I think a large part of the problem is that the very concept of climate implies a certain level of stability. We are currently in a situation where the planetary energy levels are in a state of change. If you run a model 100 times in a changing system you are going to find 100 different results. It is highly unlikely to be able to predict how all of this new energy is going to effect a million different variables. In a complex system in a state of flux chaos rules, in my limited understanding, this is where the butterfly effect is important, this is where the diversity of possible futures is highest.
Conventional wisdom is that weather is chaotic but climate is not being a boundary condition problem rather than a chaotic problem.
"In a complex system in a state of flux chaos rules" If there is a stable underlying system with chaotic noise on top, how are you ruling out small stable changes in the underlying system in order to jump to the chaos conclusion?
All models are wrong; some models are useful.
In this case, models are better at some things than others. Probably important to try to use them where they are reasonable? If you want to use them where they are bad, maybe a few bad models is better than no model, but take care.
1. Bad models are useful. Perhaps we learn most and fastest from bad models as we learn to improve them. I strongly believe in the use of a wide variety of models in any undertaking that involves thinking.
2. I do not consider my jump to the chaos conclusion as a given but as a higher probability. The primary reasons being that as total energy in the system increases the likelihood that each physical feature on the planet will react to or alter the energy in the system in an unforeseen manner increases. These unforeseen details have an increased chance of accumulating in unseen ways to produce unseen results. Each stable underlying system is increasingly likely to be altered, to become less stable. Although stable systems are likely to resist change, changing the systems around them increases the likelihood of changing them as well.
3. We are now in a system with energies for which we do not have solid historical data to check and calibrate our models with. The energies in the system are continuing to increase, decreasing our ability to check the model. The rate of increase of energy in the system is continuing to increase. Chaos, I believe, increases the probability of un-predicted (unpredictable?) conditions occurring, particularly during the periods of most rapid change. Once energy levels become more stable model predictability will increase but the next steady state may not reflect the conditions that occur during the periods of rapid change. The period of rapid change will be, at a minimum, 100 years.
4. This is the babbling of a lowly field botanist who is better suited to roaming around the woods and looking for things. I am happy to be schooled in matters physical and mathematical. I consider it a complement that my comment drew criticism from you Mr. Crandles as I have had my opinions changed more than once by your commentary.