This Changes Everything (44 page)

Boosters of Solar Radiation Management tend to speak
obliquely about the “distributional consequences” of injecting sulfur dioxide into the stratosphere, and of the “spatial heterogeneity” of the impacts. Petra Tschakert, a geographer at Penn State University, calls this jargon “a beautiful way of saying that some countries are going to get screwed.”
²⁵
But which countries? And screwed precisely how?

Having reliable answers to those key questions
would seem like a prerequisite for considering deployment of such a world-altering technology. But it’s not at all clear that obtaining those answers is even possible. Keith and Myhrvold can test whether a hose or an airplane is a better way to get sulfur dioxide into the stratosphere. Others can spray saltwater from boats or towers and see if it brightens clouds. But you’d have to deploy these
methods on a scale large enough to impact the
global
climate system to be certain about how, for instance, spraying sulfur in the Arctic or the tropics will impact rainfall in the Sahara or southern India. But that wouldn’t be a test of geoengineering; it would actually be conducting geoengineering.
²⁶

Nor could the necessary answers be found from a brief geoengineering stint—pumping sulfur for,
say, one year. Because of the huge variations in global weather patterns from one year to the next (some monsoon seasons are naturally weaker than others, for instance), as well as the havoc already being wreaked by global warming, it would be impossible to connect a particular storm or drought to an act of geoengineering. Sulfur injections would need to be maintained long enough for a clear pattern
to be isolated from both natural fluctuations and the growing impacts of greenhouse gases. That likely means keeping the project running for a decade or more.
^III
27

As Martin Bunzl, a Rutgers philosopher and climate change expert, points out, these facts alone present an enormous, perhaps insurmountable ethical problem for geoengineering. In medicine, he writes, “You can test a vaccine on one
person, putting that person at risk, without putting every
one else at risk.” But with geoengineering, “You can’t build a scale model of the atmosphere or tent off part of the atmosphere. As such you are stuck going directly from a model to full scale planetary-wide implementation.” In short, you could not conduct meaningful tests of these technologies without enlisting billions of people as guinea
pigs—for years. Which is why science historian James Fleming calls geoengineering schemes “untested and untestable, and dangerous beyond belief.”
²⁸

Computer models can help, to be sure. That’s how we get our best estimates of how earth systems will be impacted by the emission of greenhouse gases. And it’s straightforward enough to add a different kind of emission—sulfur in the stratosphere—to
those models and see how the results change. Several research teams have done just that, with some very disturbing results. Alan Robock, for instance, has run different SRM scenarios through supercomputers. The findings of a 2008 paper he coauthored in the
Journal of Geophysical Research
were blunt: sulfur dioxide injections “would disrupt the Asian and African summer monsoons, reducing precipitation
to the food supply for billions of people.” Those monsoons provide precious freshwater to an enormous share of the world’s population. India alone receives between 70 and 90 percent of its total annual rainfall during its June through September monsoon season.
²⁹

Robock and his colleagues aren’t the only ones coming up with these alarming projections. Several research teams have produced models
that show significant losses of rainfall as a result of SRM and other sunlight-reflecting geoengineering methods. One 2012 study shows a 20 percent reduction in rainfall in some areas of the Amazon after a particularly extreme use of SRM. When another team modeled spraying sulfur from points in the Northern Hemisphere for a 2013 study, the results projected a staggering 60–100 percent drop in a
key measure of plant productivity in the African countries of the Sahel (Burkina Faso, Chad, Mali, Niger, Senegal, and Sudan)—that means, potentially, a complete crop collapse in some areas.
³⁰

This is not some minor side effect or “unintended consequence.” If only some of these projections were to come true, that would transform a process being billed as an emergency escape from catastrophic
climate change into a mass killer in its own right.

