The Hockey Stick Poster Child

The Swiss Federal Institute for Forest, Snow, and Landscape Research — Wald, Schnee und Landschaft (WSL) in German — sits on top of a hill in Birmensdorf, just outside Zurich. Tree-ring research first became part of WSL’s research mission in 1971, when Fritz Schweingruber started his work there. Fritz is a botanist, an archeologist, and one of the world’s finest wood anatomists. Early in his research career he became fascinated with tree rings, and under Fritz’s leadership the WSL dendrochronology group grew to be the European equivalent of the Laboratory of Tree-Ring Research (LTRR) in Tucson.

When I started working at the WSL in 2007, after a two-year stint at Penn State, Fritz had retired, but he still visited the dendro lab a few times a week and his influence was omnipresent. Jan Esper, Fritz’s protégé, who was later involved in the discovery of Adonis, had taken over the leadership of the lab, gathering a productive group of dendroclimate researchers around him.

During my first summer at the WSL, Jan and I, along with two of these talented up-and-comers, David Frank and Ulf Büntgen,* traveled to the Spanish Pyrenees to collect tree-ring samples. The site we sampled was almost at 8,000 feet elevation, on the mountain slopes surrounding Lake Gerber in the Aiguestortes National Park. When I asked why we were sampling this particular site, the answer was simply that Fritz had recommended it. Apparently, while traveling through the Pyrenees, Fritz had spotted the site through binoculars from a road hundreds of meters below. Fritz’s renown for having an eye for old trees was well founded: as it turns out, the Lake Gerber tree-ring chronology, made up of living mountain pine (Pinus uncinata) trees and remnant wood, goes back more than a thousand years.

The Pyrenees field campaign was a swift introduction to the efficiency with which my WSL colleagues operated. We flew from Zurich to Barcelona, then the same night drove three and a half hours to the small town of Viella. At eight o’clock the next morning, we stopped for provisions and then hiked three hours straight up the mountain to the lake. Go-getters that they were, Ulf and David practically sprinted up the mountain, while Jan and I followed at a more reasonable pace. By the time we made it to the lake, it was noon and I figured we would stop for lunch before starting to core. But that was not how these guys rolled. The moment the first old tree came into sight, David started coring, with Jan quickly following suit. For the next few hours, David, Jan, and I cored one tree after another, while Ulf ran from one of us to the other collecting cores and then labeling and storing them. Finally, around three in the afternoon I flatly refused to carry on until we had lunch.

Reluctantly, my three companions agreed to take a quick break, after which we continued coring relentlessly until the sun began to set and we had to rush back down the mountain to beat the dark. The exact same thing happened on each of the following days: we would labor without respite until, famished, I refused to core any more. One evening toward the end of the week I finally asked the guys what was up with the unceasing pace, and at last the truth came out. All three of my colleagues admitted to being relieved when I demanded that we eat each day, because none of them wanted to be the first to let on that he was starving.

As a woman scientist, I got 99 problems, but at least starving or freezing to death to protect my ego ain’t one.

Apparently, before I, the first woman to join them on a campaign, came along and insisted on sense, their testosterone-driven stubbornness resulted in whole days of intense physical effort in the field, entirely without food. Thereafter, the confessions got even better. It turned out that none of them wanted to be the first to concede that he was cold at night, and so they slept in their shared room with the windows wide open even when temperatures in the mountains dropped to near freezing. At that moment, I was very happy with my gender. As a woman scientist, I got 99 problems, but at least starving or freezing to death to protect my ego ain’t one.

The research focus of the WSL dendroclimate team was to use tree rings to reconstruct climate over past centuries. To study past climate prior to the early twentieth-century start of the instrumental climate record — the meteorological data derived from daily measurements at weather stations around the world — we make use of paleoclimate proxies.

These biological or geological archives, such as the layers in ice cores, lake sediments, trees, and corals, record climate conditions and can therefore be used as sources of climate information. Tree-ring records provide good value for money in that context. They are relatively easy and cheap to acquire and analyze. With trees and forests covering large swaths of the earth’s surface, tree-ring data are some of the most commonly used climate proxies, especially for the most recent 1,000 to 2,000 years, the period with the most dendrochronological data.

In that context, our Lake Gerber field campaign was aimed at developing a long-term climate reconstruction for the Pyrenees. With a 1,000-plus-year tree-ring chronology in hand, our odds for achieving this goal initially looked good. But as we discovered, mountain pine trees in the Pyrenees are sensitive to a mix of limiting climate factors. The trees grow at high elevation and are thus limited by cold temperatures that persist even in summer. But they also grow in the seasonally dry Mediterranean region and so are also limited by a lack of summer rain. As a result, the pines form narrow rings both when it is cold and when it is dry. Their narrow rings can indicate cold summers or dry summers, and their yearly ring width cannot be used to reliably reconstruct either temperature or precipitation independently. It was looking as if our hard-earned tree-ring data were not going to pay off in a Pyrenees climate record after all.

