Youtube: https://m.youtube.com/watch?v=PHXLP1FrZmw&pp=ygUHd3czdG9vbA%3D%3D
WAVEWATCH III visualization and run tool (WW3Tool) is a pre-processing workflow tool for the WAVEWATCH III model. It helps you complete the basic WAVEWATCH III workflow.
This tool includes:
- Multiple forcing fields: wind (ERA5, CFSR, CCMP), currents (Copernicus), water level (Copernicus), sea ice (Copernicus), with automatic fixes (latitude ordering, time fixes, variable fixes).
- gridgen rectangular grid generation, JIGSAW generate triangular unstructured grid,supports up to two nested grids (Python version, no MATLAB dependency,Two-Way Nesting,same forcing field).
- Regional runs, 2D spectrum point runs, and track runs.
- Slurm script configuration.
- Automatic configuration for ww3_grid.nml, ww3_prnc.nml, ww3_shel.nml, ww3_ounf.nml, ww3_multi.nml, etc. (compute precision, output precision, time range, 2D spectrum points, track runs, partition output, forcing setup).
- Wave height plots, wave height videos, contour plots, 2D spectrum plots, JASON3 satellite tracks, 2D spectrum plots.
This software can run on Windows, Linux, and Mac, and is almost entirely composed of Python.
You must install WAVEWATCH III yourself on local or server environments. This tool does not provide an installer,please read https://github.com/ZxyGch/WAVEWATCH-III-INSTALL-TUTORIAL
My undergraduate major was not oceanography, and I am currently a first-year graduate student. The only WAVEWATCH III methods I know are these. If you have any more ideas, please contact me at atomgoto@gmail.com or issue.
If you find this tool useful, please give me a star 🥳
python run.pyIf any packages fail to install or are missing, please install them manually.
Python ≥ 3.8 is supported.
Tested on:
- Windows 11
- Ubuntu 24
- macOS 15
WAVEWATCH III does not need to be installed locally. Local runs are optional and not recommended.
For actual runs, make sure the server has:
-
WAVEWATCH III
-
Slurm workload manager
You must choose or create a work directory when the app starts. This step is required and cannot be skipped.
The default new work directory name is the current time. Up to 3 recent work directories are shown.
A work directory is just a folder for files generated during runs, such as grid files, forcing files, and WAVEWATCH III configuration files.
The default work directory is WW3Tool/workSpace. You can change it on the settings page.
Wind fields can use data from ERA5, CFSR, and CCMP.
For other forcing fields, I only tested Copernicus currents, water level, and sea ice.
I have pre-prepared a few forcing files under WW3Tool/public/forcing. You can select them directly (for testing only).
WAVEWATCH requires latitude to be ascending. ERA5 wind data is descending by default, so the app checks and automatically reverses if needed.
CFSR wind variables are also auto-fixed to match WW3 requirements.
Copernicus forcing timestamps are auto-fixed during this process as well.
Forcing files are automatically copied (or cut, configurable in settings) to the current work directory and renamed to wind.nc, current.nc, level.nc, ice.nc. The log panel shows file info.
Usually only wind forcing is needed. The software does not allow using other forcing fields without wind.
If a single file contains multiple forcing fields, the related buttons are auto-filled. The file is named like current_level.nc in the work directory to indicate the contained fields.
The reference_data is mandatory; otherwise, grid files cannot be generated.
The bundle includes bathymetry (e.g. gebco, etopo1, etopo2) and coastline data, extracted under reference_data.
When the app reports missing reference_data, click Download: it will automatically fetch four parts from the WW3Tool GitHub release (part_aa–part_ad), merges them into reference_data.zip , and extracts to WW3-Grid-Generator/reference_data (or your configured path).
If that is slow or fails, use the mirrors: OneDrive or BaiduNetdisk.
For manual downloads, extract the contents to the directory: WW3Tool/WW3-Grid-Generator/reference_data.
