My paper Classical Decision Rules and Adaptation to Climate Change suggests ways of thinking about adaptation decisions in practical climate change settings where probability information is poor.
The idea is to identify the net benefits from climate change when policies are applied and they work in averting catastrophic outcomes, when climate change does not produce catastrophic outcomes even without policy and when policies fail to be effective even though adaptation policies are applied. Then one can implement a version of the Precautionary Principle and try to avert the maximum possible damage - this is a minimax rule or, alternatively, try to minimise the regret one experiences in the future - the minimax regret rule. The minimax rule suggests adapting only if one can be sure policies will work. The minimax regret rule always provides a case for enacting policy if losses consequent to climate change greatly exceed the costs. This rationale justifies 'all weather' policies but can justify trying a mix of policies only if employing the mix eliminates the prospect of policy failure. The analysis is applied to a two period model of the Murray-Darling River system.