The width (or density) of annual rings can be measured, absolutely dated, and compared on a year-by-year basis with instrumental weather-station data

Luckily, we can also measure other parameters in tree rings, apart from their width. Using tree-ring radiometry, for instance, we can measure the wood density of individual rings, which often captures summer temperature variations better than the ring’s width. The maximum density of the latewood portion of a ring, in particular, reflects how much the cell walls in that ring have thickened by the end of the growing season. This in turn is strongly determined by the temperature that the tree has experienced during its growing season: trees form denser latewood in hot summers than in cold summers. Therefore, maximum latewood density provides a very good record of summer temperature in the year that the ring was formed, and wood-density measurements can be used as proxies for past summer temperatures.

Like ring-width measurements, tree-ring-density measurements are absolutely dated and give a data point each and every year. When we measured maximum latewood density in our Pyrenees tree-ring chronology, we found that it was strongly and uniquely influenced by summer temperature, which allowed us to use the chronology to reconstruct past summer temperatures after all.

The concept behind using tree rings to reconstruct past climate is fairly straightforward. The width (or density) of the annual rings can be measured, absolutely dated, and compared on a year-by-year basis with instrumental weather-station data. In our Pyrenees project, we collected cores in the summer of 2006, so the last fully formed ring on the cores was the 2005 ring. The oldest sample in the Pyrenees tree-ring chronology dates back to 924 CE, and at least five samples date back to 1260 CE. Because of the more robust replication of five samples, we consider the year 1260 to be the start year of our reconstruction.

Fortunately, instrumental temperature measurements started early in the Pyrenees, and temperature data from the nearby mountain observatory Pic du Midi are available from 1882 onwards. We can thus compare our maximum-latewood-density data for each year with the summer-temperature data for the same year collected at the observatory for the period 1882–2005. As a result, we have 124 years of tree-ring data to compare with 124 years of summer-temperature data. Given that our maximum-latewood-density (MXD) measurements are good recorders of summer temperature, we can keep it simple and link maximum latewood density for each year to summer temperature (Tsummer) for that year in a linear equation or model:


Tsummer(t) = a*MXD(t) + b.


This equation states that summer temperature in year t can be expressed as a function of maximum latewood density of the ring formed in that year. Because density measurements are made in grams per cubic centimeter (g/cm3) but we want to reconstruct summer temperature in degrees Celsius, we need the constants a and b to transform g/cm3 into degrees Celsius. We use the MXD and Tsummer data for the period 1882–2005 to calculate the values of a and b. To test the strength of the relationship between MXD and Tsummer, we calculate how many of the 124 years of overlap feature hot summers (high Tsummer) corresponding to high density (high MXD), and vice versa. If the relationship is strong, then we can use this same equation, with the same a and b values, to calculate summer temperature for any year for which we have a maximum-latewood-density value, all the way back to 1260. We just multiply maximum latewood density for a certain year by a and add b to it, and this will give us an estimate of summer temperature for that year.

By doing this for each year of the tree-ring record, we can reconstruct summer temperature prior to the start of the instrumental record, back to 1260. The simplest models (or equations), such as the one above, use one tree-ring chronology to predict and reconstruct one instrumental climate time series from a nearby location. Often such simple models can be improved by combining tree-ring chronologies from multiple locations.

In the Pyrenees, for instance, our Lake Gerber maximum-latewood-density record better represented the instrumental summer temperature if it was combined with a density record from a nearby timberline site in Sobrestivo, about 45 miles west. The model can also be optimized by selecting the climate variable that most strongly influences tree growth. Maximum latewood density of mountain pine in the Pyrenees, for instance, is more sensitive to temperature variability in May, August, and September than to that in June and July. In other words, our tree-ring data gave us a more reliable estimate of past May–September temperatures than of June–July temperatures.

A variety of climate variables (e.g., rainfall versus temperature, one month versus another, instrumental data from a single meteorological station versus the average of data from many stations in a region, and so forth) can thus be used on the lefthand side of the equation, and a variety of tree-ring data (e.g., tree-ring width and/or density, data from one tree species or multiple species, data from one site or multiple sites) on the righthand side. An important part of a dendroclimatologist’s job is to select the right tree-ring data and climate data for reconstruction and to determine through statistical analyses which combination gives the most reliable and robust results.


The Learned Pig


Using tree-ring data, ice- core data, and other temperature proxies, scientists have reconstructed annual temperature variability over the Northern Hemisphere for the past millennium. The graph, known as the Hockey Stick, shows slowly cooling temperatures between 1000 and ca. 1850 (the shaft), before a steep warming occurs that continues throughout the twentieth century (the blade). 1998, the most recent year of the study, was the hottest year.


