Page 1 of 1
Calibrating various scenarios beside the baseline
Posted: Sun May 24, 2020 12:28 am
by Simplifier
I have calibrated the baseline scenario in my model and I have established two more possible scenarios. Does the software need to recalibrate and re-estimate constants based on the two new scenarios or it only needs to run simulation for them using the baseline estimations?
Re: Calibrating various scenarios beside the baseline
Posted: Sun May 24, 2020 1:17 am
by tomfid
I think the answer depends on whether the scenarios differ only in terms of future policy, or also past variables that would affect the fit to history.
Normally I would think of there being a single baseline, with policy scenarios varying around that.
The way to implement this is to calibrate the model, then use the .out file from the calibration as a changes file to set up further simulations.
Re: Calibrating various scenarios beside the baseline
Posted: Sun May 24, 2020 1:24 am
by Simplifier
tomfid wrote: ↑Sun May 24, 2020 1:17 am
I think the answer depends on whether the scenarios differ only in terms of future policy, or also past variables that would affect the fit to history.
Normally I would think of there being a single baseline, with policy scenarios varying around that.
The way to implement this is to calibrate the model, then use the .out file from the calibration as a changes file to set up further simulations.
The baseline scenario is three technologies (x, y, z) interacting with one another until one becomes dominant by 2050. However, in each of the other two scenarios I only consider two of the three technologies (for example x and y, and x and z) to see the dominance pattern and result.
Re: Calibrating various scenarios beside the baseline
Posted: Sun May 24, 2020 3:07 am
by tomfid
In that case, I think you'd want to recalibrate each of those.