Dear colleagues,
We are running the ROMS model in a harbour environment, with cell sizes of about 40x40m, and a maximum depth of about 50m. We have tried several cases, with different combinations of barotropic and baroclinic timesteps. Even though the Courant number for each case is smaller than 0.5, thus assuring model stability, in some cases the model runs fine, and in others it blows-up. Is there any criteria we should take into account to estimate the best relation between both timesteps (ndtfast)?
Best regards,
Manel
LIM/UPC
Thecnical Unversity of Catalonia, Barcelona
combinations of barotropic and baroclinic timesteps
-
Manel_Grifoll
- Posts: 5
- Joined: Fri Mar 23, 2007 12:35 pm
- Location: Laboratori d'Engineyria Maritima/Universitat Politècnica de Catalunya (LIM/UPC)
Perhaps it would help if you knew what was causing the blow-ups. Can you plot fields just before the explosion? Do you know where the trouble is happening? I have had some luck with putting stations in trouble spots and recording to the stations file every timestep so that nothing is missed. Have you tried shorter timesteps? The WET_DRY option? What exactly is happening?