The Learned Pig


In 1998, climatologist Michael Mann, paleoclimatologist Ray Bradley, and dendrochronologist par excellence Malcolm Hughes took this simple concept and moved it a giant step forward. By the late 1990s, the extraordinary character of twentieth-century global warming was unmistakable, and Mann, Bradley, and Hughes aimed to put the recent warming in a historical context in order to find out whether it could be part of a natural climate cycle. For this purpose, they developed a year-by-year reconstruction of Northern Hemisphere temperatures for the past 600 years. They combined tree-ring data with ice-core data and other proxies to reconstruct yearly temperature variations averaged over the Northern Hemisphere. They applied a new statistical approach to produce a single reconstruction for the entire hemisphere that covered the period back to 1400 CE, which was published in the scientific journal Nature. Their reconstruction demonstrated that twentieth-century global warming was unprecedented over the past 600 years. In a follow-up paper a year later, they extended the reconstruction even further back in time to 1000 CE. The key figure of their paper was a plot of Northern Hemisphere temperature change over time that resembles a hockey stick (above). The graph shows slowly cooling temperatures between 1000 CE and about 1850 CE (the hockey stick’s “shaft”) before a steep warming that continued throughout the twentieth century (the “blade”). The hottest year of the 1,000-year-long Hockey Stick was 1998, also the most recent year of their record.

The Mann, Bradley, and Hughes Hockey Stick paper was the first to show that twentieth-century warming was unprecedented in a 1,000-year perspective and thus unlikely to be part of a natural cycle. Given the relevance of this finding, the Hockey Stick graph was featured prominently in the 2001 report of the Intergovernmental Panel on Climate Change (IPCC). The IPCC is a United Nations scientific body entrusted with providing a comprehensive, scientific, and objective overview of climate change and its impacts on society. It won the 2007 Nobel Peace Prize, jointly with former US vice president Al Gore.

The IPCC does not carry out original research, but every five years or so it produces a hefty report based on the published scientific literature. The report is written by volunteering scientists and reviewed by governments before publication. The 2001 IPCC report was an 800-page tome. Because no sane policymaker carries 5.5 pounds of paper around in her briefcase (or has time to read such a lengthy report), a handy, approximately 30-page summary for policymakers was published that highlighted the most important findings with a handful of graphs. The Hockey Stick graph was prominent in this summary, and it subsequently drew global media attention when a large poster of it was used as the backdrop to a television announcement presenting the 2001 IPCC report.

As Mann, Bradley, and Hughes’s 1998 and then 1999 Hockey Stick papers were going through the review process, the three highly respected scientists anticipated that their findings would make a media splash, but they were unprepared for the media frenzy that followed. The Hockey Stick story was picked up by all the major media outlets, and in a New York Times interview Mann reemphasized that “the warming of the past few decades appears to be closely tied to emission of greenhouse gases by humans and not any of the natural factors.”* What followed was almost two decades of relentless political inquisition and intimidation.

Two of the first and foremost political mudslingers were James Inhofe, a US senator from Oklahoma and chair of the Senate Committee on Environment and Public Works, and Joe Barton, a congressman from Texas and chair of the House Energy Committee. Inhofe is well known for repeatedly calling man-made global warming “the greatest hoax ever perpetrated on the American people.” In their sustained efforts to oppose potential restrictions on greenhouse-gas emissions and to dismiss the ideas behind anthropogenic climate change, both Republican politicians honed in on the Hockey Stick, the poster child of the IPCC and of climate-change policy. In the years 2003 through 2006, Inhofe and Barton convened multiple congressional hearings to which they invited the Hockey Stick scientists and, more importantly for them, a range of climate-change skeptics. Debate between scientists, in which controversial scientific methods and results are double-checked and the validity of conclusions is argued, is a crucial part of the scientific process, but the political arena is not the place for such debate. In science, you cannot win a majority vote on who is right or who is wrong; scientific facts are not decided by the winning number of votes. Or as Sherwood Boehlert, the chairman of the House Science Committee, himself a conservative Republican, phrased it in a letter to Barton: “My primary concern about your investigation is that its purpose seems to be to intimidate scientists rather than to learn from them, and to substitute congressional political review for scientific review.”**

Michael Crichton’s fictional thriller State of Fear [became] required reading for the Senate Committee on Environment and Public Works.

