I have a question about continuation or merged runs in EE7. When we have an interrupted model we have the option of saving it in order to assign the last hydrodynamic conditions of the first model as initial conditions for a second model. However, the model comparison results between the full run and the stopped/saved run show distinct differences for many model parameters. I think, when we use this option the initial settings of the new model is not so good. This is not a question regarding the RESTART FILE. The restart file option is works well.
If we have this situation:
Base: is base model.
Run_01: Run for a certain time.
Run_02: Continuation run from Run_01 to the end.
Merged: merged the Run_01 and Run_02 results.
Then comparing Base and the Merged there are differences in salinity and temperature for example.
This is not really a problem as EE is not creating a “hot” start in this case. When we use EE to create the startup files, they contain the IC fields for all constituents and bathymetry. However, the velocities and internal shears are not saved. Therefore, EE is creating a “warm” start, not a true “hot” start. We would like to eventually add the ability to create a “hot” start using this feature but that is not how it is now.
So, for example a model that is a tidal system will have various energy waves moving though the domain. Starting the run using velocities equal to zero everywhere will cause an initial condition shock to the system resulting in differences in the continuous run versus the merged run for at least a week or maybe more. After an initial “ramp up” period the two models should be nearly the same.