Andrew Bolt on Climate Predictions

Three years ago Andrew Bolt, a political columnist for the Herald-Sun, ran a 'blog article entitled "The Worst Ten Warming Predictions". It is opportune to take stock of the predictions and to illustrate how the debate can be unreasonably skewed by those who are unfamiliar with the scientific method or are politically partisan.

This response will go test Mr. Bolt's claims point-by-point.

1. OUR CITIES WILL DIE OF THIRST

Mr. Bolt selects a number of comments by Professor Tim Flannery concerning the status of water supply in dams. For example in March 2008: "The water problem is so severe for Adelaide that it may run out of water by early 2009". Flannery made this comment in a very brief interview in the March 2008 of Jetstar magazine. In response, Bolt points out that Adelaide's reservoirs were 75 per cent full at the time of writing.

Bolt also criticised Flannery for saying "water supplies are so low they need desalinated water urgently, possibly in as little as 18 months". This is actually a truncated quote from his editorial in New Scientist. The remark was actually for Adelaide, Sydney and Brisbane. As evidence against Prof. Flannery's proposition, Mr. Bolt points out "its [Brisbane's] dams are 46 per cent full after Brisbane’s wettest spring in 27 years." etc.

Bolt he hasn't given sufficient (indeed any) consideration to the fact that Flannery's remarks are guarded propositions; "may run out", "possible in as little" etc. Scientific statements should always be expressed in this form as it deals with trends and other variables may be introduced (such as the introduction of a desalination plant). Knowledge, at least to those who know that they don't know everything, is tenuous.

Further, in seeking to counter Flannery's claims, Bolt cherry-picks particular examples, rather than using trends, a tactic typical in political rhetoric. He points out that Sydney's dams at the time were 63% full after a particularly wet June. He does the same with Brisbane which had the "wettest spring" in 27 years and that Perth had its wettest November in 17 years. Such instances are meaningless in themselves (as is pointing out individual particular hot or dry months); trends are what is is important from a scientific perspective.

Finally, as should be obvious from Flannery's remarks and the New Scientist editorial, the professor is strongly advocating the introduction of desalination plants (and/or usage reduction) to make up for declining rainfall and natural supplies. With no sense of contradiction Bolt himself notes the success of this project: "Perth now has the lowest water restrictions of any state capital [actually incorrect, that is Hobart], thanks to its desalination plant ....". So whilst criticising Flannery's concerns, Bolt is actually using Flannery's solution. I am quite unable to fathom what sort of mind engages in such a confusion.

What about data for actual rainfall over time, which is the real contributing issue? The information is easily available from the Bureau of Meteorology. My general recommendation would be to pick the longest-term data closest to catchment areas, a detailed analysis I'll leave to the people paid to do this. Data analysis is of course difficult, as there can be significant variation in individual years; perhaps the most effective method would be to take a five-year averages and a rolling mean. But a cursory look does suggest that, as being predicted, tropical areas (e.g., Darwin, Brisbane) are experiencing an increase in rainfall, and even flooding, whereas temperate areas (Perth, Adelaide, Melbourne) are facing significant declines. Further when looking at water inflow, the water collected by reservoirs independent of the desalination plants, it doesn't look so good for Melbourne and Western Australia is in serious trouble.

So yes, some cities - those without desalination plants - will indeed 'die' of thirst with current trends. Half a point to Mr. Bolt, because Prof. Flannery suggested some tropical cities which will initially receive increased rainfall due to a warming climate.

2. OUR REEF WILL DIE

Mr. Bolt criticises Professor Ove Hoegh-Guldberg for warning in 1999 of coral bleaching on the Great Barrier Reef, and in 2006 warning "between 30 and 40 per cent of coral on Queensland’s great Barrier Reef could die within a month". Mr. Bolt notes that the Global Coral Reef Monitoring Network in 2008 said there had been no big damage to the reef caused by climate change in the four years since its last report. He also refers that "veteran diver Ben Cropp said this week that in 50 years he'd seen none at all". Apparently the lesson is "Reefs adapt, like so much of nature."

