I am running a coupled model configuration (OpenIFS + NEMO 4.2 eORCA05.L75). I keep running into the error:
stp_ctl: |ssh| > 20 m or |U| > 10 m/s or S <= 0 or S >= 100 or NaN en
counter in the tests
kt 1 |ssh| max Infinity at i j 618 149 found in 8 MPI tasks, spread o
ut among ranks 18 to 84
kt 1 |U| max 110.8 at i j k 564 528 1 MPI rank 81
kt 1 Sal min 0.1000 at i j k 564 527 60 MPI rank 81
kt 1 Sal max 5174. at i j k 564 528 62 MPI rank 81
I have tried several things including:
- recreate the coupling files
- flooding the land with temperature and salinity in the input files
- changing the coupling timestep of the ocean
Although I am trying to run a configuration with ice-shelf melting enabled, the error persists even when I disable cavity melting and/or iceberg calving. I am aware this is a very broad question, but can somebody let me know what else I could look for?
sounds like the barotropic sub-timestepping going loose but I suspect that a forcing issue underlies the crash. Can you check the utau/vtau and other forcing fields of the abort file?
Hello,
I have been running a piControl starting from 1850. I thought I solved the problem when I realized my input files (temperature and salinity that I received from somewhere else) were interpolated to eORCA05.L75 using a mesh_mask that didn’t have ice-shelf cavities. I interpolated those files to match my tmask of the mesh_mask and it worked well, until it produced the same SSH error after about 30 years.
I have attached some images from the output.abort file. Could you please let me know where I could be going wrong?
Thanks for the images. I think I can see small scales in taux/y of the ACC. It might be coming from the interactions with sea-ice. Do you have sea ice fields as well in the abort file?
Here’s what the sea-ice cover field looks like in the abort file.
sorry if I cannot tell from the figures due to absence of axes and tics, but does that confirm that the crash is related to interactions with the sea ice? If yes, do you have also the ice velocities?
