If you're going to insist on following the scientific 'consensus', you might note that a fundamental consensus, and one of very long standing, is that you don't accept assertions like 'last year was the hottest on record' unless the measured temperature is at least three standard deviations above the previous high. It isn't.

In rhetoric, ad hominem can be effective. In science, it's inadmissible.

Agreed, we should make sensible plans to reduce emissions. If you want to focus on CO2, fine. In that case, you need to consider alternatives to coal for base-load power. PV solar and wind are not alternatives for base-load power. Methane, coal-seam gas, and nuclear are the only alternatives which can deliver base-load power (I'm not sure whether to include 'ckean coal' or not, but my guess would be it's no nearer than geothermal to commercial viability, hence in the too-hard basket for the next few years at least). Methane, coal-seam gas and nuclear technologies won't attract investment to replace coal without explicit government intervention. That intervention is what Abbott calls 'direct action'. Dis it if you like, but unless and until you show how a Carbon Tax will drive replacement of coal-fired base-load power by cleaner technologies, WITHOUT government intervention of any sort, I'd suggest your thinking isn't sufficiently ... scientific.

Just out of curiosity, donkeygod where do you get this information?

"If you're going to insist on following the scientific 'consensus', you might note that a fundamental consensus, and one of very long standing, is that you don't accept assertions like 'last year was the hottest on record' unless the measured temperature is at least three standard deviations above the previous high. It isn't."

Where is this long standing 'consensus'? Where did you get the standard deviation calculations information from?

From my understanding, isn't standard deviation just a test of statistical significance, assuming a Gaussian distribution? I think you can say an average is higher, without it necessarily being statistically different.

Hi, Bugsy!

The GISS-NASA press release in January 2011 said:

“Global surface temperatures in 2010 tied 2005 as the warmest on record, according to an analysis released Wednesday by researchers at NASA's Goddard Institute for Space Studies (GISS) in New York. The two years differed by less than 0.018 degrees Fahrenheit. The difference is smaller than the uncertainty in comparing the temperatures of recent years, putting them into a statistical tie.”

That’s a roundabout way of saying that the difference is statistically ‘insignificant’. Doesn’t mean the observation is useless: they’re saying 2005 and 2010 temperatures fall within a bracket four standard deviations wide. No, you certainly can’t say one average is ‘higher’ than another without showing statistical significance. If you could, then if I tossed a coin and heads came up three times running, I’d have ‘proven’ it was biased ... even though by tossing it 30 times more I’d ‘prove’ it was NOT biased. You do need to know something about data distribution, but it needn’t be Gaussian to estimate significance.

Estimating global temperature is a horrific problem, and the statistical implications are wicked. NASA’s records run from 1880, but of course 100 yrs ago far fewer measurements were made, the technology was different, and the environment was VERY different, i.e., much smaller urban heat-island effects (UHI). Comparing such different data sets dramatically increases uncertainty. UHI adjustments are particularly contentious — but that’s a whole book in itself. Then there’s the problem of estimating ‘global surface temperature’ based on a very biased system of measurements: urban areas are hugely over-represented, while much larger non-urban areas (Russia, Africa, South America, THE SEA!) are grossly under-represented. Adjustments can be greater than the purported effects. If so, what you have may be suggestive, even useful, but it's not yet science.

It’s a huge subject, and 350 words won’t begin to even list the challenges. But, having worked for some decades as a scientist & engineer, I can promise you that significance testing IS de rigeur. If you don’t have it, and can’t get it, don’t believe it!

Oh, so it was the EQUAL hottest on record.

That's a lot of words protesting the omission of one word.

robbyh "Amicus,

You didn't hear correctly. Hundreds of years? Why then was last year the hottest on record?"

What's your point?

Tim Flannery said it not me, like many others you attack the messenger, since you don't like the message .. go ask Tim, I heard it and read the transcript.

This is a natural thing going on, it's not a "problem", it may be contributed to by man's land clearing and other activities, that's progress, but I do not subscribe to the view that taxing CO2 in Australia is the "solution", to a non problem.

The difficulty for you alarmists is that now people feel they need do nothing else about pollution or reducing energy use, after all, the government has told us it will all be solved with a tax .. so I'll pay the tax, then go on consuming as normal.

It's like you think Labor in power will not add to taxes, of course they do, they will find a way.

As to the hottest year syndrome, if this year is not as hot as last year, what will you do?

If it's cooler, we can all go "phew, dodged a bullet there eh" yes? Our tax must be working, yes?

What utter rubbish, you try to exploit convenient events to support your belief, that's all it is .. geologically one year is irrelevant.

I can imagine in the church of AGW though everyone will be praying for another hot year, to "get" those sinners who don't believe .. they will get their comeuppance! Repent ye fools!

I also do not offer a solution, as there is no problem (to those who insist solutions MUST be offered, bonmot, and we all look forward to your predictions and solutions).

No, bugsy. It's not 'equal' hottest. GISS-NASA have calculated a 0.18 F difference between 2005 and 2010. That's a best case, because they're using more recent, more copious, and more accurate data. Even so, they've clearly stated that their calculations STILL aren't accurate enough to distinguish a change that small from background noise. If calculated global surface temperatures for the last 130 years have, say, an average standard deviation of 0.5 F (and it'd be nowhere near that low for the early data), then there may well be 60+ years in that time period which, by your criteria, are 'equal hottest'. If five of those 'equal hottest' years happened to fall in the 1930s, would you conclude that the world is cooler now than it was then? You wouldn’t, of course, but if you did, you'd be quite wrong.

This is the problem we run into when statistical significance isn’t part of the discussion. You can sometimes find out more if you check out the original sources, but nowhere near often enough. Physicists, drug manufacturers, even economists use scientifically valid statistical analysis to support their data; it’s quite often required by law. Climate scientists, admittedly, face special problems in establishing statistical significance, but that doesn’t excuse pretending that those problems don’t exist. That’s what was so terribly wrong with Michael Mann’s ‘Hockey Stick‘ — he used proxies rather than temperatures, cherrypicked his data, and used methods which were easily shown to be totally lacking in statistical significance. And got away with it, for far too long.

I’m not arguing against AGW here. I’m just pointing out that there are established, demonstrably valid criteria for determining whether some scientific observations have a better-than-even chance of being true. When it’s feasible to use them, scientists are professionally and ethically obliged to do so. If journalists and activists don’t want to bother, that’s their call. But I notice, and form my judgements accordingly.