SWAN encountered 'Unexpected end of file while reading DUMMY', produces constant SWH

[kenlagamayo]

Good day!

I am not sure if this is the right forum to ask this question, but I would like to take my chances, and maybe someone knows how to fix the problem.

I am trying to run a 14-month simulation in a South Korean estuary by coupling ROMS and SWAN. When the wind is provided using the line below, the SWAN model stops calculating after the 7th month and only provides the same wave data to ROMS. This resulted in constant significant wave heights starting from the 8th month until the end. On the other hand, ROMS did not encounter any problem as tides were well-correlated with the observed data from the start until the end.

swan.in code
READINP WIND 1 'Data/SWAN4/Geum_Wind_swan3.dat' 5 0 FREE

When this line is commented out, the SWAN model runs until the end of the 14th month, but the significant wave heights were underestimated. Thus, it is concluded that the wind is essential in the estuary for wave propagation.

I tried several methods to fix the problem but to no avail. I varied the parameters in this particular line, but the SWAN still stops midway. I also double-checked if there is an error in my wind file. I believe that they are correct since if I did a model run starting from the 8th using the restart files of SWAN and ROMS, the model gives me reasonable significant wave heights until the end. This has been my band-aid solution to have a complete run of my model, but I think this is not the proper way to solve it.

Just today, we tried to look at the PRINT01 file. It states that there is an error starting the 8th month. The error states 'Unexpected end of file while reading DUMMY.' However, I have no idea on which file to look to be able to fix this.

Any help or direction on where to look for this source of error to have a complete 14-month run will be highly appreciated.

Thank you very much.
Kenneth

[jcwarner]

i can give this a try.
suggest you look in the swan manual for the syntax on the wind lines.
There are more lines for the wind, something like:

&& KEYWORD TO CREATE WIND GRID &&
INPGRID WIND REGULAR -105 10 0 220 160 0.25 0.25 &
NONSTATIONARY 20121028.000000 6 HR 20121031.000000
READINP WIND 1 'Projects/Sandy/swan_GFS_Oct2012.dat' 4 0 FREE

-105 10 is start lon lat
0 is grid angle
220 160 are number of points (1 less than the number of cells)
0.25 0.25 are the dx dy in lon lat
then the time begin, dt, time end.
this is for windx and windy
the 4 0 free means idla 4 (see swan manual, start lower left go across bottom ...)
IF yo do not have enough data in your .dat wind file, then swan will run out of data.
For this applicati0n we have: 3 days/6hr = 13 time instances.
13*221*161*2=925106 lines

[kenlagamayo]

Thank you John for your reply.

I forgot to mention that I am using a curvilinear grid, so my code looks like this:

INPGRID WIND CURVILINEAR 0 0 226 180 EXC 9999 &
NONSTATIONARY 20170101.000000 1 HR 20180228.230000
READINP WIND 1 'Data/SWAN4/Geum_Wind_swan3.dat' 5 0 FREE

I tried to check my wind file again. It has 4,640,712 lines and 181 columns.
The lines represent 424 days, 24 hours, 227 cells, and 1 more extra hour of data. There are also two lines allotted for the date; thus, the total is 4,640,712.
I attached my wind file for reference.

[jcwarner]

not sure what you mean '2 more lines for the date." please read the swan manual for the proper format.
you need to have data for each cell
424*24*227*181*2=836202624
or something close to taht. you have about half of that, so maybe that is why it stops short.