Global Climate Models (GCMs) are suggested to behave chaotically. They are spun up at the beginning of their runs. If the initial conditions are too cold, they will in effect orbit the Warming attractor below more often. If the initial condition are too hot, they will in effect orbit the cooling attractor more often. When they are done spinning up, they will be stable in most cases while it is possible many orbit swaps are happening, they average out. Below the orbit is confined to the circles and at every intersection the orbit may swap to a different attractor. When I say attractor, think of a Lorenz Butter diagram or video. Below we have 3 lobes while the Lorenz model has 2. Whether or not an attractor swap will occur and how often can be calculated with Lorenz equations but the climate is much more complicated.
I am not saying that the GCMs where specifically designed to do what I suggest, but that in effect, that’s what they do. When we see the GMSTs wandering and not following a straight upward trendline, I think we should have an explanation of why that happens. It is also possible that a GCM can be thought of as having only one lobe of the Lorenz Butterfly. In that case, the attractor would have to drift to explain warming temperatures. Above we don’t need attractor drift. Stable means, it doesn’t change up or down. Warming is explained by spending more time orbiting the Warming attractor and the atmosphere warms. It is also possible that a GCM would have 2 lobes, a Warming and Cooling one. I also think it’s possible it could have more than 3 attractors. For instance:
- Warmer and attempts to warm
- Warmer and attempts to cool
- Cooler and attempts to cool
- Cooler and attemts to warm
Also consider scaling all this down to regional and local climate.