Run the app, choose a domain from wind.nc, and click Generate Grid. This calls WW3Tool/WW3-Grid-Generator to generate grid files into the work directory.
Smaller DX/DY yields higher accuracy because DX/DY is the grid spacing.
Four files are produced in the work directory: grid.bot, grid.obst, grid.nml (WW3 grid description from write_ww3meta), grid.mask_nobound.
Choose type: Nested grid.
For nested meshes, we use two-way nesting, rectangular meshes, and a spherical coordinate system. Other meshes are not currently supported (if you know how to use other meshes, please let me know and I will add them to the software).
We define a nested grid shrink factor on the settings page. The default is 1.1x.
When setting the outer grid, it expands from the inner grid by 1.1x.
When setting the inner grid, it shrinks from the outer grid by 1.1x.
Nested grid generation runs twice: once for the outer grid and once for the inner grid.
In nested mode, two folders are created under the work directory: coarse (outer) and fine (inner).
When coarse and fine exist, opening the directory automatically switches to nested mode. This affects later operations, so if coarse/fine or other grid files already exist, switching the grid type is disabled.
To avoid repeated computation, each grid is cached under WW3Tool/WW3-Grid-Generator/cache.
A cache key based on parameters becomes the folder name. When generating, the cache is checked first and reused if available.
Each cache folder includes params. json:
{
"cache_key": "c161115dfd8bde7b30fd01826a3c292ada7835df377a81b9ee59f73acc28328b",
"source_dir": "/Users/zxy/ocean/WW3Tool/workSpace/2026-01-11_23-18-38",
"parameters": {
"dx": 0.05,
"dy": 0.05,
"lon_range": [
110.0,
130.0
],
"lat_range": [
10.0,
30.0
],
"ref_dir": "/Users/zxy/ocean/WW3Tool/WW3-Grid-Generator/reference_data",
"bathymetry": "GEBCO",
"coastline_precision": "full"
}
}For global grids, the wind field file's range is generally not a standard global range, and we'll be prompted to generate a global grid.
Currently, global simulations still have issues, and I know very little about this. I hope to get your help.
My current approach to global grids involves modifying IS_GLOBAL = 0 to 1 in the grid.nml file of MATLAB Gridgen,
and changing 'RECT' T 'NONE' to 'RECT' T 'SMPL' in the generated grid.meta.
This way, in the fourth step of parameter confirmation, we'll detect 'RECT' T 'SMPL' in grid.meta. I've also modified GRID%CLOS = 'SMPL' in ww3_grid.nml.
&GRID_NML
GRID%NAME = 'I Love ZhangXiaoPeng'
GRID%NML = 'namelists.nml'
GRID%TYPE = 'RECT'
GRID%COORD = 'SPHE'
GRID%CLOS = 'SMPL'
GRID%ZLIM = -0.1
GRID%DMIN = 2.5
/
We originally used the gridgen version from Ifremer. The NOAA version is at https://github.com/NOAA-EMC/gridgen.
Because they are MATLAB-based, it was cumbersome, so I converted it to a Python version. The default is now Python.
If you really want to use MATLAB Gridgen (not recommended), you can switch it in settings and configure the MATLAB path.
Gridgen settings allow you to change many parameters:
-
GRIDGEN version: default is Python (much faster than MATLAB, no dependency).
-
Default DX/DY for regular grids is the X/Y spacing. Larger DX/DY means a smaller grid and lower precision.
-
Nested grid shrink factor: controls auto scaling for inner/outer grids.
-
Bathymetry data: gebco, etop1, etop2. We typically use gebco (highest precision), then etop2, then etop1.
-
Coastline boundary precision: typically the highest precision.
The three run modes have similar computational cost, but outputs differ. Spectrum point mode and track mode look like they compute only a few points, but they still compute the entire domain.
Regular regional mode corresponds to ww3_ounf output.
Spectrum point mode adds ww3_ounp.
Track mode corresponds to ww3_trnc.
Regular output mode.
