This is a writeup of a shallow investigation, a brief look at an area that we use to decide how to prioritize further research.

In a nutshell

  • What is the problem? Asteroids occasionally hit Earth, and a sufficiently large asteroid could cause enormous humanitarian damage. NASA reports that all of the near-earth asteroids as large as the one believed to have killed the dinosaurs have been detected, and substantial progress has been made in detecting other dangerously large asteroids. A 2010 National Research Council report, the most authoritative we’re aware of, estimated remaining actuarial risk due to asteroid strikes equivalent to less than 100 fatalities a year.
  • What are possible interventions? To substantially reduce residual risk from asteroid strikes, we would need better detection technology – a new telescope – followed, in the unlikely event of detecting a hazard, by action to divert it or minimize humanitarian damage. The B612 Foundation is currently raising funds to attempt to design and launch such a space-based telescope, which would be the first ever deep-space satellite launched by a private group. Philanthropic funds could also potentially also be used to lobby a government to fund such a satellite, instead of funding it directly.
  • Who else is working on it? Although the most recent asteroid tracking goal set by the U.S. Congress is not expected to be met, the issue appears to remain on the agenda of NASA and Congress to some extent.


Published: March 2013

Why did we look into this area?

  • Asteroid impact is a possible global catastrophic risk (i.e. it could conceivably endanger the survival of humanity).  The potential catastrophe from such an impact is so great that an investment in asteroid detection could conceptually have high returns.
  • Unlike other GCRs (e.g., nuclear war), asteroid risk is extremely quantifiable: scientists have estimated the number and size of near-earth asteroids and are able to track how many have been discovered.
  • Multiple asteroid deflection mechanisms have been designed and, according to advocates, could be carried out with components of existing technology,1 suggesting that a focus on asteroid detection is appropriate. Early detection could also help minimize damage from an impact that couldn’t be avoided by giving sufficient warning for people to be evacuated from a given area.

 

What is the problem?

The risk of asteroid impacts is quite well-understood relative to other long-term risks to humanity like climate change or nuclear war, though the humanitarian impact of asteroid strikes is more difficult to assess: there has only been one asteroid impact in recorded history that caused serious damage, so there are not many past examples to study,2 and the effects of impact depend on many characteristics of the asteroid and the way it collides with Earth.3 Our understanding is that the risk from asteroids is believed to be much larger than the risk from comets, which are harder to detect, so we focus here on asteroids.4

NASA reports that all near-earth asteroids larger than 10 kilometers in diameter—the size of impact thought to have caused the extinction of the dinosaurs—have already been identified.5 As of September 2011, roughly 93% of asteroids larger than 1km in diameter—large enough to have potentially global effects—had also been tracked, according to NASA.6 This eliminates much of the estimated risk due to asteroids, since extremely harmful low-probability impacts dominate expected-value calculations of asteroid risk.7 We have not vetted these claims, but our rough understanding is that the NASA study indicating that 93% of asteroids larger than 1 km in diameter have been found is likely to be robust.8

Taking the frequency of impacts by asteroids of different sizes, the proportion of asteroids of different sizes that have been successfully tracked, and the expected humanitarian effects of such impacts, it is possible to estimate the expected number of deaths due to asteroids each year. Scientists have conducted these calculations several times as near-earth asteroid tracking has progressed.

In a 2003 report, NASA estimated that the expected value of asteroid risk would be roughly 150 global deaths per year in 2008.9 In an update published in 2010 by the National Research Council, the most authoritative report on this issue that we’re aware of, the estimated remaining actuarial risk due to asteroid strikes declined to 91 fatalities a year.10 (A 2012 presention by the scientist cited in the National Research Council report suggested an even lower current value—64 expected fatalities per year, on track to decline to 33 by 2030—though we have not seen a published estimate to that effect.11)

What are possible interventions?

The B612 Foundation is a non-profit organization that is currently raising funds to launch a satellite to detect potentially hazardous near-earth asteroids.12 They estimate that their mission will cost $450 million and be able to catalog 90% of asteroids large enough to create a crater on earth (140 meters).13 No deep-space satellite has ever before been launched by a private body,14 and the B612 Foundation’s satellite would need to be custom-built for the mission.15

Instead of directly financing a telescope, as the B612 Foundation is attempting to do, a philanthropist could also attempt to lobby a one or more governments to fund a new telescope. We do not have a sense of how expensive such a lobbying effort might need to be or how it might compare in likelihood of success to directly funding a telescope.