This politicization of the Hockey Stick hit rock bottom in September 2005, when Senator Inhofe invited Michael Crichton, the creator of the popular TV series ER and author of fictional thrillers such as Jurassic Park, to testify before the Senate on the legitimacy of climate change during a hearing on the role of science in environmental policy making. Inhofe referred to the novelist Crichton as a scientist and made his fictional thriller State of Fear required reading for the Senate Committee on Environment and Public Works. In his novel, Crichton imagines a world where climate change is not an ecological reality but rather an evil eco-terrorist conspiracy. In his two-hour-long testimony before the committee, Crichton expressed his strong doubts about “whether the methodology of climate science is sufficiently rigorous to yield a reliable result.”† He then sat back as Senator Hillary Rodham Clinton expressed her opinion that his views “muddied the issues around sound science.”

To me, a fiction writer as a star witness before a US Senate committee on the validity of scientific research seems just as unbelievable as a T-Rex running amok in a Disney-esque theme park. In an additional attempt to break the Hockey Stick, Inhofe’s greatest ally, Barton, demanded from the three scientists full records of all of their climate-related research. The laundry list of demands included exhaustive information about all financial support they had received during their long careers, the source of funding for every study they had conducted, and all data and code for every paper they had published. In a letter asking Barton to withdraw his requests, Democratic congressman Henry Waxman wrote, “These letters do not appear to be a serious attempt to understand the science of global warming. Some might interpret them as a transparent effort to bully and harass climate change experts who have reached conclusions with which you disagree.”*

The goal of Inhofe, Barton, and their disciples in this “misguided and illegitimate investigation,”† according to the Republican Boehlert, was to ensure that legislation to regulate greenhouse-gas emissions in the US never passes. As reported by Naomi Oreskes and Erik Conway in their 2010 book Merchants of Doubt, this tactic, which keeps a controversy alive by spreading doubt and confusion despite a scientific consensus, has successfully been applied in the past by the tobacco industry in its denial of the connection between smoking and cancer. The ruthlessness of the climate-change denier’s tactics became very evident in November 2009, when hackers broke into the server of the Climate Research Unit at the University of East Anglia and stole and made public thousands of researchers’ private email correspondences. The emails showed some researchers to be rude, others cocky or petty. What they did not show was a sweeping scientific conspiracy perpetrated by a shadowy cabal of global proportions. Yet, that is what the climate-change-denial league behind the hacking claimed.

It is no coincidence that this illegal hacking occurred just weeks before the UN Climate Summit in Copenhagen, at which an international framework for climate-change mitigation was to be agreed upon. The media were quick to dub the hacking incident “Climategate,” focusing not on the hacking crime itself but on the language used in the emails. The Climate Research Unit — backed by multiple professional scientific organizations — rejected the accusations. No fewer than eight independent committees, including one from the US Environmental Protection Agency, have investigated the emails and allegations, and all came to the same conclusion: that there was no evidence of fraud or scientific misconduct. Nevertheless, Climategate was very effective in drawing media attention away from the important goals of the Copenhagen climate summit, while dragging some of the world’s most renowned climate scientists through the mud. Mann, Bradley, and Hughes are still dealing with the aftermath of the Hockey Stick controversy and Climategate, almost two decades after their paper was first published. Over the past twenty years, more and more scientific studies showing the unprecedented nature of the current climate have confirmed and advanced the concept of the original Hockey Stick paper.

As time goes on, reality is also catching up with us. Nineteen ninety-eight, the most recent and hottest year of the Hockey Stick record, now only is the tenth hottest year on record, with nine even hotter years occurring since. Needless to say, the relentless intimidation inspired by those who stand to lose much if actions are taken to limit emissions has required an inordinate amount of my colleagues’ attention, time, and energy. Attention, time, and energy that they were not able to spend on research, on sampling more and older trees, on crossdating more samples, or on publishing more scientific results. And therein seems to lie a key motivation for this dogged political inquisition and intimidation: to keep climate scientists from doing their job, from studying natural and man-made climate change and sharing their findings with the world.



Shortening the Candle's Wick

* David is now the director of the LTRR in Tucson; Ulf is a professor of geography at Cambridge University.
** Boehlert to Barton, 14 July 2005
† “The role of science in environmental policy making,” Hearing before the Committee on Environment and Public Works, US Senate, 28 September 2005

This is an extract from chapter six of Tree Story: The History of the World Written in Rings by Valerie Trouet, published by Johns Hopkins University Press.


Image credit: Simon Harrod, Tree rings, 2012, via Flickr

This is part of ROOT MAPPING, a section of The Learned Pig devoted to exploring which maps might help us live with a clear sense of where we are. ROOT MAPPING is conceived and edited by Melanie Viets.


The Learned Pig

The Learned Pig

Launched in November 2013, The Learned Pig is an online arts magazine with a love for plants and animals. Our thinking has grown out of our engagement with landscape aesthetics and environmental ethics and these days we try to bring together multiple perspectives on bodies and places, interspecies interactions, and alternative conceptions of that which has rather too frequently been defined as “nature” or “the animal”.