You can pick points from the map to open a window.
Click points on the map. The blue dashed rectangle shows the grid extent; points must be within it. Then click Confirm and add points.
When confirming parameters in step 4, a points.list file is generated in the work directory.
117 18 '0'
126 21 '1'
127 20 '2'
115 15 '3'
128 14 '4'
126 18 '5'The three columns in points.list are longitude, latitude, and point name. If a work directory contains points.list, the app switches to spectrum point mode and loads the points.
After running WW3, you can get ww3.2025_spec.nc in the plotting page.
Then you can plot 2D spectra.
Similar to spectrum point mode, but adds a time column. Step 4 generates a file: track_i.ww3, with the format below:
WAVEWATCH III TRACK LOCATIONS DATA
20250103 000000 115.4 19.7 0
20250103 000000 127.6 19.7 1
20250103 000000 127.6 15.6 2Finally, ww3_trnc outputs a ww3.2025 file.
The example below uses track mode, which produces more logs than regular regional mode.
✅ Copied 10 public/ww3 files to current work directory
✅ Successfully synced grid.meta parameters to ww3_grid.nml
✅ Modified spectral partition output scheme in ww3_shel and ww3_ounf
✅ Updated server.sh: -J=202501, -p=CPU6240R, -n=48, -N=1, MPI_NPROCS=48, CASENAME=202501, ST=ST2
✅ Updated ww3_ounf.nml: FIELD%TIMESTART=20250103, FIELD%TIMESTRIDE=3600 seconds
✅ Updated ww3_shel.nml: DOMAIN%START=20250103, DOMAIN%STOP=20250105, DATE%FIELD%STRIDE=1800s
✅ Modified ww3_prnc.nml: FORCING%TIMESTART = '20250103 000000', FORCING%TIMESTOP = '20250105 235959'
✅ Copied and modified ww3_prnc_current.nml: FORCING%FIELD%CURRENTS = T
✅ Copied and modified ww3_prnc_level.nml: FORCING%FIELD%WATER_LEVELS = T
✅ Copied and modified ww3_prnc_ice.nml: FORCING%FIELD%ICE_CONC = T
✅ Copied and modified ww3_prnc_ice1.nml: FORCING%FIELD%ICE_PARAM1 = T
✅ Modified ww3_shel.nml: Updated INPUT%FORCING%* settings
✅ Generated track_i.ww3 file
✅ Modified ww3_shel.nml: Added DATE%TRACK (Track Mode)
✅ Modified ww3_trnc.nml: TRACK%TIMESTART = '20250103 000000', TRACK%TIMESTRIDE = '3600'
First, all files under WW3Tool/bin/public/ww3 are copied into the current work directory.
✅ Copied 9 public/ww3 files to current work directory
Next:
✅ Successfully synced grid.meta parameters to ww3_grid.nml
We convert the grid.meta section:
'RECT' T 'NONE'
401 401
3.00 3.00 60.00
110.0000 10.0000 1.00Into the ww3_grid.nml section:
&RECT_NML
RECT%NX = 401
RECT%NY = 401
RECT%SX = 3.000000
RECT%SY = 3.000000
RECT%SF = 60.000000
RECT%X0 = 110.000000
RECT%Y0 = 10.000000
RECT%SF = 60.000000
/
Then modify the partition output plan:
✅ Modified spectral partition output scheme in ww3_shel and ww3_ounfTYPE%FIELD%LIST in ww3_shel.nml:
&OUTPUT_TYPE_NML
TYPE%FIELD%LIST = 'HS DIR FP T02 WND PHS PTP PDIR PWS PNR TWS'
/FIELD%LIST in ww3_ounf.nml:
&FIELD_NML
FIELD%TIMESTART = '20250103 000000'
FIELD%TIMESTRIDE = '3600'
FIELD%LIST = 'HS DIR FP T02 WND PHS PTP PDIR PWS PNR TWS'
FIELD%PARTITION = '0 1'
FIELD%TYPE = 4
/Then update server.sh:
✅ Updated server.sh: -J=202501, -p=CPU6240R, -n=48, -N=1, MPI_NPROCS=48, CASENAME=202501, ST=ST2
#SBATCH -J 202501
#SBATCH -p CPU6240R
#SBATCH -n 48
#SBATCH -N 1
#SBATCH --time=2880:00:00
#wavewatch3--ST2
export PATH=/public/home/weiyl001/software/wavewatch3/model/exe/exe:$PATH
MPI_NPROCS=48
CASENAME=202501✅ Updated ww3_ounf.nml: FIELD%TIMESTART=20250103, FIELD%TIMESTRIDE=3600 seconds
Then update ww3_ounf.nml and find:
&FIELD_NML
FIELD%TIMESTART = '20250103 000000'
FIELD%TIMESTRIDE = '3600'
FIELD%LIST = 'HS LM T02 T0M1 T01 FP DIR SPR DP PHS PTP PLP PDIR PSPR PWS TWS PNR'
FIELD%PARTITION = '0 1'
FIELD%TYPE = 4
/FIELD%TIMESTART is the start time, and FIELD%TIMESTRIDE is the output stride.
✅ Updated ww3_shel.nml: DOMAIN%START=20250103, DOMAIN%STOP=20250105, DATE%FIELD%STRIDE=1800s
Update ww3_shel.nml:
&DOMAIN_NML
DOMAIN%START = '20250103 000000'
DOMAIN%STOP = '20250105 235959'
/
&OUTPUT_DATE_NML
DATE%FIELD = '20250103 000000' '1800' '20250105 235959'
DATE%RESTART = '20250103 000000' '86400' '20250105 235959'
/The dates define the start/stop range, and '1800' in DATE%FIELD is the time step.
Then update the time range in ww3_prnc.nml:
&FORCING_NML
FORCING%TIMESTART = '19000101 000000'
FORCING%TIMESTOP = '29001231 000000'
FORCING%FIELD%WINDS = T
FORCING%FIELD%CURRENTS = F
FORCING%FIELD%WATER_LEVELS = F
FORCING%FIELD%ICE_CONC = F
FORCING%FIELD%ICE_PARAM1 = F
FORCING%GRID%LATLON = T
/Based on the selected forcing fields, we generate ww3_prnc_current.nml and ww3_prnc_level.nml. For ice, concentration and thickness become ww3_prnc_ice.nml and ww3_prnc_ice1.nml.
We toggle forcing switches based on the selected fields. Each forcing switch can only enable one, but ww3_prnc is used multiple times later.
We also update forcing file names and variables:
&FILE_NML
FILE%FILENAME = 'wind.nc'
FILE%LONGITUDE = 'longitude'
FILE%LATITUDE = 'latitude'
FILE%VAR(1) = 'u10'
FILE%VAR(2) = 'v10'
/
✅ Modified ww3_shel.nml: Updated INPUT%FORCING%* settings
Then update ww3_shel.nml based on selected forcing fields:
&INPUT_NML
INPUT%FORCING%WINDS = 'T'
INPUT%FORCING%WATER_LEVELS = 'T'
INPUT%FORCING%CURRENTS = 'T'
INPUT%FORCING%ICE_CONC = 'T'
INPUT%FORCING%ICE_PARAM1 = 'T'
/
Based on the current track points or spectrum point list, we generate:
✅ Generated track_i.ww3 file
✅ Modified ww3_shel.nml: Added DATE%TRACK (Track Mode)
We also add to ww3_shel.nml:
&OUTPUT_DATE_NML
DATE%FIELD = '20250103 000000' '1800' '20250105 235959'
DATE%TRACK = '20250103 000000' '1800' '20250103 000000'
DATE%RESTART = '20250103 000000' '86400' '20250105 235959'
/
✅ Modified ww3_trnc.nml: TRACK%TIMESTART = '20250103 000000', TRACK%TIMESTRIDE = '3600'
In track mode we also update ww3_trnc.nml:
&TRACK_NML
TRACK%TIMESTART = '20250103 000000'
TRACK%TIMESTRIDE = '3600'
TRACK%TIMESPLIT = 8
/
✅ Updated namelists.nml: Changed E3D from 0 to 1 (Spectral Point Calculation Mode)
In 2D spectrum point mode, we also modify namelists.nml:
&OUTS E3D = 0 /We first generate nested grids, creating coarse and fine under the work directory, then select 2D spectrum mode.
