I'm assuming by error you mean the mass balance or percent discrepancy error reported by MODFLOW in the output file. Check the output file for warnings and errors. You can reduce your convergence criteria in the solver options to reduce this error. PEST will not help reduce this error. You need to figure out why your model is unstable first. Review your conceptual model to see that your model boundary conditions are representative of your understanding of the groundwater system. Good luck.

Thank you response Roemer. how ever when i say error i meant to say this. in the attached image below you will realize the PEST parameter estimation error is very high as i am trying to calibrate my model. on the othervside the model has some observation wells within the calibrated target where as others do not. my question is in reference to how to get all if not possible most within the target. Alternative approaches to calibrate my model

When the error model is too high, there may be an error in the data input. So, in the first step, check the data and conceptual model. The PEST can not reduce the error. Basically, in the first step, we calibrate manually to suffer. Then, using these sufferings, we start the calibration automatically. This is not always a good automatic calibration because I believe the model is really far from real.

Hi,

I experienced this problem before, and I think that this is due to several challenges that can face your model. If you are pretty sure of your conceptual model settigns as boundary conditions, recharge, ....etc.n, you should check these points:

1) Convergence cirteria, if it's very high the model will converge rapidly and don't go to the extent of reducing discrepancy errors to an acceptable level.

2) Probem of local minima: this keeps calibration trapped to solutions that the model cosiders as the best solutions and it will not evolve to more accurate solutions

3) Calibration of unconfined aquifers is very difficult because of nonlinearity of transmissivity/conductivity during computation iterations. So you can calibrate the model in confined aquifer and assume that top layer of the aquifer as the sought water levels. And after that turn to unconfined aquifer and assess the error of diercet simulation or go further with inverse modelling to enhance calibration -start the inconfined steps with results of confined aquifer calibration-

4) Before launching the automatic calibration, you can begin with manual calibration in order to find the same trend of errors, this could considerably reduce problem of local minima and approach the sought calibration parameters values.

5) Just for confirmation, you should be sure that the degree of liberty of your parameters (such number of parameters and value ranges) are enough to reproduce water levels targets.

Good luck

Maybe a special piezometer generates high error results. Check the errors of the all piezometers seperately, and if one was high, eliminte it.

Hi,

how can i add specific yield to LPF package in GMS?

Please help how i can do calibration for grid approach in GMS software?