Do climate and biogeochemical models simulate past changes well enough that we can have confidence in their ability to tell us the effects of different options for the future?

It is pointless to make ever more detailed projections unless we have good confidence in the models – and right now they are still a ways away from being able to understand even the relatively recent past. We are far from being in a position to move away from a natural science agenda, as the Science editorial suggested!

8 comments

1. Alastair says:

   at

   The problem with this question is that the answer is a value judgement and not factual.
   I suspect those who want this question answered already believe that the models are not good enough, as I do.
   But how do we persuade ICSU editor we are correct?
2. Kalense says:

   at

   Clim101 and georgii are correct, in my view. Though I wouldn’t formulate the question quite as georgii has, it is a more honest way of looking at things than the original one.

   Perhaps “Is the confidence interval of models sufficiently narrow for any practical uses of their predictions?”
3. Ntchoua says:

   at

   I think this topic is strongly research oriented if it’s well addressed.
   Thank
4. georgii says:

   at

   This question presumes that model ability to simulate the past implies the ability to predict the future. This presumption is questionable.

   If ICSU would like to include this question into the agenda of ESS research, it would be better to formulate it as follows. Is the confidence interval of model-based predictions too wide for choosing between the strategies of world development?
5. jasc says:

   at

   Models are only one of the tools to understand the environment and should be validated against real data, not other models. Looking back is not only desirable, it is a scientific need.
6. clim101 says:

   at

   Note also that achieving good past hindcasts does not guarantee success at future climate prediction. Many climate models can simulate changes in Earth’s climate due to past changes in orbital forcing etc however I don’t believe it is clear that some ranking based on their performance at this will yield better future predictions. It may indeed put too much of a constraint on our future predictions.
7. africain says:

   at

   “we are far from being in position to stop working on the fundamental laws of the Earth system” : yes indeed, but we are unfortunately unable to wait for a perfect knowledge of the Earth system before beginning any action : it would be too late.
8. joel.savarino says:

   at

   Perfectly right,

   now that the global change is on top of the political agenda and the international scene, our policy markers want fast and concrete actions for the “good” (= their reelection) of the people, it doesn’t matter if solutions to be offered are detrimental as long as they show actions.
   The truth is that we still know little about the fundation of the Earth system and the complexities of the interactions between the different spheres (bio, hydro, atmo, cryo, etc). Yes we are far from being in position to stop working on the fundamental laws of the Earth system.

   It is frightening to read such kind of ICSU editorial.