Climate model tuning

[rconnor]

rb1957,

Granted, models that diverge greatly from historical data are re-examined but, again, the tuning is done to more accurately represent the subsystem, not as a blind attempt to match historical data.

From the paper:

The MPI-ESM was not tuned to better fit the 20th Century. In fact, we only had the capability to run the full 20th Century simulation according to the CMIP5-protocol after the point in time when the model was frozen.

 

[rb1957]

in that case, why match these isolated effects ? my sense (and it seems others as well) tells me that something was done specifically to create the effect in the model.

how does modelling these isolated events "improve sub-system representation" ?

another day in paradise, or is paradise one day closer ?

 

[TGS4]

In defense of the models and modelers (I know, I know...) there are some events which are externalities to climate system. Changes in TSI (small though they may be) are external forcings and are input as time-varying boundary conditions. Similarly, vulcanism in not something intrinsic to the coupled atmosphere-ocean system, and hence would be a boundary condition input.

However, dissipative self-organizing features such as thunderstorms or hurricanes/typhoons are intrinsic to the coupled atmosphere-ocean system and should not be forced but should be allowed to evolve.

One of the fascinating features of vulcanism is that they provide an excellent insight into how the coupled atmosphere-ocean system reacts in the short-term to step changes in incoming solar irradiance (via particulate/aerosol blocking). They provide excellent opportunities to learn about our world.

However, because the models poorly represent regional data (ref. IPCC), it is not surprising that they don't handle these regional effects, which can have global impact, well.