Hi everyone,

I’m having an issue with scenario inheritance and calibration in WEAP.

I created a new Baseline scenario as a child of the Reference scenario, using the same time period (2018–2024). However, when I check the results, the Baseline scenario loses the calibration, even though it should inherit everything from Reference. I realized that maybe my PEST calibration results are stored only inside the PEST set (e.g., “set4”) and not in the actual data of the Reference scenario.
From what I understand, WEAP does not automatically write the optimized PEST parameters into the scenario. They must be explicitly applied.

If this is true, how can I permanently save the PEST-optimized parameter values into the Reference scenario (or Current Accounts) so that all child scenarios, including Baseline, inherit the calibrated data?

There is a second issue:
When I create future scenarios, the Reference scenario’s time period also changes, and the calibration becomes invalid or lost. I assumed new future scenarios would only modify themselves, not the Reference scenario, but it seems changes to the time horizon affect the whole project.
How can I create future scenarios without altering the Reference scenario’s time period or affecting its calibrated results? Is there a recommended way to “lock” or preserve the Reference period so it remains untouched while exploring future conditions?

Simbarashe,

While I can't speak much to PEST, I can tell you how to run both historical and future scenarios. I recommend to use the YearOffset feature in ReadFromFile, adjust the Base Year Data setting in the ReadFromFileWizard in your historical scenarios. Using this, you could for example set your model period to run from 2030-2050, but have your historical scenarios look at time-series data from 1990-2010. See the Help entry for ReadFromFile wizard within WEAP for more detailed information on this. Note you could also do it vice versa- run your model in historical years and modify the YearOffset for future scenarios.

Another option is to modify the timestep back and forth from past to future and run either only historical or future scenarios at a time.

Cheers,
Doug