Criticism can be levelled, once again, of Mr. Bolt's extremely selective and narrow use of dates and the confusion between warnings and certainties. In contrast Mr. Bolt's blithe "she'll be right, reefs adapt", is nothing short of ignorant and irresponsible. As Professor Hoegh-Guldberg responded on his own 'blog big coral deaths occurred in the western Indian Ocean in 1998 following high temperatures in 1998. There was genuine concern that the same could happen in 2006. Almost needless to say, the long-term trends indicate that the concerns are, in fact, quite real, especially given the sensitivity of coral to temperature change.

It is extraordinary that Mr. Bolt would select a mere four years from the Global Coral Reef Monitoring Report and pick, apparently as verification, the words of a "veteran diver". Instead, one should actually refer to even basic material from the GCRMN itself which quite the opposite of the pollyanna reports provided by Mr. Bolt. Indeed, one of the lead authors of the Australian and PNG chapter of the GCRMN report disputes the basic claims of Mr. Bolt.

Actual scientific research shows that in over the last 30-40 years 80% of coral in the Caribbean has been destroyed and 50% in Indonesian waters and the Pacific. Globally about 1% of coral is dying out each year. This is a recent trend; with a number of studies showing no signs of bleaching over thousands of years.

So yes, our reef may very well die with current trends.

3. GOODBYE, NORTH POLE

Mr. Bolt starts with a criticism of Dr. David Barber of Manitoba University for saying "We're actually projecting this year that the North Pole may be free of ice for the first time" and points out that the pole has bee ice-free in the past (yes, there were forests in Antarctica as well). Context is everything and Dr. Barber was referring to in historical time, as was recognised in the original article where the claim was made. As the article notes, the comment was made because much of the North Pole ice is now annual rather than perennial ice.

Mr. Bolt also targets Flannery (evidently a man he loves to hate) for claiming "this may be the Arctic's first ice-free year". A modicum of research, which should be expected of a journalist, would have revealed that it was in fact "a senior adviser to the Norwegian government was warning that this may be the Arctic's first ice-free year". Flannery's opinion is the very next line: "As I write, the rate of loss, while still well above average, has slowed somewhat. Yet even now it's impossible to predict."

In an effort to bolster his argument, Mr. Bolt points out "the Arctic’s ice cover this year was almost 10 per cent above last year’s great low" and that "sea ice in the Southern Hemisphere has been increasing", because that has a great deal to do with the North Pole apparently. In any case once again Mr. Bolt is engaging in cherry-picking. Again, this must be countered by use of trends. The U.S. National Snow and Ice Data Center, for example, displays quite clearly what the trend actually is.

If it worth mentioning there is a modest increase as a trend with South Sea Ice, an increase in (thinner, weaker) annual ice due to the ozone hole affecting weather patterns; where the hole is not present (e.g., the Antarctic Peninsula) "the effects of warming have been dramatic".

So yes, the North Pole may very well become ice free with current trends.

4. BEWARE HUGE WINDS

In a short response, Mr. Bolt criticised Al Gore's global warming film, "An Inconvenient Truth" for have a poster that included a hurricane. To help Mr. Bolt out, it is also worth mentioning that some artistic license was given combining emissions with a cyclonic image in the poster. Criticism at Bill Clinton is levelled for saying "It is now generally recognised that while Al Gore and I were ridiculed, we were right about global warming ... It’s going to lead to more hurricanes." The Clinton quote is unsourced, so there's no way to check how the first sentence of the quote related to the second.

However the main point for Mr. Bolt is: "In fact, there is still no proof of a link between any warming and hurricanes". But this simply isn't true; there just isn't conclusive proof. The NOAA's Geophysical Fluid Dynamics Laboratory suggests a 5-12% increase because if sea level temperature increases by 2 degrees, because "Most hurricanes do not reach their maximum potential intensity before weakening over land or cooler ocean regions. However, those storms that do approach their upper-limit intensity are expected to be slightly stronger in the warmer climate due to the higher sea surface temperatures." The debate between Landsea and Emmanuel in Nature in 2005 on how significant this increase will be illustrates some main differences (see Nature 438, E11-E12, 22 December 2005 and Nature 436, 686-688, 2005)

Will global warming cause an increase in hurricanes? We just don't know yet. Half a point to Mr. Bolt. But it is foolish and dangerous not to prepare oneself for the possibility.

