The basic science is accepted but it is still unclear whether or not some key elements of the model have been missed out or are poorly approximated. This continues to drive the academic endeavour.
Does this matter?
The simple question soon reveals that in their tremendous efforts to make good physics the basic question has not been answered: are any of the models fit for purpose? If not, will they ever be?
Fit for purpose implies that someone needs to be making some kind of decisions based on these models. But who?
Politicians, insurers, estates managers, businesses, charities among others.
But in each region each stakeholder will have different action thresholds, and each individual threshold is a different function of estimated change; and the precision of that estimate.
For example, if your population largely lives within 1m of high tide then a model which predicts local sea level change to within 5m is not much use. A predicted rise of 0.5 m ±0.1m is good to work with but only if the time-scale is pinned down to better than 20 years. Perhaps the politician would set the action threshold at 0.3 m if the precision is 0.1m but would be say 0.2 m if the precision was less good. A refugee charity would view the matter quite differently.
And so on for rainfall, wind speeds, fog, oxygenation levels, turbidity etc. etc..
Until the stakeholders publish or share these thresholds, their tolerances and derivatives then no-one will know if any of the models are meeting a strategic/commercial need.
- Politicians at least should be exposing their views on these key parameters to public scrutiny. The recent climate change risk assessment in the UK at least takes a step along that path.
- Boards should be asking their executives to be explicit on this. Many have done so. Many haven’t. many are hoping that insurance will come to their aid if they need it.
- People developing commercial risk tools should first prove the validity of those key parameters to their clients, and, once the modelling data is crunched, explain the remaining scope for risk appetite.
What may be good enough to drive political choices may not be good enough for infrastructure investment decisions or insurance strategies.
Each stakeholder should develop their understanding of the key parameters and then ask the modellers, can you deliver what we need? Only by subjecting the various models to the selection pressures created by the stakeholders will the models be identified as fit for purpose. Academic excellence is no guarantee of utility.
Being the best yet, and being commercially available, don’t imply fitness for purpose.