SWAN - WRF interpolation #424
Description
Hello,
I have a question on the following aspect. When I run a WRF-SWAN coupled simulation I set:
INPGRID WIND REGULAR -20.0 40.0 0 240 120 0.25 0.25 &
NONSTATIONARY 20210723.000000 6 HR 20210731.000000
READINP WIND 1 'swan_wind_d02.dat' 4 0 FREE
for the spin up run, where I run SWAN alone compiled through the COAWST framework. This wind grid corresponds to the ERA5 data with certain lat/lon bounds and a frequency of 6 hours. Then, for the actual run, I comment out the READINP directive so that SWAN and WRF can exchange data.
Now, am I understanding correct that the following statement applies?
When commenting the READINP directive, the actual entries for the INPGRID WIND directive are ignored. These are the spatial and temporal grids. This means that I can leave all the settings that were correct for the spin up run (i.e. ERA5 data spatial and time discretization) and they will be ignored for the coupled run by COAWST. In the coupled run, the wind grid will be the WRF grid (handled internally by the code) and the time does not matter, as data will be exchanged based on the settings of the coupling.in file.
If the above is the case (which I think it is because of some previous cases that I conducted), then why the simulation gives a segmentation fault if the dates after the NONSTATIONARY directive (2nd line) are past the simulation start date for the coupled run? Aren't these dates being ignored? I only realized this by chance as I left these dates from a previous case that was later in time.
I just want to make sure that I understand what the code is doing.
Thanks
Sebastiano