Blow up error due to max density

[caca329]

I'm really new to ROMS and now I've been hitting so many blow up error when running ROMS. I want to get the data of whole year of 2020, but I can only get until 28th Jan. For month Jan 2020 I already had 3 blow up, first at 22nd and second at 25th, I get to continue the running from those blow up after I smooth the bathymetry of my grid, but then for the third blow up even if I smooth the bathymetry but the blow up keep coming up, so I hope if anyone can help me on this error so that I can run the model well without blowing up, hereby attached the log I get after I run the model in txt.file. Btw I also tried to visualize the output data for those 28 days of temperature and salinity but seems like the values looks unrealistic too cuz for example the range of temperature start from 1.5 to 111.5 degree celsius.

[ebpbenke]

I had the same problem in the past. To smooth the bathymetry, I now used the gradient rx0=0.1 in GridBuilder. Good luck.

[jcwarner]

it is important to identify where the blow up is occurring. is it at the open boundary? is it at a location where the land/sea mask meet? etc.
The location of the high temp (111) might be where the instability is occurring.

[caca329]

I've noticed the problem was my boundary and climatology files were being extracted wrongly cuz I accidentally used only the condition of first day in 2020 to get to output of 366 days for boundary and climatology files. After I corrected this and get the correct boundary and climatology files, however it can't even run the model with the error blowup shown below:
MAIN: Abnormal termination: BLOWUP.
REASON: KEchar = NaN, PEchar = NaN
Note: The following floating-point exceptions are signalling: IEEE_INVALID_FLAG IEEE_OVERFLOW_FLAG IEEE_DENORMAL
ERROR STOP

So i went to check my boundary files and i saw all NaN values in zeta, ubar, and vbar. Any suggestion to solve this?

[jcwarner]

i am not sure which tools you are using to create bndry file, but you can not have nans in any forcing or bndry files.
suggest you try different tools or different source of data.

[caca329]

i am not sure which tools you are using to create bndry file, but you can not have nans in any forcing or bndry files.
suggest you try different tools or different source of data.

I was using pyroms to get the forcing from MERRA-2 and boundary, clm, and ic files from HYCOM