Question from Forum

[lrennels]

Original Question

Hi,

Good day! In MimiFAIRv1_6_2, I noticed that the Montecarlo simulation function (run_mcs) is only associated with the SSPs (defined by the get_model), whereas the get_model_original (defined for the RCPs) do not have a corresponding run_mcs function. Is this right? Please if not, is there an existing function for running mcs with the get_model_original ? If yes, does it also apply to MimiFAIR as used in MimiGIVE?

Thank you for your assistance!

[lrennels]

Since the original version was kept around primarily for testing, but the primary version developed in this repository is intended to match the AR6 calibration, which is also used in MimiGIVE, we did not further develop an additional run_mcs and did not have another set of calibrated parameters intended for use with the original version. Insofar as the two models have the same parameters defined you functionally could run the same mcs function on the original version, but the results would not be particularly useful because you would be running a model with a set of calibrated parameters that are not consistent with the settings for that original version. I would not use the original version in practice.

Per the RCPs and SSPs, it is true that the input for the run_mcs is an SSP, because the scenarios defined by AR6 are defined by SSP, but note that each SSP in this case is modeled alongside an RCP, so for example ssp245 means SSP2 run to meet RCP4.5

As to your last question about MimiGIVE I'm not sure I understand the question. MimiGIVE was originally written to use the climate model and parameters that best matched AR6. One could decide to force it with a different climate model altogether, or a different version/parameterization of FaIR, if they chose to.