Who else is working on this?

In 1998, the U.S. Congress set a goal of finding and tracking 90% of near-earth asteroids larger than 1 kilometer in diameter by 2008. Although it was not completed by 2008, NASA reports that that goal has since been met.16

In 2005, Congress set an additional goal: identifying 90% of asteroids larger than 140 meters in diameter by 2020.17 The National Research Council estimates that $50 million per year would be required to reach the new 90% goal a decade late, by 2030, but current funding is only approximately $20 million per year, and the goal is not expected to be met.18

Despite the lack of funding to date, it seems to us that this issue continues to be on NASA and Congress’ agenda to some extent. For instance, in February 2013, two Democrats on the House subcommittee on space wrote an op-ed in the Washington Post calling for more spending on asteroid risk mitigation, and the committee chairman, a Republican, announced a hearing to discuss the risk.19 (Both were triggered by the coincidence of a meteor strike that caused many injuries—but no deaths—in Russia and a near-miss of Earth by a 45 meter-wide asteroid on February 15.) In addition, over the past decade, NASA has conducted multiple cost-effectiveness analyses of further work to detect near-earth asteroids.20

Questions for further investigation

Our research in this area has been relatively limited, and many important questions remain unanswered by our investigation.

Amongst other topics, our further research on this cause might address:

  • How quickly will the remaining 7% of near-earth asteroids larger than one kilometer in diameter be found at current rates?21 How reliable are the underlying estimates of the likely numbers of different sizes of near-earth asteroids?
  • How precise and reliable are extant estimates of humanitarian damages for different sizes of asteroids? How much is known about the likely humanitarian damages and what are the chances that scientists are mistaken?
  • At current levels of funding, how long would it take NASA to find the remaining undiscovered near-earth asteroids larger than 140 meters in diameter?
  • What might be the cost and likelihood of success of efforts to lobby a government to fund further detection capacity?

Sources

  • 1.

    “Yet, contrasting with the irrational perceptions of the impact hazard, it potentially can be mitigated in much more concrete ways than is true of most hazards. An impact can be predicted in advance in ways that remain imperfect but are much more reliable than predictions of earthquakes or even storms, and the components of technology exist—at affordable costs given the consequences of an actual impact—to move any threatening object away and avoid the disaster altogether.” Clark R. Chapman, “The Hazard of Near-Earth Asteroid Impacts on Earth,” Pg 13.

    • “Nuclear standoff explosions are assessed to be 10-100 times more effective than the non-nuclear alternatives analyzed in this study. Other techniques involving nuclear explosives may be more effective, but they run an increased risk of fracturing the target. They also carry higher development and operations risks.
      Kinetic impactors are the most mature approach and could be used in some scenarios, especially for objects that consist of a relatively small, solid body.
    • Slow push deflection techniques are the most expensive, and their ability to both travel to and divert a threatening object is limited unless mission durations of many decades are available.
    • Deflection campaigns may need to be 100-1,000 times more reliable than current space missions to meet mitigation requirements.
    • Many potentially hazardous objects (30-80%) are in orbits that are beyond the capability of current or planned launch systems. Therefore, if these objects need to be deflected, swingby trajectories or on-orbit assembly of modular propulsion systems may be required to augment launch vehicle performance.”

    NASA, “2006 Near-Earth Object Survey and Deflection Study,” Pg 16.

  • 2.  

    “Even the diameter of the asteroid that created the only known serious historically-recorded damage, the 1908 Tunguska explosion that leveled about 2000 km2 of Siberian forest, has been recently revised downward—according to a single study— from 50-75 meters to about 30-40 meters, thus causing a reassessment of the apparent risk from airblast due to impacts by small asteroids.” National Research Council 2009, Pg 3.

  • 3. “The hazard associated with NEA impacts – that is, the probability for an individual of premature death as a consequence of an impact – depends on the frequency of occurrence as well as the destructive effects. Quantitative estimates of this risk were presented in the NASA Spaceguard Survey Report (Morrison, 1992) and amplified by Chapman and Morrison (1994) and Morrison et al. (1994). Such estimates are substantially uncertain due to lack of precision (and possible time dependence) in the impact flux as a function of projectile energy, possibly wide variability in the environmental effects that depend on properties of the impactor and target, and especially ignorance of the response of society to disasters of a magnitude never experienced. However, we can estimate the order of magnitude of the risk and its approximate dependence on impact energy, as a guide to possible societal responses and efforts at mitigation.” Morrison 2002, Pg 12.
  • 4.