One might think all of this alarming research would be enough to put
a serious damper on the upbeat chatter surrounding the Pinatubo Option. The problem is that—though computer models have proven remarkably accurate at predicting the broad patterns of climate change—they are not infallible. As we have seen from the failure to anticipate the
severity of summer sea ice loss in the Arctic as well as the rate of global sea level rise in recent decades, computer models have tended to underestimate certain risks, and overstate others.
³¹
Most significantly, climate models are at their weakest when predicting specific regional impacts—how much more southern Somalia will warm than the central United States, say, or the precise extent to which
drought will impact crop production in India or Australia. This uncertainty has allowed some would-be geoengineers to scoff at findings that make SRM look like a potential humanitarian disaster, insisting that regional climate models are inherently unreliable, while simultaneously pointing to other models that show more reassuring results. And if the controversy were just a matter of dueling
computer models, perhaps we could call it a draw. But that is not the case.

Without being able to rely on either models or field tests, only one tool remains to help forecast the risks of sun blocking, and it is distinctly low-tech. That tool is history, specifically the historical record of weather patterns following major volcanic eruptions. The relevance of history
is something all sides of the debate appear to agree on. Ken Caldeira has described the 1991 eruption of Mount Pinatubo “as a natural test of some of the concepts underlying solar radiation management” since it sent so much sulfur dioxide into the stratosphere. And David Keith assured me, “It’s pretty clear that just putting a lot of sulfur in the stratosphere isn’t terrible. After all, volcanoes
do it.” Likewise, Lowell Wood, Myhrvold’s partner in the invention of the StratoShield, has argued that because his hose-to-the-sky would attempt to imitate a natural volcano, there is “a proof of harmlessness.”
³²

Levitt and Dubner have stressed the relevance of historical precedent most forcefully, writing in
SuperFreakonomics
that not only did the earth cool after Pinatubo, but “forests around
the world grew more vigorously
because trees prefer their sunlight a bit diffused. And all that sulfur dioxide in the stratosphere created some of the prettiest sunsets that people had ever seen.” They do not, however, appear to believe that history offers any cautionary lessons: aside from a reference to the “relatively small” number of deaths in the immediate aftermath of the eruption due to
storms and mud slides, they make no mention in the book of any negative impact from Pinatubo.
³³

Critics of sun shielding also draw on history to bolster their arguments, and when they look back, they see much more than pretty sunsets and “proof of harmlessness.” In fact, a great deal of compelling research shows a connection between large volcanic eruptions and precisely the kinds of droughts
some computer models are projecting for SRM. Take the 1991 eruption of Mount Pinatubo itself. When it erupted, large swaths of Africa were already suffering from drought due to natural fluctuations. But after the eruption, the situation grew much worse. In the following year, there was a 20 percent reduction in precipitation in southern Africa and a 10–15 percent reduction in precipitation in South
Asia. The United Nations Environment Programme (UNEP) described the drought as “the most severe in the last century”; an estimated 120 million people were affected. The
Los Angeles Times
reported crop losses of 50–90 percent, and half the population of Zimbabwe required food aid.
³⁴

At the time, few linked these disastrous events to the Pinatubo eruption since isolating such climate signals takes
time. But more recent research looking at rainfall and streamflow patterns from 1950 to 2004 has concluded that only the sulfur dioxide that Pinatubo sent into the stratosphere can account for the severity of the drop in rainfall that followed the eruption. Aiguo Dai, an expert in global drought at the State University of New York, Albany, stresses that though the drought had additional causes,
“Pinatubo contributed significantly to the drying.” A 2007 paper cowritten by Dai and Kevin Trenberth, head of the Climate Analysis Section at the Colorado-based National Center for Atmospheric Research, concluded “that the Pinatubo eruption played an important role in the record decline in land precipitation and discharge, and the associated drought conditions in 1992.”
³⁵

If Pinatubo was the
only large eruption to have been followed by severe
and life-endangering drought, that might not be enough to draw clear conclusions. But it fits neatly into a larger pattern. Alan Robock, a leading expert on the effect of volcanoes on climate, points in particular to two other eruptions—Iceland’s Laki in 1783 and Alaska’s Mount Katmai in 1912. Both were sufficiently powerful to send a high volume
of sulfur dioxide into the stratosphere and, like Pinatubo, it turns out that both were followed by a series of terrible, or badly worsening regional droughts.

