Running both probabilistic and deterministic segments of a model together

Answered

Follow

Carly Schultz

• May 13, 2024 20:37

My GoldSim water balance model includes a calibration portion using historical data and a predictive portion using forecasted data. Is it possible to shorten the run time of the model by running the calibration portion (4 years) deterministically and the predictive portion probabilistically?

For more background, the client is only interested in the predictive portion of the model. However, the calibration portion affects many parameters, and it would be a lot of work to manually reset these parameters to the values they are at the end of the calibration. If we could run the model deterministically for these years only, it would shorten the run time, reduce the chance of error in copying over values and save a lot of work. 

0

• Official comment

  

  Rick Kossik

  • May 13, 2024 21:47

  I don't understand what you mean by running one portion deterministically and one probabilistically.  A run is either deterministic (1 realization) or probabilistic (multiple realizations).  Typically, you would run a deterministic calibration, and then use those parameters and change the simulation settings to do the predictive simulation. Why can't you do that?

  Comment actions Permalink

Please sign in to leave a comment.