Judith Curry is a science-denying blogger who is, by day, a climate scientist. Today on her blog, she copies and pastes huge slabs from a couple of interesting blog posts and twists them about to suit her own agenda, writing (archived here):
The IPCC has a very bad case of confusing the probability inside their argument with the probability of the question as a whole (e.g. 20th century attribution, 21st century projections, climate sensitivity). Dangerous anthropogenic global warming is one possible scenario of the future; there are many other possible scenarios that the IPCC completely ignores (heck, we cant predict solar variations, volcanic eruptions, and natural internal variability so we might as well ignore them).But Judith Curry doesn't know her inside from her outside. And if she's read the AR5 reports, she's telling fibs. If she hasn't then what business does she have making that claim?
The blog article Judith quotes from is on a blog called LessWrong and has the title: Confidence levels inside and outside an argument. The argument draws on examples from political polling, the Large Hadron Collider and creationism, not climate science. Here is a segment, so you can see what Judith is on about (my bold italics):
When an argument gives a probability of 999,999,999 in a billion for an event, then probably the majority of the probability of the event is no longer in "But that still leaves a one in a billion chance, right?". The majority of the probability is in "That argument is flawed". Even if you have no particular reason to believe the argument is flawed, the background chance of an argument being flawed is still greater than one in a billion.
More than one in a billion times a political scientist writes a model, (th)ey will get completely confused and write something with no relation to reality. More than one in a billion times a programmer writes a program to crunch political statistics, there will be a bug that completely invalidates the results. More than one in a billion times a staffer at a website publishes the results of a political calculation online, (th)ey will accidentally switch which candidate goes with which chance of winning.
So one must distinguish between levels of confidence internal and external to a specific model or argument. Here the model's internal level of confidence is 999,999,999/billion. But my external level of confidence should be lower, even if the model is my only evidence, by an amount proportional to my trust in the model....
... This is not a fully general argument against giving very high levels of confidence: very complex situations and situations with many exclusive possible outcomes (like the lottery example) may still make it to the 1/10^20 level, albeit probably not the 1/10^4478296. But in other sorts of cases, giving a very high level of confidence requires a check that you're not confusing the probability inside one argument with the probability of the question as a whole.
Judith Curry is Wrong No. 1
So far so good. I see no basis for disagreeing with that article from LessWrong. Thing is, Judith Curry does this sort of thing all the time in reverse. Instead of overstating confidence she understates it and often by a very wide margin. She does it in her very article on the subject (archived here). Judith Curry pretends that climate science is not presented in context (it is), that unknowns and uncertainties are not described (they are), that probabilities are not quantified (they are), that assumptions and constraints are not listed (they are). So while Judith writes "The IPCC has a very bad case of confusing the probability inside their argument with the probability of the question as a whole" - she's dead wrong.
For one thing, the IPCC uses a number of different models. It is not a single model that is used for projections. Therefore, while any single model may have a "bug" that completely invalidates the results, the chances of multiple models all having the same "bug" is much smaller. The chance that a staffer will accidentally switch an equation relating to total solar irradiance with one relating to CO2 forcing becomes much less. And when the model outputs are considered against what is known about the behaviour of the earth system, the chance of internal "bugs" causing "all the models to be wrong" becomes even less.
For another thing, I've not read anywhere in the IPCC documents that the chances of any one thing happening have a probability of 999,999,999 in a billion. I don't know of any science, let alone climate science, where that sort of probability would be assigned - at least not without 999,999,999 provisos and caveats :)
Judith Curry is Wrong No. 2
Not only is Judith wrong about applying that argument to the CMIP5 models, her "many other possible scenarios that the IPCC completely ignores...solar variations, volcanic eruptions, and natural internal variability" is wrong. This is from page TS-46 of the IPCC AR5 Technical Summary in the section on model projections (my bold italics):
There is low confidence in projections of natural forcing. Major volcanic eruptions cause a negative radiative forcing up to several W m–2, with a typical lifetime of one year, but the possible occurrence and timing of future eruptions is unknown. Except for the 11-year solar cycle, changes in the total solar irradiance are uncertain. Except where explicitly indicated, future volcanic eruptions and changes in total solar irradiance additional to a repeating 11 year solar cycle are not included in the projections of near- and long-term climate assessed. {8, 11.3.1}
So the IPCC report did not ignore other possible scenarios. It just didn't see fit to model things like a major volcanic eruption next year, or five, ten, twenty five or 89 years from now. What would be the point? It's addressed the amount of negative radiative forcing such an eruption would bring and Judith can do her own estimate if she wants to bet there will be one.
