Slanting Science

Science is supposed to be unbiased, seeking to understand the workings of nature untainted by the personal beliefs or prejudices of its practitioners. Nature alone is the arbiter of truth—when science and nature disagree it is science that is wrong. But science is practiced by human beings, who cannot keep their beliefs, whether engendered by religious, philosophical, or political leanings, from skewing any result that is equivocal or highly complex. Presented here are two examples taken from the pages of Nature, perhaps the world's primer scientific journal. One is a rehash of temperature history in northern latitudes with a new statistical twist, the other a report on a study regarding fracking. One shows how what scientists leave out of their studies may be more important than what they put in. The other shows how a headline can spin the results of a report even when its authors are carefully neutral in their conclusions.

In “Recent temperature extremes at high northern latitudes unprecedented in the past 600 years,” Martin P. Tingley and Peter Huybers, both from Harvard University's Department of Earth and Planetary Sciences, present a statistical analysis of historical temperature data. Most of these data are inferred by proxies. The arresting title is obviously intended to bolster the climate catastrophists' claims that anthropogenic global warming is harming the environment. To their credit, the authors do not include the gratuitous reference to AGW that is almost universally found in such papers today. The background and motivation are presented in the paper's abstract quoted below.

Recently observed extreme temperatures at high northern latitudes are rare by definition, making the longer time span afforded by climate proxies important for assessing how the frequency of such extremes may be changing. Previous reconstructions of past temperature variability have demonstrated that recent warmth is anomalous relative to preceding centuries or millennia, but extreme events can be more thoroughly evaluated using a spatially resolved approach that provides an ensemble of possible temperature histories. Here, using a hierarchical Bayesian analysis of instrumental, tree-ring, ice-core and lake-sediment records, we show that the magnitude and frequency of recent warm temperature extremes at high northern latitudes are unprecedented in the past 600 years. The summers of 2005, 2007, 2010 and 2011 were warmer than those of all prior years back to 1400 (probability P > 0.95), in terms of the spatial average. The summer of 2010 was the warmest in the previous 600 years in western Russia (P > 0.99) and probably the warmest in western Greenland and the Canadian Arctic as well (P > 0.90). These and other recent extremes greatly exceed those expected from a stationary climate, but can be understood as resulting from constant space–time variability about an increased mean temperature.

The authors then cite past palaeoclimate reconstructions, which demonstrate that these temperatures are anomalous. They claim, however, that ascertaining whether such extremes are unprecedented requires a treatment beyond that supported by standard measures of uncertainty. They propose to bolster such claims using Bayesian analysis, a statistical method that has been quite the rage recently. This is partially because Nate Silver, heralded prognosticator of election results and baseball statistics, recently announced a move from the New York Times to a major sports network. Silver has become America’s top prediction guru, fawned over in TV interviews and generating an entertaining best-seller, “The Signal and the Noise.” in which he discusses the challenges and science of prediction in a wide range of domains, covering politics, sports, earthquakes, epidemics, economics, and climate change.

Bayesian statistics is a subset of the field of statistics in which evidence of the true state of the world is expressed in terms of degrees of belief or preexisting probabilities. It was the brainchild of the eponymous Thomas Bayes. In 1763, Bayes published a paper on the problem of induction. In it he introduced his famous mathematical formula stating that for two random quantities y and θ ,

p(θ|y)=p(y|θ)p(θ)/p(y),

where p(⋅) denotes a probability distribution, and p(⋅|⋅) a conditional distribution. When y represents data and θ represents parameters in a statistical model, Bayes Theorem provides the basis for Bayesian inference. Unsurprisingly, the Bayesian approach is much less helpful when there is no consensus about what the prior probabilities should be. As Radford M. Neal, Professor and Canada Research Chair in Statistics and Machine Learning at the University of Toronto's Deptpartments of Statistics and Computer Science, states: “Unfortunately, many "Bayesians" don't really think in true Bayesian terms. One can therefore find many pseudo-Bayesian procedures in the literature, in which models and priors are used that cannot be taken seriously as expressions of prior belief.”

That not withstanding, Tingley and Huybers have taken the Baysean plunge and cranked out a new analysis of extreme temperatures in northern latitudes. Their study seems to reinforce the dire warnings of the most strident warmists. In the paper they provide this explanation of their methodology:

The probability that a given interval contains the most extreme temperature in the span of a reconstruction is generally quantified using ensemble-based reconstruction methods because they can be used to estimate simultaneous, or pathwise, uncertainty intervals, and provide direct probabilistic assessments of extremes. Ensemble-based reconstructions have recently been used to evaluate extremes in spatially averaged temperature using bootstrap methods, Bayesian principal component regression and realizations drawn from global climate model simulations

The problem here is that they depend on multiple, dispirit proxy data sets, all of which have been “corrected” by researchers before. They then munged ice-core, tree-ring and varve data together with more recent instrument based information and used that to drive a climate model. Combining the output of this “ensemble” creates a temperature curve that looks remarkably familiar to anyone who has followed the climate change debate for the past decade or so.

