolumne.noahmp.nldas2 | Coupled LIS-WRFHYDRO Tuolumne NoahMP v3.6 |
| coupled_tuolumne.noahmp_v4.0.1_cold.nldas2 | Coupled LIS-WRFHYDRO Tuolumne NoahMP v4.0.1 |
| coupled_tuolumne.noahmp_v4.0.1_da.nldas2 | Coupled LIS-WRFHYDRO Tuolumne NoahMP v4.0.1 with DA |
| coupled_tuolumne.noah.nldas2 | Coupled LIS-WRFHYDRO Tuolumne Noah v3.6 |
| coupled_tuolumne_noMed.noah.nldas2 | Directly Coupled LIS-WRFHYDRO Tuolumne (No MED) |
| sbys_tuolumne_hyd.ens002.ldas | Uncoupled LIS-WRFHYDRO Tuolumne Ensemble (2 member) |
| stdalone_frontrange_hyd.ldas | Standalone WRFHYDRO Frontrange |
| stdalone_irene_lnd.nldas2 | Standalone LIS Hurricane Irene |
| stdalone_tuolumne_hyd.ldas | Standalone WRFHYDRO Tuolumne |
Setup Run Directory
Preconfigured use cases can be set up using the setuprun.sh script.
For more information run ./setuprun.sh --help. For a list of use cases
run ./setuprun.sh --list-usecases.
$ ./setuprun.sh USECASE
This will create and populate a run directory in run/<usecase>.
Submit Run
The batch script run.sh can be modified if needed, for example to change the number of MPI tasks or the project number.
$ cd run/<usecase>
$ sbatch run.sh # Discover - SLURM
$ qsub run.sh # Cheyenne - PBS
This will submit the run to the batch queue. Output will appear in the same directory.