    “The relative constancy of the long-period comet discovery rate over the past 300 years, the results from the Sekanina and Yeomans (1984) analysis, the Marsden (1992) type analysis and the above reality check all suggest that the threat of long-period comets is only about 1% the threat from NEAs. Levison et al. (2002) note that as comets evolve inward from the Oort cloud, the vast majority of them must physically disrupt rather than fade into dormant comets; otherwise, vast numbers of dormant long-period comets would have been discovered by current NEO surveys. This conclusion would strengthen the case against there being a significant number of dormant long-period or Halley-type comets that annually slip past the Earth unnoticed. While Earth impacts by long-period comets are relatively rare when compared to the NEA impact flux, the present number of Earth-crossing asteroids drops very steeply for asteroids larger than 2 kilometers in diameter, more steeply than the flux of cometary nuclei (Weissman and Lowry 2003). Hence, it is possible, perhaps even likely, that long-period comets provide most of the large craters on the Moon (diameter > 60 km) and most of the extinction level large impacts on Earth (Shoemaker et al., 1990).
    The conclusion is that, while a newly discovered Earth-threatening, long-period comet would have a relatively short warning time, the impact threat of these objects is only about 1% the threat from NEAs. More generally, the threat from all long-period or short-period comets, whether active or dormant, is about 1% the threat from the NEA population. The limited amount of resources available for near-Earth object searches would be better spent on finding Earth- threatening NEAs with the knowledge that these types of surveys will, in any case, find many of the Earth-crossing, long-period comets as well.” NASA Near-Earth Object Science Definition Team 2003, Pg 16.

  • 5.

    “It is believed that all near-Earth asteroids approximately 6 miles (10 kilometers) across, as big as the one thought to have wiped out the dinosaurs, have been found.” NASA, “NASA Space Telescope Finds Fewer Objects Near Earth.”

  • 6.
    • “The new data [on NEAs 1-km or larger in diameter] revise their total numbers from about 1,000 down to 981, of which 911 already have been found. None of them represents a threat to Earth in the next few centuries.” NASA, “NASA Space Telescope Finds Fewer Objects Near Earth.”
    • 911/981 = 92.9.
    • “By the 1990s, available research indicated that the impact of a 1.5- to 2-kilometer-diameter asteroid or comet anywhere on Earth had the potential to produce global effects that would seriously impact human civilization (e.g., a significant reduction in the total food yield, perhaps for several years). Because there were substantial uncertainties in the threshold impactor size needed to produce global effects, a team of NEO experts selected 1-kilometer-diameter objects as the threshold for the most dangerous objects to human civilization.
      The uncertainties in the damaging effects of asteroids increase as the size of the asteroid increases. A 1-kilometer-diameter asteroid is generally accepted as the lower boundary for an impactor with global consequences—asteroids below this size probably will not have globally catastrophic effects, although most estimates place the boundary for catastrophic effects starting at around 1.5 to 2 kilometers. Such an asteroid would be expected to produce a continent-sized fireball and form a crater approximately fifteen times the diameter of the asteroid, similar in size to many craters known from the geologic record; it could instead produce a devastating tsunami if it hit in an ocean. On average, such craters form at about 1-million-year intervals, but there is no known association between impact craters of this size and biologic extinctions. However, modern human civilization, with its strong dependence on agricultural crops and intricate distribution networks, is presumably much more fragile than the mere survival of humans or other animals as a species. We would thus want to avoid any impact that caused a large fraction of surviving humans to die of starvation, even though humans as a species would endure.” National Research Council 2009. Pgs 2-3.
  • 7.

    “Although giant impacts are very rare, when the threshold for globally destructive effects is exceeded (NEAs >1.5 – 3 km diameter) then the potential mortality is unprecedentedly large, so such impacts dominate mortality, perhaps 3000 deaths per year worldwide, comparable with mortality from other significant natural and accidental causes (e.g., fatalities in airliner crashes). This motivated the Spaceguard Survey. Now the estimated mortality is somewhat lower, ~1000 annual deaths due to somewhat lower estimates of the number of NEAs >1 km diameter and somewhat higher estimates of the threshold size for destructive global effects. Since most of that mortality has been eliminated by discovery of 55% of NEAs >1 km diameter and demonstration that none of them will encounter Earth in the next century, the remaining global threat is from the 45% of yet-undiscovered large NEAs plus the minor threat from comets. Once the Spaceguard Survey is complete, the residual global threat will be < 100 annual fatalities worldwide, see Table 1.” Chapman 2004, Pgs 8-9.

