I want to put down one of my “ridiculous ideas” here. But am having trouble phrasing it in such a way as to avoid being obviously wrong.
We all agree that global warming is not synonymous with climate change, right?
We also all agreee that global warming is more than 90% due to greenhouse gases, right?
Two main aspects of climate change are ‘extreme events’ and ‘regional changes in rainfall’. It’s fair to say that those are the two aspects of climate change that most affect farmers, right?
Now I’m going to change tack. When doing computatioinal fluid dynamics, a huge concern is how to separate short term fluctuations (turbulence) from long term fluctuations (variation in mean flow). It can’t be done, so we artificially select a time scale to separate the two. The equiations are solved by integrating over that time scale. That isn’t really satisfactory, and a more advanced technique is to use what is called ‘ensemble averaging’, which involves creating alternative theoretical universes where the same experiment can be run over and over again.
Changing tack again. There are at least three different types of “random”. One familiar type of random has a mean and a standard deviation. A second type of random is chaos, where there is a mean but beyond that things get complicated. A third type of random is the random walk. There is no time-independent mean or standard deviation. A familiar example of the random walk is evolution. You can’t stop evolution because you can’t stop mutation. There is no “attractor” (using a term from chaos theory) forcing evolution to tend towards a final state.
How does this relate to climate and climate change?
Climate change is caused by deterministic factors, such as global warming. But some aspects of climate change are also caused by random factors.
Going back to fluid dynamics, how to separate short term fluctuations (weather) from long term fluctuations (climate change). We know that weather is chaotic, and the equations of chaos give us the exponential rate of separation between adjacent ensemble events, often called the butterfly effect.
It’s no stretch of the imagination to allow the butterfly effect to extend to timescales beyond the timescale for the formation of hurricanes to the timescale of climate change, for those aspects of climate change not directly explained by global warming.
Now here comes my “ridiculous idea”.
What if what we actually have here is the “random walk” type of randomness rather than the “chaos” type of randomness? To put is another way, what if the “strange attractor” of chaos only operates on timescales much longer than human experience, on timescales of 100,000 years or a million years or longer?
In a random walk there is no steady state mean, only permanent change. So those aspects of climate change that are not deterministically caused by global warming would be in permanent change.
This could affect most of ‘extreme events’ and ‘regional changes in rainfall’, the two aspects of climate change that most affect farmers.