- Comment: Global warming was not important in my analysis
- Comment: CO2 was not important or relevant
- Comment: Natural variation was not important
- I now understand why researchers tend to “look where the light is” and often ignore the big picture.
We ought to be thinking about climate drivers both natural and human as being like a “plumb pudding” – a lot of things all potentially having an impact rather than one “control knob”.
In yesterday’s article I showed that my hypothesis that reduced pollution had caused warming was not supported by the evidence and that instead it pointed to a clear link between US holidays and warming in Canada (as above). This and a range of other evidence this points strongly toward contrail warming as causing at least part of the 1970-2000 warming (and for example part of the discrepancy between land and sea temperatures).
However, this created a dilemma. Because now I had to reconcile my understanding that contrails are very likely to be responsible for at least part of the rise in temperature from the 1970s to 2000, with the fact that there is no discernable change in global temperature beyond what we would expect naturally. Had I lost site of the wood for the trees?
OK Yes – when looking at the global temperature on its own there is no change that can be distinguished from noise, but when there are other very specific changes such as the discrepancy between land and sea, the geographical pattern and now the very clear warming during holiday periods, these specific things when combined are unique enough that they can be attributed to a specific cause even if the overall change cannot be.
But reflecting further, I realise that the whole process of trying to “divine cause” is a very seductive lure. The result is that it tends to mean you lose sight of the wood for the trees. For in that analysis I’ve entirely lost site of the following:
- Global warming – I did not find it necessary to adjust any data or even to investigate any change in the Canadian data because it had no impact on the quality of data analysis. IT DID NOT MATTER! The variation each day even when averaging over 50years was still much greater than any presumed warming.
- CO2 – I keep thinking I ought to do more work looking at a possible relationship between CO2 a temperature. I keep thinking I ought, but I still cannot find a way to attribute any change to CO2. So the best I can do is say: “the best estimates of the effect of the measured increase in CO2 is …”
- Natural variation – likewise, I find that the more I look at possible connections between various parameters and temperature, the more I lose sight of the bigger picture – which is that the climate always varies quite naturally. But like CO2, it’s almost impossible to work out a way to attribute any variation to natural variation. And I understand why researchers try to attribute every last change to “something” – rather than stepping back and admitting that we’ll never be able to attribute every change to something concrete.
But my worst sin was this:
Looking where the light is.
There’s an old joke about someone looking for a sixpence under a lamp – and it turns out they lost the sixpence out in the dark. So why look where the lamp is? Because that’s where the light is. Likewise, I find that when analysing global temperature, there is a very strong temptation to look at parameters I understand – which in my case is economic and “physicy” type parameters – what I’m not looking at is possible biological changes. So, for example when I saw a week of higher temperatures in Canada – I assumed human agency. What I did not assume was some biological agency (e.g. pollen from certain crops – although technically that could be human agency as well).
Along similar lines of “not yet having looked at CO2″ – I’ve not yet looked at the possible contribution from solar (mainly because I haven’t spent the time to convince myself I could reproduce the data the “solar guys” use.
As a result I found myself being tempted to say “the 1970-2000 change was caused by …. ” instead of “at least part of the change was caused by …”. A subtle but very important difference. As such I think I should be thinking about the climate and global temperature using a very different model. One I describe as the:
A Climate Plumb Pudding
I’m more and more heading toward what I would call a “plumb pudding” view – which is that there are a lot of very different things which are having an effect and whilst some like the 1970-2000 warming may be analysable and even possibly quantifiable, the actual real effect of some like CO2 and the scale of “natural variation” seem unlikely to be accurately quantified in my lifetime.
Also, I’m increasingly aware that the same pollutant or driver can affect the climate in dramatically different ways depending how it interacts. So, for example, the old approach used by researchers of plotting “Sulphate” emissions as one parameter and trying to attribute change to this one single variable as if it only had one effect on the climate was always doomed – because it is almost as important how sulphate gets into the atmosphere and even geographically where as how much there is present.