Hello,
I am running a PEST inversion for solute transport with HYDRUS. After several iterations, PEST inevitably runs a simulation which produces NaN values.
I have altering the number of spatial and temporal nodes several times to prevent the non-convergence issue which produces the NaN error with limited success.
I am wondering if there is a way to supress NaN value output (i.e. output negative values or 0) instead of NaN values when there is numerical non-convergence.
Thank you, John