It was while checking on Black's Twitter feed that I saw mention of a story that says the Russian heatwave can be explained as part natural and part global warming. Or in Black's Tweet:
Latest on Russian heatwave - natural or human-driven? New AGU paper says a bit of both http://t.co/rIicn4Fu - extreme temp risk x3 from AGWFollow the link and you get a story on the AGU website:
Russian heat wave had both manmade and natural causesNow that's a fairly definite headline - and it's in contrast to previous studies that explained the heatwave as being due to natural weather phenomena and not AGW. Attribution - assigning causes to specific weather events - is a big deal. If scientists can pin the blame on climate disasters to CO2 then that's a major step forward in convincing the unconvinced masses that the AGW hyopthesis is not just a massive scam.
So how have they untangled the complex mix of factors that can partially pin the blame for the heatwave on human CO2 emissions?
Well, the short answer is that they haven't. What they've done is run a series of simulations and compared results. In one set of simulations they used data said to represent the 1990s, in the other they ran data said to represent the 2000s. The results showed that the chances of a comparable heatwave occuring were three times greater in the later set of simulations.
Yes, that's all there is to it. You run a model one way, tweak the inputs and run it again and then compare the results. You don't actually look in detail at the specifics of the real heatwave that happened. You don't look at the weather patterns or anything like that. You run two sets of simulations and just note that in the second set you can get some of the characteristics of a heat wave occuring more often.
I don't need to labour the obvious inadequacies of this. I don't need to point out the inherent uncertainties involved in such a process.
But where is the hint of this uncertainty in that headline or in Richard Black's Tweet? Where is there anything to indicate that this is all based on software and models?