I’ve got this interesting piece of data – it’s polling data measuring the proportion of the public that utilised a particular government service in country X in the previous month, and a particularly important government service at that. The polls were taken fortnightly for a period of around 5 years, where the results were seasonally adjusted.

The other interesting thing is that the five year period involved is over the first 5 years of a particular government. When that government first came to power, a long term government study had the usage rate at officially 50% to the nearest decimal point – which makes it pretty easy to trace the change in demand for this service over time.

All up there were 6 pollsters measuring fortnightly, giving us 131 polls per pollster, for a total of 786 separate polling observations.

First up, it’s worth noting that there were house effects involved here.

housebias1 housebias2

Pollsters A, B and C picked up a consistently higher level of use than Pollsters D, E and F – yet, all pollsters “moved together” through time, approximately all of the time.

We can see that by throwing them all in the same chart.

psuse1

It’s pretty easy to determine whether demand for this service is growing over time, and what any local trend might be by running a linear regression for the former, and a LOESS regression for the later. Once we do, the trends become pretty damn obvious.

psuse2

That linear regression is statistically significant to so many decimal places that I had to go into the guts of my stats software to find a value that wasn’t zero. The p-value of the linear regression is 0.000001 (rounded up), meaning that there is less than a 1 in one million chance that the long term increase in demand we are witnessing could be due to chance alone. It’s a slow grinding increase, but a consistent one regardless of the local, short term variation involved.

If you were a government facing that kind of data, there would be absolutely no hesitation at all, on either side of politics, to plan for a future which contains that increasing demand for this particular service. There might be political differences over how to do it, maybe even on what ought to be done – but there would be universal acceptance of the empirical reality of increasing demand, albeit one that contains significant variation at any given time. Similarly, if you were a business or an industry facing that sort of data on the demand for your own product, there would be universal action to deal with it. In fact – you would struggle to find anyone that would suggest such data doesn’t have an observable and statistically significant linear trend over time strong enough to warrant acting on.

Yet – the data above isn’t actually polling data about demand for a government service. It’s temperature data from 1999 through to November 2009 that measures the monthly deviation from the long term temperature average of 1979 to 1998. You would have most likely seen the data stated like this:

uah2

The bloke who makes this chart only uses the global average series (and he’s not to be confused with the other bloke that regularly posts this chart – the fella that’s worried about people sticking calculators up his bum), whereas I used hemispherical data.

Pollsters A, B and C are Northern Hemisphere, Northern Hemisphere land and Northern Hemisphere ocean average departures respectively, while Pollsters D, E and F are Southern Hemisphere, South Hemisphere land and Southern Hemisphere ocean average departures. You can see the raw data set here.

All I did was change the scale by making each observation =50 + (15 x actual temperature deviation);  a transformation needed to make it appear like polling data. The regression work doesnt change a bit.

What these data points represent is the departure of any month’s temperature from the long term average value taken between 79-98.

This is what the raw data of the 6 series I used from 1979 through to 2009 looks like:

uah1

I merely looked at the period from 1999 onwards – the period after the reference period, to the right of the blue vertical line.

For any other data series showing exactly the same behaviour, no one would be arguing against the long term trend and the need to act on it for their industry or government. But climate change skepticism isn’t about the data – it’s never been about the data. For most skeptics, especially those that are professionally unqualified in any form of quantitative science or statistics (i.e. most of them), it’s just about the politics.

Why we’re supposed to listen to these people in the name of balance is beyond me – ignorance isn’t a balance against empirically observable and statistically measurable reality.

When we discuss politics, do we make sure there’s always a person in the debate that represents the viewpoint that there’s a secret one world government ruled by aliens?

When we discuss bushfire management, do we always make sure there’s some bloke there banging on about  phlogiston?

Of course we dont – we treat it like the horsefluff it is. Yet, for some reason when it comes to climate change, the mainstream media (including the ABC, an organisation that should know better) regularly include some bonehead in the debate that reckons temperature isn’t increasing – some bonehead that can’t move beyond a first base that reality passed many, many years ago.

There are plenty of angles in the climate change debate where there is disagreement and where arguments and empirical evidence for different views do exist – from the magnitude of mans contribution to global warming through to the benefits vs costs of any and all action and the kinds of action that might involve.

But low rent denialism – it’s cheap fiction. It’s time we all told it to fuck off.