That's right, it is our old friend the hockey stick, a soothing sigil to all hidebound climate change true believers. It is totally unsurprising that a couple of researchers would take old data, inaccurate and unreliable for all the reasons presented in critiques of older papers; use that to drive new models that, if anything, incorporate more of the authors' preconceived expectations than older programs; and the end result is the same meaningless pap climate science has been foisting off on an unsuspecting public for thirty years. The only thing this paper can state with real confidence is that it appears that the world has been warming recently compared with the past 600 years.

That brings us to perhaps the most egregious deception used in this paper. Notice how the time-line starts at around 1400? Care to ask why? Because prior to 1400 the world was in the noted warm period known as the Medieval Warm Period. Despite substantial uncertainties, the warmest period of the last 2,000 years very likely occurred between 950 and 1100. Things were still rather balmy in the Northern Hemisphere until the climate descended into the Little Ice Age, from which it is still recovering. Pick your data, pick your time span and pick a modeling methodology sure to take your biases into account and magically you get the results you expected. This exercise is simply an attempt to borrow some of Bayesian statistics' current cachet to prop up climate change's faltering fortunes.

Moreover, Silver and the Bayesian methodology he promotes are not without their critics. Gary Marcus and Ernest Davis, both professors at New York University, recently penned a piece in The New Yorker online. In it they capture the fundamental problem with both Silver's book and Tingley and Huybers' paper. “Switching to a Bayesian method of evaluating statistics will not fix the underlying problems; cleaning up science requires changes to the way in which scientific research is done and evaluated, not just a new formula.” Amen, brothers.

You must be fracking joking

The second example of science that has been through the spin cycle also comes from Nature. In an article with the strident title, “Gas drilling taints groundwater,” Nature staff writer Jeff Tollefson cites a recent study done by researchers at Duke University and published in PNAS. Judging from the title, the report must have conclusively found that drilling for natural gas—let alone hydraulic fracturing, or fracking—must be unavoidably detrimental to local groundwater supplies. But in “Increased stray gas abundance in a subset of drinking water wells near Marcellus shale gas extraction,” Robert B. Jackson et al. make no such absolute proclamations.

Based on isotropic analysis of the methane, which is naturally found in the areas drinking water, the report concluded: “Overall, our data suggest that some homeowners living <1 km from gas wells have drinking water contaminated with stray gases.” Yet there are many other choice items in this “latest salvo” in the ongoing debate about shale-gas extraction and the environment.

  • Trace amounts of methane are often found in drinking water.
  • The team did not find evidence that chemicals used in fracking migrated from depth to contaminate aquifers.
  • The data also suggest that some gas migrated up from geological layers between the Marcellus Formation and the groundwater table.
  • The levels identified by Jackson’s team do not represent a health threat.

So, while some of the methane is probably from leaks in well casings, other methane could have found its way into the aquifer naturally and still other methane is present from biological sources. And regardless of its source, the traces of methane found do not pose a health risk.

Jackson says that his results do not necessarily mean that all drilling operations will have problems. Other research by his team, now in the press in Applied Geochemistry, found no evidence of contamination in a shale formation in Arkansas. More importantly, he says, the results suggest that the problem is relatively simple to fix.

This has not stopped excitable news reporters from linking fracking to flaming faucets, popular fodder for You Tube. Since most news organizations have the attention span of a goldfish practically none have investigated the flaming faucet phenomenon sufficiently. If they had they would have discovered that methane migration has been setting faucets on fire for decades. Given that Hollywood has turned flaming explosions into a major revenue source, is it any wonder that Matt Damon couldn't resist spreading the lie that fracking causes sinks across the nation to burst into flames?


Col. Drake's 1859 oil well in Titusville, Pennsylvania.

In reality, having your well water burst into flame is nature's way of saying, “drill here for natural gas and oil, you fool!” There are many places where natural gas percolates up through the ground from deep formations. Sometimes, great discoveries are made by the marginally observant. It is not surprising that the world's first oil well was drilled in Pennsylvania, not Texas.

There are many things worse than having trace amounts of methane in your groundwater. In many parts of the world groundwater is much more dangerous to human health than the groundwater in northern Pennsylvania. A basement carved out of granite can fill you house with levels of radon gas that can give residents lung cancer. Breathing the air in any major city in China will certainly shorten your life. Why did this mild mannered report deserve the threatening headline?

These days you cannot go wrong attacking the evil energy industry—at least as far as the eco-lobby and the mindless minions of Hollywood and the EPA are concerned. What fools they are. Natural gas is the cleanest of all fossil fuels. It is a net positive because it pollutes less and allows more harmful energy sources to be eliminated. And from a purely nationalistic point of view, it is domestically produced, lessening dependence on oil shipped from unstable parts of the world. This headline was nothing more than a cheep shot by a liberal hack. Nature should be ashamed of itself.

Be safe, enjoy the interglacial and stay skeptical.