There are potentially thousands of interdependent variables that impact climate, on both a local and global scale.
Everything from the slight variations in the suns energy, the extent of ice (reflects heat), chemical composition of the atmosphere/land/oceans (including everything we emit), existing distribution of energy (GHG’s trap more energy in the system), geography and terrain, to the plants and animals.
The reason we can’t reliably predict the weather beyond a week in advance is due to this enormous amount of variables. In statistics, attempting to model the behaviour of 1 variable via its dependent variables becomes extremely difficult beyond a relatively small number, especially when there are multiple confounding variables — the weather and climate is that, except there’s hundreds/thousands of dependent and confounding variables that impact each other. All we can do is provide a probability distribution or range.
That’s also why climate scientists stress for warming to be limited as much as possible, because every 0.1 c increase in temp, increases the volatility and uncertainty of the modelling; feedback loops exponentially increase the complexity (predictability) and, if triggered, could render all our existing climate modelling devastatingly inaccurate.
NOTE: GHG’s are the strongest variable that we have any control over, and the probability distribution of climate modelling over the last 50 years has been extremely accurate — that is to say that the majority of historic modelling predicted where we are today as the most probable.
There are potentially thousands of interdependent variables that impact climate, on both a local and global scale.
Everything from the slight variations in the suns energy, the extent of ice (reflects heat), chemical composition of the atmosphere/land/oceans (including everything we emit), existing distribution of energy (GHG’s trap more energy in the system), geography and terrain, to the plants and animals.
The reason we can’t reliably predict the weather beyond a week in advance is due to this enormous amount of variables. In statistics, attempting to model the behaviour of 1 variable via its dependent variables becomes extremely difficult beyond a relatively small number, especially when there are multiple confounding variables — the weather and climate is that, except there’s hundreds/thousands of dependent and confounding variables that impact each other. All we can do is provide a probability distribution or range.
That’s also why climate scientists stress for warming to be limited as much as possible, because every 0.1 c increase in temp, increases the volatility and uncertainty of the modelling; feedback loops exponentially increase the complexity (predictability) and, if triggered, could render all our existing climate modelling devastatingly inaccurate.
NOTE: GHG’s are the strongest variable that we have any control over, and the probability distribution of climate modelling over the last 50 years has been extremely accurate — that is to say that the majority of historic modelling predicted where we are today as the most probable.