Bunyip’s failed Flannery gotcha

The Australian government’s Climate Commission this week released a controversial report projecting “impacts of climate change on the health of Australians.”

The report was launched by Chief Commissioner Tim Flannery, which was duly reported in The Australian. From among the report’s projections, the paper reported:

Without international action on climate change to limit temperature rises to [2 degrees celsius], the number of predicted temperature related deaths in Australia is predicted to rise from just over 6000 in 2020 to about 10,000 in 2070.

Admirably taking this statement with a sceptical grain of salt, blogger Professor Bunyip armed himself with “a pocket calculator and a copy of some recent Bureau of Statics [sic] projections” to set about testing the Climate Commission’s predictions.

In 2020, according to the ABS, Australia will be home to some 30 million people, of which Flannery insists roughly 6000 will be carried off by dengue fever and other curses that thrive in the heat. By 2070, the same ABS projection posits a likely population of between 46 million and 54 million, depending on which curve you choose to track.

So let’s see how that works out: 6000 deaths per 30 million means a 1-in-5000 chance of being done in by nasty weather as of 2020.

And 10,000 deaths in a 2070 population of 54 million? Well that comes in at 1-in-5400 climate casualties.

So the warmer it is, at least by Flannery’s reckoning, the safer and healthier we will be.

Neat... except Bunyip has made a couple of erroneous assumptions — one trivial, and one that’s somewhat more serious.

Trivially, he attributes the “reckoning” to Tim Flannery, when in fact the report’s authors are Lesley Hughes and Tony McMichael. (Hughes is a commissioner with the Climate Commission, and head of the Department of Biological Sciences at Macquarie University. McMichael is professor of population health at Australian National University.) True, Flannery as Chief Commissioner did launch the report, but the centrality attributed to Flannery by Bunyip and his fellow travellers is about an agenda.

More serious, however, is Bunyip’s simplistic (albeit pocket calculator-friendly) assumption that the health-related effects of hotter temperatures will be linearly uniform across Australia.

Whereas the modelling from Hughes & McMichael is somewhat more complex:

Under a worst-case scenario, unmitigated climate change may modestly reduce temperature-related deaths in Victoria, Tasmania, South Australia and NSW . . . but deaths could increase markedly in Queensland and the Northern Territory.

That’s an actual quote from the report that was in the article Bunyip relied upon for the above, so he has no excuse other than agenda-driven zeal.

So, Bunyip will need to factor into his calculations some sort of inverse differential between Qld/NT and the other states and territories, based on population projections for each, factoring in some assumptions about relative risk in each, etc. etc. etc.

The take-home message of Bunyip’s foolishness is that modelling of complex systems is ... well, complex ... and requires somewhat more than a pocket calculator and an agenda.

Oh, and beware when someone like fact-averse Andrew Bolt thinks well of you.

UPDATE 06/12

Bunyip has rebutted my remarks here in an update to his post, in which I’ve been soundly bitch-slapped, had my arse ass handed to me on a plate, whilst simultaneously having it kicked, on my way out the door. That’s fixed my little red wagon, yessah, and I won’t be dragging my arse ass back in a sling for more, nossah!

Kidding. He called me some names and had a little whine about standards.