  • 8.

    The estimates are derived from a fairly straightforward sampling procedure: “The results come from the most accurate census to date of near-Earth asteroids, the space rocks that orbit within 120 million miles (195 million kilometers) of the sun into Earth’s orbital vicinity. WISE observed infrared light from those in the middle to large-size category. The survey project, called NEOWISE, is the asteroid-hunting portion of the WISE mission. Study results appear in the Astrophysical Journal. 

    ‘NEOWISE allowed us to take a look at a more representative slice of the near-Earth asteroid numbers and make better estimates about the whole population,’ said Amy Mainzer, lead author of the new study and principal investigator for the NEOWISE project at NASA’s Jet Propulsion Laboratory (JPL) in Pasadena, Calif. ‘It’s like a population census, where you poll a small group of people to draw conclusions about the entire country.’ 

    WISE scanned the entire celestial sky twice in infrared light between January 2010 and February 2011, continuously snapping pictures of everything from distant galaxies to near-Earth asteroids and comets. NEOWISE observed more than 100 thousand asteroids in the main belt between Mars and Jupiter, in addition to at least 585 near Earth.” NASA, “NASA Space Telescope Finds Fewer Objects Near Earth.”

  • 9.
    • Chapman 2004, Table 1, Pg 9. Estimated annual worldwide deaths from impacts: nominal residual hazard: 155. The “minimum” and “maximum” values for residual hazard of worldwide deaths from impacts are 36 and 813, respectively, though it is not clear what confidence should be associated with these ranges; most of the variation comes from estimates of catastrophic risk potential rather than from smaller asteroids.
    • “The most comprehensive analysis of the risks of NEA impacts is that of the NASA NEO Science Definition Team (SDT) [6]. 
      Although giant impacts are very rare, when the threshold for globally destructive effects is exceeded (NEAs >1.5 – 3 km diameter) then the potential mortality is unprecedentedly large, so such impacts dominate mortality [64], perhaps 3000 deaths per year worldwide, comparable with mortality from other significant natural and accidental causes (e.g., fatalities in airliner crashes). This motivated the Spaceguard Survey. Now the estimated mortality is somewhat lower, ~1000 annual deaths [6] due to somewhat lower estimates of the number of NEAs >1 km diameter and somewhat higher estimates of the threshold size for destructive global effects. Since most of that mortality has been eliminated by discovery of 55% of NEAs >1 km diameter and demonstration that none of them will encounter Earth in the next century, the remaining global threat is from the 45% of yet-undiscovered large NEAs plus the minor threat from comets. Once the Spaceguard Survey is complete, the residual global threat will be < 100 annual fatalities worldwide, see Table 1 [6].
      The SDT [6] also evaluated two other sources of mortality due to NEO impactors smaller than those that would cause global effects: (a) impacts onto land, with local and regional consequences analogous to the explosion of a bomb and (b) impacts into an ocean, resulting in inundation of shores by the resulting tsunamis. The SDT evaluated fatalities for land impacts using (a) a model for the radius of destruction by impactors >150 m diameter [65] that survive atmospheric penetration with most of their cosmic velocity (although 220 m may be more nearly correct [66]) and (b) a map of population distribution across the Earth. A thorough analysis of the tsunami hazard [67], based on reanalysis of wave and run-up physics combined with analysis of coastal populations, provided an estimated number of ‘people affected per year’ by impact-generated tsunami. As the SDT notes, historically only ~10% of people in an inundation zone die, thanks to advance warning and evacuation. Hence, in Table 1, which summarizes mortality from land impacts, ocean impacts, and globally destructive impacts, I divide the SDT’s estimated tsunami hazard by a factor of 10. 