======================================================================
🔄 【Work Directory】Starting to process public files...
✅ Copied server.sh, ww3_multi.nml to work directory: /Users/zxy/ocean/WW3Tool/workSpace/nest
✅ Updated server.sh: -J=202501, -p=CPU6240R, -n=48, -N=1, MPI_NPROCS=48, CASENAME=202501, ST=ST2
✅ Updated ww3_multi.nml: Start=20250103, End=20250105, Compute precision=1800s，Forcing Fields=Wind Field、Current Field、Level Field、Ice Field、Ice Thickness，Compute Resources: coarse=0.50, fine=0.50，ALLTYPE%POINT%FILE = './fine/points.list'，ALLDATE%POINT = '20250103 000000' '1800' '20250105 235959'，ALLTYPE%FIELD%LIST = 'HS DIR FP T02 WND PHS PTP PDIR PWS PNR TWS' (spectral partition output)
======================================================================
🔄 【Outer Grid】Starting to process outer grid...
✅ Copied 9 public/ww3 files to current work directory
✅ Modified spectral partition output scheme in ww3_shel and ww3_ounf
✅ Successfully synced grid.meta parameters to ww3_grid.nml
✅ Updated ww3_ounf.nml: FIELD%TIMESTART=20250103, FIELD%TIMESTRIDE=3600 seconds
✅ Updated ww3_shel.nml (spectral point calculation mode): Start=20250103, End=20250105, Compute step=1800s，Added TYPE%POINT%FILE = 'points.list'，Added DATE%POINT and DATE%BOUNDARY
✅ Modified ww3_prnc.nml: FORCING%FIELD%WINDS = T, FILE%FILENAME = '../wind.nc'
✅ Modified ww3_prnc.nml: FORCING%TIMESTART = '20250103 000000', FORCING%TIMESTOP = '20250105 235959'
✅ Copied and modified ww3_prnc_current.nml: FORCING%FIELD%CURRENTS = T
✅ Copied and modified ww3_prnc_level.nml: FORCING%FIELD%WATER_LEVELS = T
✅ Copied and modified ww3_prnc_ice.nml: FORCING%FIELD%ICE_CONC = T
✅ Copied and modified ww3_prnc_ice1.nml: FORCING%FIELD%ICE_PARAM1 = T
✅ Modified ww3_shel.nml: Updated INPUT%FORCING%* settings
✅ Updated namelists.nml: Changed E3D from 0 to 1 (Spectral Point Calculation Mode)
✅ Created points.list file with 4 points
✅ Updated ww3_ounp.nml: POINT%TIMESTART = '20250103 000000', POINT%TIMESTRIDE = '3600' (Spectral Point Calculation Mode)
======================================================================
🔄 【Inner Grid】Starting to process inner grid...
