Brian Hoskins and Steve Smith, advisers to the Committee on Climate Change, have written (yet another) riposte to David Rose's article in the Mail on Sunday. This is crazy, crazy stuff:
A chart of observed global temperatures against climate model outputs is the main evidence provided in the article. It claims that the chart “blows apart the scientific basis” for reducing emissions. This is simply incorrect, and reveals a misunderstanding of what the chart shows – a pattern of observed temperature over the last sixty years within the range of model outputs (see detailed notes below).
As Rose correctly notes in the version of the graph published at the Mail, it is only the final few years of the graph that are predictions, but Hoskins and Smith are actually asking us to draw comfort from a hindcast!
With accurate models we would expect about 3 years of observed temperature in the past 60 to be below the 90% range and 3 years above it. The observed temperature behaviour is consistent with model outputs, even during the last 15 years of negligible temperature increase, when temperatures have remained within the 90% range.
It's a similar story when Hoskins and Smith move onto climate sensitivity, with much attention devoted to climate models and to studies based on long-term temperature records, which even the IPCC says are extraordinarily uncertain.
Then there is this:
Other work (PDF) indicates that aerosol pollutants which cool the climate may be offsetting greenhouse gas warming less than previously thought. Since this slightly reduces the amount of underlying warming that can be attributed to greenhouse gases, it too suggests the most extreme future projections are less likely, while still keeping most of them in play.
Now correct me if I'm wrong, but wasn't it the aerosols that were supposed to have prevented warming taking off at the predicted rates in the past? And now their influence is slight?
And what about all the empirical and semi-empirical studies that suggest that ECS is around 1.5degC? Forster and Gregory? Aldrin et al? Ring et al? There's only this:
But while it is true that recent studies based on temperature observations question the very highest model projections of warming, even if confirmed, these would not justify a wholesale downward revision of the range as argued in the article.
It's hard to take seriously someone who prefers models to empirical observation - as I've pointed out before, this is the basis of the scientific method. It's harder still to not to contain one's ridicule at someone who argues that the output of models should be taken into account when they do not include the latest figures for a key forcing, when they have proven entirely incapable of predicting anything, and the output of which is on the verge of falsification in an extraordinarily short space of time. But when that someone - Hoskins - has gone on record as saying that the models are "lousy" but still insists that they have a part to play in assessing climate sensitivity, you really have to wonder whether this is a scientific argument at all.