October 29, 2019 at 9:33 pm #6479James BowenParticipant
Hello EE modelers. I’ve got a question on the sediment diagenesis model. Can anyone give recommendations on how to set the variable time step parameters or other parameters to avoid numerical instabilities when running the diagenesis sub-model? We’re using variable time stepping. In our tests we’ve found that the time step parameters that work for WQ are way too long for diagenesis. We’ve seen the Tenkiller Lake model used a max step of 5 sec, w/ an initial step much less than 1 sec. We’ve tried those parameters but we still get negative ammonia concentrations in the summertime in the bottom layers of the water column. We’re assuming it’s an instability problem, as the concentrations seem very sensitive to the time step the model uses. We’d try even shorter steps but the run times are already about 2-3 hours per month of simulation. Are there sediment thickness values to adjust that could help? We’re using the standard set of diagenesis kinetic parameters. Any suggestions on how to get rid of our negative NH4 concentrations would be greatly appreciated.
Jim BowenNovember 15, 2019 at 3:39 pm #6577EEMS TeamKeymaster
The time step depends on the model configuration: grid size, calculation parameters and number of vertical layers you use. It is not always a reliable guide to use the time step from another model to your own model unless they have a very similar configurations. Our experience suggests the initial step time should be quite small to try to run the model and then gradually increase. For the setup for sediment diagenesis model you can look at section 4 at this link:
- You must be logged in to reply to this topic.