I happened to have a look at the Bishop Hill Blog and discovered that Andrew Montford (who writes the blog) was interviewed, together with Paul Williams from the University of Reading, on BBC Radio 5 Live’s Stephen Nolan show. If you want to listen, the show is available for the next 5 days, and the segment starts at about 12 minutes.
The actual segment is fine. I thought Paul Williams did well. Andrew Montford was okay, but it was the standard; climate models are running too hot, uncertainties in the aerosol forcings, we can’t model clouds properly, etc. Nothing too outrageous, just the normal there’s too much uncertainty to be sure, type of argument. Judith Curry would have been proud.
In my opinion, though, there’s something much more important to consider; which is this: if the BBC wants to interview someone about how climate models work and about the uncertainties associated with climate modelling, why do they choose to interview a blogger who – as far as I’m aware – has no formal research experience, has never published a peer-reviewed paper, and has never used a complex climate model? Why don’t they interview a professional, experienced climate modeller?
Could it be that if you did interview an experienced climate modeller, it just might not be controversial enough? They might say :
Yes, the global surface temperatures are now falling outside the 90% confidence interval of the models, but we expect that to happen about 10% of the time, so it’s doesn’t yet indicate an issue with the models. If the surface temperatures never fell outside the 90% confidence interval, we’d eliminate the models that were outliers, which would reduce the confidence interval until the surface temperatures fell outside the 90% confidence interval about 10% of the time.
Maybe that’s just too dull and agreeable. Maybe it’s better to have someone who makes it sound like this is a big issue. An actual climate modeller might also say :
Yes, the mean model trends are higher than the observed surface temperature trends, but the observed trends fall within the range of the model trends. Also, there could be a number of reasons for this. It’s possible that the models are predicting faster warming than will actually happen, but it’s also possible that natural variability has produced some cooling in recent decades and so the observed trend is slightly smaller than it would be without this cooling influence. It’s still too early to really know if the model trends really are too high.
Maybe an actual climate modeller would say :
Yes, the models didn’t predict the recent slowdown in surface warming. However, it’s very difficult to predict such variability, and what’s presented is typically ensemble averages of the model runs, which tends to smooth out short-term variability. Also, these models are not designed for decadal predictions, and so short-term mismatches are not particularly surprising.
They might add :
Yes, it’s possible that model predictions for climate sensitivity are too high, but the range is very similar to that estimated using other methods. Hence, it would be very surprising if climate sensitivity does turn out to be very different to that predicted by climate models.
They could also add that our understanding of climate change and global warming does not depend only on climate models. There are many independent lines of evidence – as pointed out by Paul Williams in the interview – that are largely consistent. Also, a major part of global climate modelling is to make predictions about how climate change will effect different regions of the globe. Of course, I’m not a climate modeller, so I can’t really say what one would actually say. I would be surprised, however, if they didn’t provide some context and didn’t illustrate the significance/relevance of the various uncertainties.
So, it’s possible that if you were to interview an actual climate modeller, it might just be too dull and uncontroversial. Maybe if you want to improve your ratings and make it more interesting, interviewing a blogger with no formal experience is better. On the other hand, if you actually want to present an informed view of our understanding of climate models and climate modelling, that may not be the best strategy. Personally, I’d rather the BBC went for the latter strategy, but maybe I’m alone in thinking this.