      In Table 1, the ‘overall hazard’ is that posed by nature, before the Spaceguard Survey started to certify that a fraction of NEAs (more larger ones than smaller ones) will not hit. The ‘residual hazard’ (see Fig. 3) is what is expected after about 2008. Whereas non-global impacts constitute < 10% of the natural impact hazard, they are nearly half of the residual hazard. The land-impact hazard is chiefly due to bodies 70 – 200 m diameter (indeed, the chances are better than 1% that such an impact will kill ~100,000 people during the 21st century; larger bodies, 150 – 600 m are mainly responsible for the somewhat smaller tsunami hazard.” Chapman 2004, pgs 8-9.

    • NASA Near-Earth Object Science Definition Team 2003.
  • 10.

    “Assuming that 85 percent of the NEOs with diameters larger than 1 kilometer have been discovered, which is close to the present state of affairs, Harris (2009) calculated the hazard statistics shown in Figure 2.7. Here the reassessed risk presented by the remaining 15 percent of the NEOs with diameters greater than 1 kilometer is comparable to that from all smaller objects. Figure 2.7 predicts that, in an actuarial sense, there is a long- term statistical average of about 91 fatalities worldwide per year due to impacts. Because the assessed statistical hazard from mid-range objects has dropped, the overall hazard has decreased as well. The drop from >1,000 to 91 expected fatalities per year clearly demonstrates the results of the Spaceguard Survey to date, which has ‘retired’ the statistical risk from most objects above the assumed global catastrophe threshold.” National Research Council 2010, Pg 24.

  • 11.

    “Alan Harris’ presentation posited an interesting question regarding the actuarial risk that we face from asteroid impact. The question: is reduction of risk really worth the cost of large surveys? When the modern effort to survey for potentially hazardous asteroids began, we didn’t know where asteroids were, only that they were out there and that an unknown one could present a hazard. Harris showed that the majority of the actuarial risk due to impacts is from undiscovered large objects. Near-Earth object surveys have found (we think) 98% of the largest objects that present the most risk, reducing the actuarial risk due to asteroid impacts from 250 fatalities per year to 64 per year. Based on past discovery rates and projecting forward through proposed future projects, over the next 16 years, we should achieve 90% completion of discovery of asteroids larger than 140 meters in diameter. The effect of this 16 years of work – at a cost of roughly a billion dollars – will be to reduce the actuarial risk to 33 fatalities per year. If you see asteroid surveys as a form of insurance, then you’re spending about two million dollars per fatality avoided. From the point of view of insurance, this is a relatively expensive effort. Harris’ point: “The hazard stuff might sell the program,” but in fact, the benefit is questionable; the real value of survey programs is in the science they produce.” Emily Lakdawalla, “DPS 2012: Future impact risks.”

  • 12.

    Wired Science 2012.

  • 13.
    • “The B612 Foundation plans to raise $450m over 12 years (or about $37m per year) to design, build, test, insure, and launch the Sentinel Space Telescope, to build and operate the control center for the duration of the mission, to carry out analysis of the observations, and to deliver the data to the people of the world. This is substantially lower than the budget would be for a similar mission in the government sector, and comparable in cost to many other philanthropic projects, including specialized medical research facilities, museums, performing arts centers, and academic buildings.” B612 Foundation, “FAQ: B612 Basics” http://b612foundation.org/about-us/faqs/faq-b612-basics/
    • “Q. How long will it take to complete the Sentinel Map of the inner solar system? How often does this survey need to be repeated?
      In 6.5 years after launch, Sentinel will discover and track half a million Near Earth Asteroids, including 90% of asteroids larger than 140 meters in diameter (i.e. those that should they hit Earth would have an impact energy of greater than 100 Megatons). This map is accurate enough to project asteroid positions 100 years into the future.” B612 Foundation, “FAQ: Aspects of the Sentinel Mission” http://b612foundation.org/about-us/faq-aspects-of-the-sentinel-mission/
    • “Rocky PHOs about 140 meters and larger would be expected to punch completely through the Earth’s atmosphere, causing a cratering event on the Earth’s surface (Hills and Goda, 1993).”  NASA Near-Earth Object Science Definition Team 2003. Pg 112.
  • 14.

    B612 Foundation, “FAQ”:

    • “Q. What the Sentinel Mission?
      The Sentinel Mission will be the first privately funded, launched, and operated deep space mission. It is an infrared space telescope to be placed in orbit around the Sun that will discover and map the locations and trajectories of a half million asteroids whose orbits approach Earth.”
    • “Q. Are there precedents for privately funding and conducting these types of missions in the private sector?
      While this is the first private deep space mission, there have been many privately funded and operated low Earth orbiting communications and Earth observing spacecraft. In addition, many large ground based telescopes have been privately funded.” 
  • 15.