✅ Copied 9 public/ww3 files to current work directory
✅ Successfully synced grid.meta parameters to ww3_grid.nml
✅ Updated ww3_ounf.nml: FIELD%TIMESTART=20250103, FIELD%TIMESTRIDE=3600 seconds
✅ Updated ww3_shel.nml (spectral point calculation mode): Start=20250103, End=20250105, Compute step=1800s，Added TYPE%POINT%FILE = 'points.list'，Added DATE%POINT and DATE%BOUNDARY
✅ Modified ww3_prnc.nml: FORCING%FIELD%WINDS = T, FILE%FILENAME = '../wind.nc'
✅ Modified ww3_prnc.nml: FORCING%TIMESTART = '20250103 000000', FORCING%TIMESTOP = '20250105 235959'
✅ Copied and modified ww3_prnc_current.nml: FORCING%FIELD%CURRENTS = T
✅ Copied and modified ww3_prnc_level.nml: FORCING%FIELD%WATER_LEVELS = T
✅ Copied and modified ww3_prnc_ice.nml: FORCING%FIELD%ICE_CONC = T
✅ Copied and modified ww3_prnc_ice1.nml: FORCING%FIELD%ICE_PARAM1 = T
✅ Modified ww3_shel.nml: Updated INPUT%FORCING%* settings
✅ Updated namelists.nml: Changed E3D from 0 to 1 (Spectral Point Calculation Mode)
✅ Created points.list file with 4 points
✅ Updated ww3_ounp.nml: POINT%TIMESTART = '20250103 000000', POINT%TIMESTRIDE = '3600' (Spectral Point Calculation Mode)
We confirm parameters in step 4 and check the log output:
✅ Copied server.sh, ww3_multi.nml to work directory: /Users/zxy/ocean/WW3Tool/workSpace/nest
We first copy server.sh and ww3_multi.nml from WW3Tool/public/ww3 to the work directory.
We use ww3_multi.nml to modify start time, precision, and forcing fields. This is similar to ww3_shel.nml.
&INPUT_GRID_NML
INPUT(1)%NAME = 'wind'
INPUT(1)%FORCING%WINDS = T
INPUT(2)%NAME = 'current'
INPUT(2)%FORCING%CURRENTS = T
INPUT(3)%NAME = 'level'
INPUT(3)%FORCING%WATER_LEVELS = T
INPUT(4)%NAME = 'ice'
INPUT(4)%FORCING%ICE_CONC = T
INPUT(5)%NAME = 'ice1'
INPUT(5)%FORCING%ICE_PARAM1 = T
/
&MODEL_GRID_NML
MODEL(1)%NAME = 'coarse'
MODEL(1)%FORCING%WINDS = 'native'
MODEL(1)%FORCING%CURRENTS = 'native'
MODEL(1)%FORCING%WATER_LEVELS = 'native'
MODEL(1)%FORCING%ICE_CONC = 'native'
MODEL(1)%FORCING%ICE_PARAM1 = 'native'
MODEL(1)%RESOURCE = 1 1 0.00 0.35 F
MODEL(2)%NAME = 'fine'
MODEL(2)%FORCING%WINDS = 'native'
MODEL(2)%FORCING%CURRENTS = 'native'
MODEL(2)%FORCING%WATER_LEVELS = 'native'
MODEL(2)%FORCING%ICE_CONC = 'native'
MODEL(2)%FORCING%ICE_PARAM1 = 'native'
MODEL(2)%RESOURCE = 2 1 0.35 1.00 F
/Note that MODEL (2)%FORCING%WINDS = 'native'. Here, native means enabled, and no means disabled.
MODEL (1)%RESOURCE and MODEL (2)%RESOURCE represent the resource allocation ratio.
Other logs are straightforward; we process the inner and outer grids in the same way as regular grids.
Notably, we set FILE%FILENAME = '../wind.nc' in ww3_prnc.nml to avoid duplicating forcing files and instead share references.
Local runs execute local.sh.
If you choose local execution, make sure WAVEWATCH III is configured locally and choose the bin directory containing the programs below.
First, you need to configure the ssh account and password. On the settings page, we find the server configuration option.
Pay attention to the default server path. This is the path where your server stores the working directory.
Click to connect to the server. After the connection is successful, a CPU usage ranking will be displayed first. This list will be refreshed every second.
If you submit the calculation task to Slurm in step 6, the task queue will also be displayed.
To view the task queue, execute squeue -l on the server.
Uploading the working directory to the server means uploading the current working directory to the server working directory. This is configured on the settings page.