Reliable records of rainfall go back only roughly one hundred years, but as Robock informed me, “There’s one thing that’s been measured for 1,500 years, and that’s the flow of the Nile River. And if you look back at the flow of the Nile
River in 1784 or 1785”—the two years following Laki’s eruption in Iceland—“it was much weaker than normal.” The usual floods that could be counted on to carry water and precious fertilizing nutrients into farmers’ fields barely took place, the devastating consequences of which were recounted in the eighteenth-century travel memoirs of French historian Constantin-François Volney. “Soon after the
end of November, the famine carried off, at Cairo, nearly as many as the plague; the streets, which before were full of beggars, now afforded not a single one: all had perished or deserted the city.” Volney estimated that in two years, one sixth of the population in Egypt either died or fled the country.
³⁶

Scholars have noted that in the years immediately following the eruption, drought and famine
gripped Japan and India, claiming millions of lives, although there is much debate and uncertainty surrounding Laki’s contribution. In Western and Central Europe, meanwhile, a brutally cold winter led to flooding and high mortality rates. Expert estimates of the global death toll from the eruption and the resulting extreme weather range widely, from over one-and-a-half million to as many as
six million people. At a time when world population was less than one billion, those are stunningly high numbers, making Laki quite possibly the deadliest volcano in recorded history.
³⁷

Robock found something similar when he delved into the aftermath of the 1912 Katmai eruption in Alaska. Once again, his team looked at the historical record of the flow of the Nile and discovered that the year
after Katmai saw “the lowest flow for the twentieth century.” Robock and his colleagues also “had found a significant weakening of the Indian mon
soon in response to the 1912 Katmai volcanic eruption in Alaska, which resulted from the decreased temperature gradient between Asia and the Indian Ocean.” But it was in Africa where the impact of the great eruption took the heaviest human toll. In Nigeria,
sorghum, millet, and rice crops withered in the fields while speculators hoarded what grains survived. The result was a massive famine in 1913–1914 that took the lives of at least 125,000 in western Africa alone.
³⁸

These are not the only examples of deadly droughts seemingly triggered by large volcanic eruptions. Robock has looked at how such eruptions have impacted “the water supply for Sahel
and northern Africa” over the past two thousand years. “You get the same story from every [eruption] you look at,” he said, adding, “there haven’t been that many big eruptions but they all tell you the same stories. . . . The global average precipitation went down. In fact, if you look at global average precipitation for the last fifty years, the three years with the lowest global precipitation
were after the three largest volcanic eruptions. Agung in 1963, El Chichón in 1982, and Pinatubo in 1991.” The connections are so clear, Robock and two coauthors argued in one paper, that the next time there is a large “high-latitude volcanic eruption,” policymakers should start preparing food aid immediately, “allowing society time to plan for and remediate the consequences.”
³⁹

So how, given
all this readily available evidence, could geoengineering boosters invoke the historical record for “proof of harmlessness”? The truth is the mirror opposite: of all the extreme events the planet periodically lobs our way—from earthquakes and tsunamis to hurricanes and floods—powerful volcanic eruptions may well be the most threatening to human life. Because the people in the immediate path of an
eruption are not the only ones at risk; the lives of billions of others scattered throughout the globe can be destroyed by lack of food and water in the drier years to come. No naturally occurring disaster short of an asteroid has such global reach.

This grim track record makes the cheerful talk of a Pinatubo Option distinctly bizarre, if not outright sinister—especially because what is being
contemplated is simulating the cooling effects of an eruption like Pinatubo not once but
year after year for decades
, which could obviously magnify the significant risks that have been documented in the aftermath of one-off eruptions.