Since I’m up early and waiting for the rest of the family to rise, I thought I might comment on this Miranda Devine article which claims that Perth electrical engineer’s discovery will change climate change debate (H/T Ben Cubby and Ketan Joshi on Twitter). The electrical engineer is David Evans, who is married to Jo Nova. The reason that I thought I would comment is that I spent some time on Bishop Hill pointing out to him that his discovery was no such thing. I won’t link to it because I managed to make a rather embarassing – but acknowledged – blunder myself at the end of that comment thread (you can probably find it if you wish 🙂 ).
David Evans has a whole series of posts on Jo Nova’s blog where he discusses his discovery. I’ll just comment on the aspect that I was discussing with him and which he discusses in this post. He says
The basic model relies heavily on partial derivatives. A partial derivative is the ratio of the changes in two variables, when everything apart from those two variables is held constant. When applied to the climate, this means everything about the climate must be held constant while we imagine how much one variable would change if the other was altered.
As far as I’m aware, this is simply untrue. A complex GCM certainly solves a set of partial differential equations, but these are the standard Navier Stokes equations which are evolved in time and space; it doesn’t, however, require holding everything constant while we check how one variable changes if another is altered. The model simply evolves all the different variables with respect to and .
The most basic climate model, on the other hand, doesn’t use partial differential equations at all; it normally simply evolves the change in temperature on the basis of a forcing time series and a feedback response that is typically assumed to depend linearly on temperature. You can introduce non-linearities and make them more complex, but even basic climate models don’t solve the partial differential equations that David is claiming that they use.
What David Evans appears to be referring to is how one might determine – for example – the feedback response from a climate model. One may indeed do so by holding everything constant, bar one thing, and then determining how the system responds to a change in another variable, such as temperature. However this does not mean that a climate model is evolving this type of partial differential equation; it simply means that this type of equation is used to analyse the output from a climate model. I encountered a similar issue when I had a discussion with Monckton a while back; confusing how one might analyse the results from a climate model, with how a climate model is actually run. Could there be a link?
So, as far as I can tell, David Evans’s startling discovery is simply him being confused about how climate models actually work. Miranda Devine’s article includes that
Dr Evans is an expert in Fourier analysis and digital signal processing, with a PhD, and two Masters degrees from Stanford University in electrical engineering, a Bachelor of Engineering (for which he won the University medal), Bachelor of Science, and Masters in Applied Maths from the University of Sydney.
Not only did David Evans bring up his qualifications in our discussion on Bishop Hill, but his expertise in Fourier analysis and digital signal processing doesn’t seem to have helped him in the past. What Miranda Devine’s article mainly illustrates is that some people will promote anything as long as it appears to suggest that there are major problems with climate science, even if it is written by someone who seems to thinks that where they got their PhD is somehow relevant. Some might call that irresponsible.