WAM Cycle 6

WAM is a third-generation wave model that describes the evolution of the wave spectrum by solving the wave energy transfer equation. WAM predicts wave direction spectra and properties and can be linked to a number of other models.

A repository of the source code with modifications for Taltech HPC, can be found here.

How to cite

The WAM Model—A Third Generation Ocean Wave Prediction Model DOI: https://doi.org/10.1175/1520-0485(1988)018<1775:TWMTGO>2.0.CO;2



Quickstart


Short jobs & one core jobs

  1. To run your first calculations, start a session on a node:

     srun -t 2:0:0 --pty bash
    
  2. Enter the following commands to set up environment and working directory:

     module load green/all
     module load WAM
    
     export WORK=$HOME/newwamtest
    
     mkdir --parent ${WORK}/tempsg
    
     cd ${WORK}/tempsg
     cp ${WAMDIR}/const/TOPOCAT.DAT .
     cp ${WAMDIR}/const/Coarse_Grid/ARD/Preproc_User .
     preproc
    
     cp ${WAMDIR}/const/Coarse_Grid/ARD/WAM_User .
     cp ${WAMDIR}/const/WIND_INPUT.DAT .
    
  3. Run WAM

     mpirun wam
    
     cp ${WAMDIR}/const/Coarse_Grid/ARD/Grid_User .
     pgrid
    
     cp ${WAMDIR}/const/Coarse_Grid/ARD/Time_User .
     ptime
    
     cp ${WAMDIR}/const/Coarse_Grid/ARD/Spectra_User .
     pspec
    
     cp ${WAMDIR}/const/Coarse_Grid/ARD/Time_User_S .
     ptime_S
    
     cp ${WAMDIR}/const/Coarse_Grid/ARD/nlnetcdf NETCDF_User
     mpirun pnetcdf
    
  4. Adapt the WORKDIR, LOGDIR and output directories to your needs!

  5. If calculations are going normally, you should have the following files in your $WORK directory:

     BLS19780907060000  Grid_Prot              OUT19780907060000  Time_Prot_S
     BLS19780908060000  Grid_User              OUT19780908060000  Time_User
     C0119780906060000  Grid_info_COARSE_GRID  Preproc_Prot       Time_User_S
     C0119780907060000  MAP19780906060000      Preproc_User       WAM_Prot
     C0119780908060000  MAP19780907060000      Spectra_Prot       WAM_User
     C0219780906060000  MAP19780908060000      Spectra_User       WAVE1978090606.nc
     C0219780907060000  NETCDF_User            TOPOCAT.DAT        WIND_INPUT.DAT
     C0219780908060000  OUT19780906060000      Time_Prot          pnetcdf_prot
    
  6. To visualise results you can open the WAVE*.nc file for example in Octave or Matlab.

     pkg load netcdf
     netcdf_open('WAVE1978090606.nc')
     ncdisp('WAVE1978090606.nc')
     hmax = ncread("WAVE1978090606.nc",'hmax_st')
     %plot field at timestep 12
     pcolor(hmax(:,:,12))
    
     hs_swell = ncread("WAVE1978090606.nc",'hs_swell')
     %plot timeseries at position 20 20
     plot(hs_swell(20,20,:))
    

Long & parallel jobs

Longer running and parallel jobs are better submitted as batch jobs using an sbatch script wam.slurm:

Click to expand
#!/bin/bash
#SBATCH --job-name=WAM-testrun
#SBATCH --mem-per-cpu=1GB
#SBATCH --nodes=1
#SBATCH --ntasks=2
#SBATCH --cpus-per-task=1
#SBATCH -t 0-01:0:00
#SBATCH --partition=green-ib
#SBATCH --no-requeue

module load green
module load WAM

export WORK=$HOME/newwamtest

mkdir --parent ${WORK}/tempsg
mkdir --parent ${WORK}/work

cd ${WORK}/tempsg
cp ${WAMDIR}/const/TOPOCAT.DAT .
cp ${WAMDIR}/const/Coarse_Grid/ARD/Preproc_User .
preproc

cp ${WAMDIR}/const/Coarse_Grid/ARD/WAM_User .
cp ${WAMDIR}/const/WIND_INPUT.DAT .

mpirun wam

cp ${WAMDIR}/const/Coarse_Grid/ARD/Grid_User .
pgrid

cp ${WAMDIR}/const/Coarse_Grid/ARD/Time_User .
ptime

cp ${WAMDIR}/const/Coarse_Grid/ARD/Spectra_User .
pspec

cp ${WAMDIR}/const/Coarse_Grid/ARD/Time_User_S .
ptime_S

cp ${WAMDIR}/const/Coarse_Grid/ARD/nlnetcdf NETCDF_User
mpirun pnetcdf



WAM long version


Starting calculations

If job is small it can be run as an interactive session:

srun -t 2:0:0 --pty bash

If calculation is long or needs several cores, it is better to gather all needed commands in one wam.slurm batch script and submit it by command:

sbatch wam.slurm

Preparation

  1. Firstly, user needs to load proper environment by commands:

     module load green
     module load WAM
    
  2. After it is needed to determine working directory and go into it

     export WORK=$HOME/newwamtest
     mkdir --parent ${WORK}/tempsg
     cd ${WORK}/tempsg
    
  3. Copy into working directory all needed data, for example:

     cp ${WAMDIR}/const/TOPOCAT.DAT .
     cp ${WAMDIR}/const/Coarse_Grid/ARD/Preproc_User .
     preproc
    
     cp ${WAMDIR}/const/Coarse_Grid/ARD/WAM_User .
     cp ${WAMDIR}/const/WIND_INPUT.DAT .
    

Running WAM

WAM calculations can be started by command WAM.

mpirun wam

To run calculations normally, such parameters as grid (pgrid), time (ptime), spectra (pspec) time step (ptime_S) and (pnetcdf) should be defined or copy from example:

cp ${WAMDIR}/const/Coarse_Grid/ARD/Grid_User .
pgrid

cp ${WAMDIR}/const/Coarse_Grid/ARD/Time_User .
ptime

cp ${WAMDIR}/const/Coarse_Grid/ARD/Spectra_User .
pspec

cp ${WAMDIR}/const/Coarse_Grid/ARD/Time_User_S .
ptime_S

cp ${WAMDIR}/const/Coarse_Grid/ARD/nlnetcdf NETCDF_User
mpirun pnetcdf

If calculations are going normally, you should have the following files in your $WORK directory:

BLS19780907060000  Grid_Prot              OUT19780907060000  Time_Prot_S
BLS19780908060000  Grid_User              OUT19780908060000  Time_User
C0119780906060000  Grid_info_COARSE_GRID  Preproc_Prot       Time_User_S
C0119780907060000  MAP19780906060000      Preproc_User       WAM_Prot
C0119780908060000  MAP19780907060000      Spectra_Prot       WAM_User
C0219780906060000  MAP19780908060000      Spectra_User       WAVE1978090606.nc
C0219780907060000  NETCDF_User            TOPOCAT.DAT        WIND_INPUT.DAT
C0219780908060000  OUT19780906060000      Time_Prot          pnetcdf_prot

Visualisation

To visualise results you can open the WAVE*.nc file for example in Octave or Matlab

	pkg load netcdf
	netcdf_open('WAVE1978090606.nc')
	ncdisp('WAVE1978090606.nc')
	hmax = ncread("WAVE1978090606.nc",'hmax_st')
	%plot field at timestep 12
	pcolor(hmax(:,:,12))

	hs_swell = ncread("WAVE1978090606.nc",'hs_swell')
	%plot timeseries at position 20 20
	plot(hs_swell(20,20,:))