5. GIANT HAILSTONES WILL SMASH THROUGH YOUR ROOF

Also coming under criticism from Mr. Bolt is Professor Ross Garnaut. Apparently under criticism from neighbours for including a curved steel roof at the rear of his property, he responded: "Severe and more frequent hailstones will be a feature of this [climate] change". Again, this is unsourced. Mr. Bolt responds that the IPCC states that "decreases in hail frequency are simulated for Melbourne...", which indeed correct. Note however that the two are not contradictory; it is quite possible to have increased hailstorms as an overall result due to climate change and a reduced level in particular parts of the world (e.g., Australia). Ross Garnaut, if such a justification was made, was correct as a general comment, but incorrect in the specific example.

In this instance, Mr. Bolt was half correct.

6. NO MORE SKIING

Again with a predilection for cherry-picking, Mr. Bolt notes that a good ski season implies that a CSIRO student that suggests that resorts such as Mt. Hotham and Mt. Buller could lose a quarter of their snow by 2020 must be incorrect. The actual report makes rather different claims:

Under the low impact and the high impact scenarios respectively, the total alpine area with an average of at least one day of snow cover decreases 10-39% by 2020, and 22-85% by 2050. The area with at least 30 days of snow cover decreases 14-54% by 2020, and 30-93% by 2050. The area with at least 60 days of cover shrinks 18-60% by 2020, and 38-96% by 2050.

At all sites, the low impact scenario for 2020 only has a minor impact on snow conditions. Average season lengths are reduced by around five days.

More seriously however (and Mr. Bolt is finally using real science) "..a study this year in the Hydrological Sciences Journal checked six climate models, including one used by the CSIRO... It found they couldn’t even predict the regional climate we’d had already: 'Local model projections cannot be credible' ... ". Except when one reads the actual paper that contains the quote used by Mr. Bolt, the snowfall study is not used. Indeed, the only local model projection in Australia that was used was from Alice Springs - a place not known for its snowfall. Further the paper does conclude that "the unsatisfactory state of the art in climatic (and hydrological) future projections does not reflect a general deadlock in related sciences, but only a wrong direction. Causality in climate and hydrology is not sequential and one-to-one but rather circular (due to feedbacks) and many-to-many (due to complexity). Such causality can be better described in probabilistic and stochastic terms (see Suppes, 1970), rather than in terms of the current deterministic climatic models and practices (see also Giorgi, 2005)."

Referring back to the originally cited CSIRO study, there are multiple references to a decline in snow depth and snow days. The hyperbolic title "No More Skiing", is indeed true, and only true, from the perspective of a 2050 high impact scenario.

7. PERTH WILL BAKE DRY

In this example, Mr. Bolt claims that CSIRO's models and warnings cannot be trusted as it claimed that Perth was "'particularly vulnerable' and had a 90 per cent chance of getting less rain and higher temperatures." In rejecting this proposition, Mr. Bolt points out that Perth "has just had its coldest and wettest November since 1991".

This remark displays extraordinary ignorance in the scientific method. Nobody should ever suggest that a single wet and cold month serves as a genuine counter to long-term evidence and trends - and nor should they suggest a single dry and hot month does so either.

A previous link has already referred to the extremely serious state of Perth's water supply. The Bureau of Meteorology explains in their 2010 report the state of Perth's temperatures (along with a wealth of other data).

If current trends continue, Perth will indeed bake dry.

8. ISLANDS WILL DROWN

Mr. Bolt claims that ABC Science Show host Robyn Williams that sea level will rise by 100m by 2100. Given that Andrew Bolt was part of the conversation his recollection could be improved. The transcript says something different:

Andrew Bolt: Dean of science...suggesting rising seas this next century of up to 100 metres, or Al Gore six metres. When I see things like that I know these are false. You mentioned the IPCC report; that suggests, at worst on best scenarios, 59 centimetres.
Robyn Williams: Well, whether you take the surge or whether you take the actual average rise are different things.
Andrew Bolt: I ask you, Robyn, 100 metres in the next century...do you really think that?
Robyn Williams: It is possible, yes.

