Thursday, 19 December 2013

Climate science has not failed

There has been a lot of fuss about an apparent difference between the temperature record for the past decade or so and the predictions of climate change models.  Most of what has been said shows little understanding of how models work, or what the meaning of an apparent difference really is.

Computer models of something like climate can never be perfectly accurate because climate is a chaotic and complex result of so many factors.  The way predictions are made is to run models to get a statistical ensemble - a large number of predictions that allow some degree of statistical confidence, and that also help to tell us which factors are significant in causing climate change.

Models will be run which deal with different aspects of climate.  Some will deal with sea temperatures, some air temperatures, some will look at rainfall and so on.  It is really important to understand this, because it shows that what appears to be a difference between reality and the predictions of models of one aspect of climate does not by any means indicate that the science of climate change modelling is flawed.  

Models will be run to answer different questions.  Some will be run to see how much human activity is resulting in global warming and others will be run to see where that warming happens.  This is important to understand because a lack of accuracy in a prediction about where the warming happens does not by any means indicate a lack of accuracy in predicting the overall amount of warming.

Models will be run over different time scales.  Some models are run to help find the parameters which fit past climate measurements so as to provide information to built and run models which can be useful in prediction.  Other models will be run to predict short term climate changes, others will be run to predict long term climate changes.  That some models may not fit well one time scale does not mean that other models won't fit a different time scale.

Of course not all the models will work with the desired accuracy - that is part of the science of modelling, in which development of modelling techniques is always ongoing, but a lack of accuracy in one area doesn't mean abandoning modelling, or rejecting all predictions.

Global warming has not stopped.  There has been a puzzle about where some of the heat energy has gone in recent years, but that puzzle has been solved, and with worrying consequences.  The heat energy has gone more into the depths of the seas than expected, and more into the Arctic than expected.  Extra warming in the Arctic will cause greater ice loss and more potential for feedback systems.  Extra warming in the deep seas will result in a greater tendency for areas of the sea to become oxygen-free, resulting in the growth there of microbes that produce toxic sulphur compounds.

Climate modelling has not failed; warming has not stopped, and the danger is real

1 comment:

Anonymous said...

You miss the point. As Richard Feynman stated, if the theory does not fit the facts, it is wrong.

Continually tweaking a theory to make it fit the facts is a good sign the theory is fundamentally wrong.

You seem ideologically committed to the idea of global warming - until the possibility that the modelling is badly wrong, there is no real point in further discussion. It has become religion, not Science.