    “The B612 Foundation is working with Ball Aerospace, Boulder, CO, which has designed and will be building the Sentinel Infrared (IR) Space Telescope with the same expert team that developed the Spitzer and Kepler Space Telescopes. It will take approximately five years to complete development and testing to be ready for launch in 2017-2018. The launch vehicle of choice is the SpaceX Falcon9.” B612 Foundation, “B612 Overview.”

  • 16.

    “New observations by NASA’s Wide-field Infrared Survey Explorer, or WISE, show there are significantly fewer near-Earth asteroids in the mid-size range than previously thought. The findings also indicate NASA has found more than 90 percent of the largest near-Earth asteroids, meeting a goal agreed to with Congress in 1998.” NASA, “NASA Space Telescope Finds Fewer Objects Near Earth.”

  • 17.

    “Finding: Congress has mandated that NASA discover 90 percent of all near-Earth objects 140 meters in diameter or greater by 2020. The administration has not requested and Congress has not appropriated new funds to meet this objective. Only limited facilities are currently involved in this survey/discovery effort, funded by NASA’s existing budget. Finding: The current near-Earth object surveys cannot meet the goals of the 2005 NASA Authorization Act directing NASA to discover 90 percent of all near-Earth objects 140 meters in diameter or greater by 2020.” National Research Council 2009, Pg 1.

  • 18.

    “In 2005 Congress set a 15-year deadline for scientists to find 90 percent of the near-Earth objects greater than about 500 feet in diameter — those large enough to cause regional or global devastation. But the mandate has been chronically underfunded. The project would require several more dedicated telescopes. Last year the project received about $20 million, far less than the $50 million that the National Research Council estimated in 2010 was needed to reach the congressional goal by 2030, a decade late. Even when this goal is met, most small asteroids and comets — too small to cause global devastation but still large enough to cause damage far worse than just occurred in Russia — will remain undetected unless funding is significantly increased.” Rush Holt and Donna F. Edwards. “We’re on notice to plan for the next meteor.”

  • 19.
    • Rush Holt and Donna F. Edwards. “We’re on notice to plan for the next meteor.”
    • Lamar Smith. “Smith: Asteroid, Meteor Stark Reminders of Need to Invest in Space Science.”
  • 20.
    • 2003: NASA Near-Earth Object Science Definition Team 2003.
    • 2010: National Research Council 2010.
    • 2012: “Alan Harris’ presentation posited an interesting question regarding the actuarial risk that we face from asteroid impact. The question: is reduction of risk really worth the cost of large surveys? When the modern effort to survey for potentially hazardous asteroids began, we didn’t know where asteroids were, only that they were out there and that an unknown one could present a hazard. Harris showed that the majority of the actuarial risk due to impacts is from undiscovered large objects. Near-Earth object surveys have found (we think) 98% of the largest objects that present the most risk, reducing the actuarial risk due to asteroid impacts from 250 fatalities per year to 64 per year. Based on past discovery rates and projecting forward through proposed future projects, over the next 16 years, we should achieve 90% completion of discovery of asteroids larger than 140 meters in diameter. The effect of this 16 years of work – at a cost of roughly a billion dollars – will be to reduce the actuarial risk to 33 fatalities per year. If you see asteroid surveys as a form of insurance, then you’re spending about two million dollars per fatality avoided. From the point of view of insurance, this is a relatively expensive effort. Harris’ point: “The hazard stuff might sell the program,” but in fact, the benefit is questionable; the real value of survey programs is in the science they produce.” Emily Lakdawalla, “DPS 2012: Future impact risks.”
  • 21.

    A 2012 presentation from Alan Harris, the scientist who produced the humanitarian impact estimates cited in the National Research Council’s 2010 report on asteroid risk might be read as implying that coverage of asteroids more than a kilometer in diameter has now reached 98%, though we are not certain of this reading and have not seen a published estimate to this effect: “Harris showed that the majority of the actuarial risk due to impacts is from undiscovered large objects. Near-Earth object surveys have found (we think) 98% of the largest objects that present the most risk, reducing the actuarial risk due to asteroid impacts from 250 fatalities per year to 64 per year.” Emily Lakdawalla, “DPS 2012: Future impact risks.”