It is possible (not even probable) that in the next century (i.e., by 2200) that sea levels could rise by 100m. Indeed, the potential loss is indeed approximately 80m.

Surprisingly, Mr. Bolt claims that "the seas have risen slowly since the last age", where instead we see a rather sudden increase (in geological time) of over hundred metres. From 3,000 years ago to the start of the 19th century sea level was almost constant, rising at 0.1 to 0.2 mm per annum; since 1900 the level has risen at 1 to 3 mm per annum; since 1992 satellite altimetry indicates a rate of about 3 mm per annum. In the IPCC's Fourth Assessment Report in 2007 the sea level is projected to rise by 18 to 59 cm by 2090-2099. Countries like the Maldives have an average ground level of 1.5m above sea level.

It is important to understand the difference between "possible" and what is likely. It is possible that sea levels could rise 80m by 2200. It is likely that they will rise between 18 and 59cm by 2099. This enough to cause serious problems to an existing country of over three hundred thousand people. Islands will indeed drown.

9. BRITAIN WILL SWELTER

Mr. Bolt considers the failure of the British Met Office to predict the 2008 winter weather as a sign that the Hadley Centre is unable to predict future climate. "If the Met can’t predict the weather three months out, what can it know of the climate 100 years hence?", he asks.

Behind this folksy humour there is a profound lack of knowledge in even the basics of meteorological prediction. Meteorologists are often incorrect in immediate predictions because of the number of variables, complexity of those variables, error involved in measuring the initial conditions, and an incomplete understanding of atmospheric processes. Weather prediction is, of course, an improving science but occasionally mistakes will be made. Sure beats the days of staring at the sky and consulting astrologers however.

Will Britain swelter? The Met provides very good historical records which does indicate that the temperature of most of the UK is indeed on the rise, albeit slowly. Is it accurate to claim that because the Met predictions are not always correct in the short-term that long-terms estimates are wrong? No, of course not.

10. WE’LL BE HOTTER

Mr. Bolt further criticises the met for predicting that 2001, 2002, 2004, 2005 and 2007 for being hotter than it actually was. He further claims that 1998 was the hottest year on record since the Medieval Warm Period and that "temperatures have slowly fallen since around 2002". The supposed lesson is that "[s]omething is wrong with warming models that predict warming in a cooling world, especially when we're each year pumping out even more greenhouse gases."

The Met making incorrect forecasts is not particularly relevant or surprising for the reasons given in the last point. It is interesting however that Mr. Bolt was quite incorrect to suggest that 1998 was the hottest year on record when in fact of the three global temperature records (HadCRUT3, NASA GISS and NCDC), only HadCRUT3 actually shows 1998 as the hottest year on record. For NASA GISS and NCDC, the hottest year on record was 2005. Further it is quite clear that global temperatures are still increasing in the longer run and from 1998 onwards with the 2010 NASA GISS data indicating an even higher global average. Obviously this means that multiple years have been hotter than the localised medieval warm period.

Yes, it is getting hotter.

Concluding Remarks

It is unsurprising that Mr. Bolt is not well versed in scientific literature, or more particularly the scientific method as this is not an area which has investigated on a formal or informal level. For example, it is basic in science that one should not cherry-pick particular examples as universal examples, a pars pro toto fallacy. Instead longer-term trends should be sought.

What is more surprising is the numerous occasions that Mr. Bolt engaged in false attributions and misrepresented qualifiers. One would expect as a fully-paid member of the chattering class that he would at least have a better level of reading comprehension than what was displayed. These mistakes can either be attributed to political partisanship or poor journalism. In either case it certainly reduces the trustworthiness of Mr. Bolt.

On items 4 and 5 Mr. Bolt is certainly partially correct, giving a grand total of 1.5 out 10 for the statements on his part. What one may make of his employer continuing to reward such poor performance is a matter deserving considered opinion.