If you ask the average person whether in the long run it is climate change or an asteroid/comet impact that's expected to kill more people annually, you'll undoubtedly get some confused replies. Those asteroid movies are scary, but there are no verified instances of an asteroid strike killing any humans, are there? Meanwhile, the science of climate change is currently being overshadowed by a media-driven public debate, mainly in the U.S.

In fact, the expected annual fatality rate due to climate change is estimated to be far higher than that due to an asteroid or comet impact—150,000 versus 91, per the World Health Organization (WHO) and Alan Harris of the Space Science Institute, respectively. You won't, however, see that 150,000 figure in the main body of the Washington, D.C.–based National Research Council report on near-Earth object (NEO) surveys and mitigation strategies. (The report was written by a total of 42 scientists.)

Instead, in a chart on page 26 of the report on "expected fatalities per year, worldwide, from a variety of causes," asteroids are compared with shark attacks (three to seven deaths), firearms accidents (2,500), earthquakes (36,000), malaria (one million), traffic accidents (1.2 million), air pollution (two million), HIV/AIDS (2.1 million) and tobacco (five million).

Meanwhile, climate change is mentioned in a note beneath the chart, regarding one of the authors: "Mark Boslough wanted an additional entry in this table for fatalities due to climate change. The steering committee disagreed with including this entry because it did not think a reliable estimate is available, among other reasons. Dr. Boslough has written a minority opinion as Appendix D."

The mysterious "Appendix D"
In Appendix D Boslough, a physicist at Sandia National Laboratories in Albuquerque, N.M., wrote that he disagreed with the steering committee's reasons for removing the climate data. The reasons: a debate on climate change could distract from the issue at hand, and the irrelevance of climate-change numbers to the NEO threat.

Boslough wrote that it's "inappropriate" to remove data from a report to avoid the potential for political controversy. And as for irrelevance, climate change is actually more relevant than the other causes in the table, he wrote. Asteroid impacts that cause global catastrophe are climate-changing events, and most of the resulting fatalities would be due to that change (which would cause social disruption that is expected to lead to starvation, disease and violence). And asteroids and climate change share some features—both can have abrupt and worldwide consequences that would have nothing to do with one's lifestyle or location on the planet. Contrast that with earthquakes, malaria, tobacco or firearms use.

Fatality estimates for both an NEO impact and climate change "are similarly impacted by uncertainty in our understanding of climate change and statistical attribution of indirect causes," Boslough and Harris wrote in a poster presented at the fall 2008 meeting of the American Geophysical Union. And when you look at the risks side by side, the researchers continued, "by any objective measure the impact threat is minuscule (by a factor of at least a thousand) compared to the threat from anthropogenic climate change."

Nevertheless, they wrote, "The asteroid-threat community has been much more successful than the climate change community in characterizing the dominant worst-case scenarios and communicating them to policymakers, the media and the public—even though the climate change threat is more than a thousand times greater…, [therefore] quantitative comparison of climate change to asteroid impact is a valuable way to put both threats into perspective."

Why was climate change scuttled?
So why were the climate change figures removed from this chart? Panelists are forbidden from discussing how they arrived at such decisions, but steering committee chair Irwin Shapiro of the Harvard–Smithsonian Center for Astrophysics says he recalls no discussion about anthropogenic climate change.

"Personally, I was concerned with the 'apples–oranges' mixtures of comparison," he says. "Specifically, some data included were rooted in actual statistics, whereas others were based upon assumptions, with some being shakier than others. The end result was a compromise to which Mark took strong exception at the time the draft report was about to be sent for review."

Panel member Michael A'Hearn of the University of Maryland, College Park, says that no single person made the decision to remove the climate figures: "I think it was a majority opinion, but not a unanimous one."

He added: "I don't think there was any disbelief in anthropogenic climate change, as such. I suppose there may be some doubts about our quantitative predictions of climate change, and there were certainly doubts about the quantitative predictions of deaths due to it."

The sociopolitical climate played a part in the decision, he says: "There was definitely also the independent concern that this would become a distraction to the report."

A'Hearn actually would have preferred an even more reduced set of comparisons in the table. "My personal hesitation did not apply solely to the climate change entry but to several other entries, as well," for example, air pollution and possibly even tobacco deaths, he says, "and I espoused the general principle that we should list only discrete events with identifiable, individual deaths in order to have a fair comparison. Clearly that opinion did not carry the day."

Boslough accepts the fatality estimates for climate change (the WHO's estimate was made for the agency by epidemiologist Tony McMichael, of the Australian National University's National Center for Epidemiology and Population Health) as well as those for asteroid impacts, the latter of which are based lately on work by Harris.

The WHO/McMichael mortality rate is not a prediction. It is an estimation of the current (as of 2000) existing rate of annual deaths reasonably attributable to climate change, albeit from a limited subset of climate-related health impacts: malaria due to an increase in the geographic range of disease-bearing mosquitoes, malnutrition associated with loss of agricultural productivity, water-borne diarrheal diseases, and deaths from flooding, McMichael says.

Regarding the strength of McMichael's climate change fatality estimates, Boslough says: "I do not dispute that there are uncertainties in the numbers, but that's true of every number in the table, including (and perhaps especially) the impact number."

Boslough also rejects the idea that the comparison should list only discrete events with identifiable, individual deaths because that's not the case for the fatalities figures for an asteroid impact. "About half the deaths attributed to impact in our risk chapter are actually from events above an assumed 'global catastrophe' threshold," he says, "and most of those people don't get killed by the impact itself, but from starvation, exposure or violence associated with agricultural and civilization collapse in a world that would look a lot like Cormac McCarthy's The Road. Their death certificates would not say, 'killed by impact.' Their direct cause of death would be varied or ambiguous but ultimately traceable back to the impact just as the current increases in malaria and malnutrition are traceable back to climate change."

The methodology of speculative corpse counting
Human risk assessments for asteroid impacts are confusing in part because there has never been an asteroid impact in historic times that has caused even a local disaster (setting aside the 1908 flattening of more than 2,000 square kilometers of sparsely populated taiga near the Tunguska River in Siberia), let alone a global catastrophe. Nevertheless, scientists can estimate the risk and the frequency of such events quite well, Harris says.

Here is a rough outline of how it's done, he says: "In round numbers, [this risk] is dominated by very large events that would kill about a billion people, and happen about once in a million years—thus, in round numbers, about 1,000 per year. Present [NEO] surveys have found about 90 percent of that risk, and provided assurance that no such event will happen in the next century or so from the 90 percent we have found. So the leftover short-term risk is around 100 [per] year, as tabulated."

Harris says he agrees with much of what Boslough wrote in Appendix D: "It was important as a matter of 'due process,' the scientific endeavor cannot tolerate suppression of dissent. On the other hand, it is also unfortunate because…it has distracted attention from the central messages of the report."

The distraction is unfortunate, Boslough says, adding that he only insisted on the minority opinion because it was more important to include the climate change information than to avoid the potential for distraction.

The report, which came out in January and with which Boslough otherwise fully agrees, reveals that the scientific inventory of Earth-threatening space objects (asteroids and comets), especially the smaller ones that are most likely to impact our planet, is far from complete and unlikely to improve significantly without a greatly increased funds for NEO search programs. And our preparedness, if scientists found a moderate to large asteroid with our name on it, is weak—it'd take at least a decade to mount a space mission to deflect the object. And that might not be soon enough.