As for changes in total solar irradiance, this is from page TS-21 of the IPCC Technical Summary (my bold italics and para):
Solar and volcanic forcings are the two dominant natural contributors to global climate change during the industrial era. Satellite observations of total solar irradiance (TSI) changes since 1978 show quasi-periodic cyclical variation with a period of roughly 11 years. Longer-term forcing is typically estimated by comparison of solar minima (during which variability is least). This gives a RF change of –0.04 [–0.08 to 0.00] W m–2 between the most recent (2008) minimum and the 1986 minimum. There is some diversity in the estimated trends of the composites of various satellite data, however. Secular trends of TSI before the start of satellite observations rely on a number of indirect proxies. The best estimate of RF from TSI changes over the industrial era is 0.05 [0.00 to 0.10] W m–2 (medium confidence), which includes greater RF up to around 1980 and then a small downward trend. This RF estimate is substantially smaller than the AR4 estimate due to the addition of the latest solar cycle and inconsistencies in how solar RF was estimated in earlier IPCC assessments. The recent solar minimum appears to have been unusually low and long-lasting and several projections indicate lower TSI for the forthcoming decades. However, current abilities to project solar irradiance are extremely limited so that there is very low confidence concerning future solar forcing.
Nonetheless, there is a high confidence that 21st century solar forcing will be much smaller than the projected increased forcing due to GHGs. {5.2.1, 8.4; FAQ 5.1}.
Now what other "possible scenarios" does Judith want to factor in for the future? Geo-engineering? The IPCC has addressed that one. What else? A pandemic wiping out half the human race? While there may be more than an infinitesmal chance of that occurring, what is the point of factoring in such an event?
Judith may think she is queen of uncertainty but if she does know the first thing about it, she hides her knowledge well.
Judith is queen of obfuscation and science denial. That's all -1!
Judith Curry Plays Politics with Uncertainty
The next slab that Judith Curry quotes relates to the value of scientific authority. It's from another blog article on the LessWrong blog. Ironically, Judith commits the very sin discussed in the article itself. The article is about how to convey scientific probability to a lay audience. Something that Judith Curry does her best *not* to do. Judith copied a huge amount of the article but for some reason left off the last sentence, which is:
In fact, it seems to me that to prevent public misunderstanding, maybe scientists should go around saying "We are not INFINITELY certain" rather than "We are not certain". For the latter case, in ordinary discourse, suggests you know some specific reason for doubt.
Judith writes about "communicating uncertainty" and "playing politics with uncertainty" - she is a fine one to talk! She writes:
I found both of these essays to provide substantial insights into reasoning about climate uncertainty, confidence levels, communicating uncertainty to the public, and playing politics with uncertainty and confidence levels.
Judith makes a habit of playing politics with uncertainty. That's what her blog is for. It's not to communicate uncertainty in a scientific sense. You only have to look at this article and half a dozen of her other articles to see that her aim is to pretend "we are not certain" means "scientists don't know nuffin'".
Judith blogs to confuse her readers who are less knowledgeable about scientific probability. To try to persuade them that "there are many other possible scenarios" while not acknowledging that those "other possible scenarios" are much less likely and/or have already been quantified or addressed.
All Judith ever promotes is the "we are not certain" part, while waving her Italian flag - and implying that certainty expressed in all aspects of climate science is much less than it really is.
Here's a link to an archived copy of Judith's blog post again.
That's all - 2!
