Tutorial for nested ww3 runs for ww3 6.07

[JMSnyder1980]

Are there any tutorials or examples on how to do a full nested run off ww3 first starting with a broad nest that has boundary conditions for a smaller nest? I cannot find any such thing in the regtests and would like an example of how to do this so I can apply to a global nest going into a smaller regional nest.

[EmilioEchevarria]

There's a couple of ways to accomplish this, and the manual discusses them, see here: https://github.com/NOAA-EMC/WW3/blob/develop/manual/app/nest.tex

One option is to run your global grid specifying that you want boundary conditions for a regional (nested) grid. You define that in ww3_grid.inp (see here for example). Also make sure you output boundary data (see here). Once this is set up, run your global grid and it should generate a file nest1.ww3, rename that to nest.ww3 and put it in the directory where you're going to run your regional grid.

Another option, use spectral output data (it could be from your global simulation) and use ww3_bounc to generate the file nest.ww3 to be used in your regional run.

Hope that helps.
Emilio

[JMSnyder1980]

I am having problems applying the ww3_grid.inp to a global bathemetry file. How do I modify the ww3_grid.inp to be able to be compatible with the global bathemetry files? Can you guide me in thiis or help modify the file to work successfully with the bathemetry files I am sending you?
global_1p0_mask_new.dat.txt
global_1p0_obst_new.dat.txt
global_1p0_bathy_new.dat.txt

[EmilioEchevarria]

I will try to have a look later and run ww3_grid with these files. But please tell me, which error are you getting?

[JMSnyder1980]

This is the error I am getting:

Output boundary points :

*** WAVEWATCH III ERROR IN W3GRID :
ERROR IN READING FROM INPUT FILE
IOSTAT = 0

[EmilioEchevarria]

Ok I just saw your other question, the ww3_grid.inp you're using seems to be taken from here, which is a test. Naturally this won't work in a different setting with a different grid. The idea is that you define ww3_grid.inp based on your specific grid and specific configuration you want to use (for example, if you are using regular grids, unstructured grids, SMC grids, etc. there are options for each of these in ww3_grid.inp and you need to define them based on your case). This can be taken as a set of instructions that you can follow to generate your own input file correctly.

The error you got indicates that there is an issue with your ww3_grid.inp file and WW3 failed trying to read it. If you keep getting it, it means you need to change your input file to make it have the correct format that WW3 expects.

I haven't tested your files, but I imagine you would need something like this maybe:

$ -------------------------------------------------------------------- $
  'Name of your grid'
$
  1.100000 0.041180 25 24 0.
$
  F T T T F T 
$
  1800  600  900  30
$
   &MISC FLAGTR = 4, CICE0 = 0.25, CICEN = 0.75/
   END OF NAMELISTS
  'RECT' T 'NONE'
  361	 159 
  1.0	 1.0  1.00 
  0   -78   1.00
  -0.10   1.0  20  1.0  1  1 '(....)'  'NAME' '/path/to/your/depth/file' 
  30  1.0  1  1  '(....)'  'NAME'  '/path/to/your/obst/file'
  40            1  1  '(....)'  'NAME'  '/path/to/your/mask/file'
$
$ Nest boundary conditions
$
nest_xW nest_yN 0.0 -DY -NY
nest_xW nest_yS DX 0.0 NX
nest_xE nest_yS 0.0 DY NY
nest_xE nest_yN -DX 0.0 NX
$
$
      0.      0.    0.    0.       0
$
$ ----------------- 
$ End of input file
$ ----------------- 

But again, you need to work it out for your own grid and configuration.

[CarstenHansen]

The namelists version of the ww3_grid input, ww3_grid.nml, contains more lately updated instructions. It may for example be instructive to read the section here about how to 'Define the timesteps parameterization'. I don't remember if this is reflected in the manual.

Judging from your global bathymetry data, it seems that you need to set a scaling (division) factor of value 0.1 rather than the value 1.0 in @EmilioEchevarria's example.

[abayhaqi]

Hi @EmilioEchevarria I have successfully produced the two nest files (nest1.ww3 and nest2.ww3) following your answer. However, I am just wondering how to rename it to a nest.ww3. Then It could be two files with the same name (?)

[mickaelaccensi]

you have to reduce your cfl time step because the distances in km for each grid cell are going smaller when are closer to the pole for a same resolution in degree. Is your resolution at 0.5deg ?

[abayhaqi]

Hi @mickaelaccensi ,
My resolution is 1 degree

[mickaelaccensi]

so your DX at the northern latitude (lat=89.5) for a spatial resolution at 1deg is :
dx=1cos(89.53.14/180)185260
dx=1057m

To estimate your CFL timestep Tcfl, the formula is : (I've considered that your lower frequency is 0.0373Hz)
Tcfl = dx / (g / (Fmin4Pi) ) with the constants Pi=3,14 and g=9.8m/s²;
Tcfl = 1057 / (9.8 / (0.037343.14) )
Tcfl = 50s

so your dtcfl should be set a bit lower, like :
dtcfl=45s
and dtmax, 3 times dtcfl, so :
dtmax=135s

As you can see, having a global grid going really close to the pole will require a very model timestep, that's why we usually reduce the global grid around latitudes 80N and then we create a curvilinear grid for the North pole, which has regular dx and dy in km units.

[abayhaqi]

Hi @mickaelaccensi
Thank you so much for your explanation. It helped me.
Now my result is fine with reasonable values.

[mickaelaccensi]

ok great ! so I close the ticket

[nYifeiLee]

Dear @EmilioEchevarria ，
I went through a series of preparations for WW3 self-nesting based on your conversation with @abayhaqi .Now my large nest.ww3 file has been generated. But while ww3_shel in a small area, an error occurred in the program.I used python code to change the "1" to "2" of the boundary in the mask file of the small area.But the error was still reported.Can you help me see where my problem is appearing?Below is my input file and mask file.
Large scale grid:
ww3_grid.txt
ww3_shel.txt
rect.mask.txt
Sarge scale grid:
ww3_grid.txt
ww3_shel.txt
rect1.mask.txt
ERROR: