The broken hockeystick

The arguments against the hockey stick graph are stacking up, so I thought I’d summarise them.

Tree rings:

Most hockeystick graphs use tree rings as a proxy for temperature, seeing as how thermometers didn’t exist a thousand years ago. This relies on the assumption that there is a good, linear correlation between temperature and the width of the corresponding tree ring. Is temperature the only factor influencing the growth of a tree? Sunlight, soil salinity, pH, nutrient level, the presence of heavy metals, rainfall and water availability can all influence the growth of a tree, not to mention the carbon dioxide level. Most skeptics accept that this has risen with mostly anthropogenic causes. Some hockeysticks could simply be measuring the carbon dioxide level in the atmosphere, while other factors muddy-up the correlation with temperature.

Statistical method:

It took a retired mathematician to actually take a critical look at the original hockeystick graph. Steve McIntyre and colleague Ross McKitrick attempted to identify the statistical method behind the hockeystick, and after numerous FOI requests, which were constantly fended off, they found some shocking truths. Feeding random data into Mann’s program invariably resulted in a hockeystick shape, but as we know science has to be repeatable. McIntyre and McKitrick could not replicate Mann et al, but their experiment could be repeated. Climategate 2 reveals how Rob Wilson tried to do so…and suceeded. Furthermore an independent team of statisticians, led by Edward Wegman, found McIntyre and McKitrick’s results to be “valid and compelling”.

The “decline”:

When I first heard about “Mike’s Nature trick”, I thought it was some complicated statistical technique. Then I learnt that he just took an eraser to the last twenty years of his proxy series and stuck in the instrumental record. He did this to “hide the decline” that his graph showed in the temperature. This is confusing because we know that in this same period temperatures were rising, so which is right; the trees or the thermometers? Mann contends that despite this divergence, his data remains an accurate description of temperatures prior to this decline (which he felt inconvenient enough to “hide”). I don’t know about Mike, but when I was in high school, when I had an unexpected result, my science teachers told me to explain it, rather than leave it out.

Evidence of the Medieval Warm Period:

For a long time, the idea of a prominent medieval warm period and little ice age was the unchallenged consensus in the geological community. This was until 1998, when Mann, Bradley and Hughes invented the first hockeystick, and the warmist community accepted it at face value, without looking into the statistical methods employed and overlooking the fact that it conflicted with years of solid work in paleo-reconstructions. So which is a more accurate description of the global climate over the past thousand years; the hockeystick graph or the sinusoidal-like graph showing the medieval warm period? Loehle uses 18 non-tree rings proxies to corroborate a prominent medieval warm period, a view which CO2 science finds to be the dominant one in the peer-reviewed literature.

Lack of agreement among graphs:

Tom Wigley in the second round of climategate emails notices an issue with the hockeystick theory that he mysteriously decides to keep within the team. While most hockeysticks show agreement on the temperatures post-1850 with the sudden rise, they conflict on the temperature prior to the industrial revolution. Compare this to the 18 non-tree ring proxies compiled by Craig Loehle, which finds remarkable conformity over 2000 years.

Conflict with high climate sensitivity:

Warmists argue that we have a highly sensitive climate, meaning even minor changes can result in drastic climate change. While the rise post-1850 would support this contention, it would seem that this would come into conflict with the first part of the graph, where temperatures are shown to be remarkably steady. Carbon dioxide on its own is a relatively minor forcing factor, a doubling of which causes only about a degree of warming directly. It is implausible that a similar natural forcing could not have led to a similar rise in temperatures to the one we see post-1850. In this way, the hockeystick is a self-contradicting paradox.