Submitting a computing task means executing the server.sh script on the server. If it runs successfully (all instructions run normally), a success. log will be generated in the server working directory, containing all execution logs. If it fails, a fail. log will be generated, which also contains all execution logs.
Checking whether it has completed is detecting whether there is success. log or fail. log
Clearing the folder means clearing the current server working directory folder.
Downloading the results to the local computer will automatically download all ww3*nc files. If it is nested grid mode, only the result files in fine will be downloaded.
Downloading the log file means downloading success. log or fail. log
When opening a work directory, the app auto-detects converted forcing files and fills the related buttons.
It auto-reads grid extent and precision to fill step 2, detects coarse/fine folders, and switches to nested mode.
It auto-detects points.list to switch to point output mode, and track_i.ww3 to switch to track mode.
It auto-reads server.sh Slurm parameters to fill step 4, and detects ww3_shel.nml precision, time range, and partition scheme.
Most settings are saved automatically, except the partition output scheme.
Run mode only controls whether some UI elements are shown on the home page. It does not affect core logic.
For example, when local mode is selected, Slurm parameters are hidden.
Forcing selection auto-links: if a file contains multiple forcing fields, other buttons are auto-filled.
File handling follows your forcing file preference: copy or move.
The JASON data path is used for plotting, e.g., comparing simulated wave heights with JASON 3 observations.
WW3 configuration corresponds to step 4 defaults. Confirming parameters updates ww3_shel.nml and ww3_multi.nml precision, and ww3_ounf.nml, ww3_ounp.nml, ww3_trnc.nml output precision.
File splitting is the TIMESPLIT value in ww3_ounf.nml, ww3_ounp.nml, ww3_trnc.nml. If you compute a 3‑month range, monthly or yearly splits make sense; daily split creates a file per day.
Spectrum parameters, numerical integration time step, and nearshore settings are in ww3_grid.nml. Changes here update both WW3Tool and the current work directory's ww3_grid.nml (if present).
Partition output is configured in ww3_shel.nml, ww3_ounf, and ww3_ounp.
Run on the server:
sinfoThis shows CPU information (if Slurm is configured).
Then go to Settings → Slurm parameters → CPU management, and set your server CPUs.
Fill in your SSH account and work directory path.
Every work directory is uploaded to this work directory path.
This is for managing different compiled WAVEWATCH versions. Just fill in their paths.
https://cds.climate.copernicus.eu/datasets/reanalysis-era5-single-levels?tab=download
The images below show ERA5 download steps. You need to register an account first. Use a real English name, not random letters.
http://tds.hycom.org/thredds/catalog/datasets/force/ncep_cfsv2/netcdf/catalog.html
Find cfsv2-sec2_2025_01hr_uv-10m.nc and note the uv-10m suffix.
To download global full-year data, click:
HTTPServer://tds.hycom.org/thredds/fileServer/datasets/force/ncep_cfsv2/netcdf/cfsv2-sec2_2025_01hr_uv-10m.nc
To download a specific region and time range, click NetcdfSubset: //ncss.hycom.org/thredds/ncss/grid/datasets/force/ncep_cfsv2/netcdf/cfsv2-sec2_2025_01hr_uv-10m.nc
After opening, select wndewd and wndnwd on the left, then choose Output Format: netCDF.
If you cannot input lat/lon, uncheck Disable horizontal subsetting.
Then click Time range, enter the time range, and submit.
https://data.remss.com/ccmp/v03.1/
This is straightforward. Just download directly.
Choose Variables below. If you don't need water level, uncheck Sea surface height above geoid.
Then enter the range and time and click DOWNLOAD.
You can download sea ice and currents.
Sea ice includes Sea ice area fraction and Sea ice thickness.
ftp://ftp-oceans.ncei.noaa.gov/nodc/data/jason3-gdr/gdr
This software is licensed under the GNU General Public License v3.0 (GPL-3.0), as it is based on and links against GPLv3-licensed frameworks.