As more and more of these objects are found, the search for the missing ones gets tougher and tougher. So while we are more than half way toward meeting the Spaceguard goal in terms of the number of objects, we are not halfway there in terms of the time it will take to find the goal of 90 percent of them. Nevertheless, it seems likely that we will have discovered 90 percent of the NEAs larger than 1 kilometer by sometime not much beyond the 2008 deadline.

But there are a vast number of NEAs smaller than 1 kilometer in diameter. There are likely to be more than 300,000 that are about 100 meters long -- larger in diameter than a football field. Because there are many more of them, they would be expected to hit Earth far more frequently than the near-Earth asteroids larger than 1 kilometer -- once every few thousand years for the 100 meter asteroids, as opposed to every half million years for the 1 kilometer or greater asteroids.

Clark Chapman: Despite the fluctuating headlines in often inaccurate media stories, estimates of the relative risks due to comets and asteroids of various sizes has changed little in the last decade. Roughly 80 percent of the impact hazard is due to asteroids between 1 and several kilometers in diameter. About another 10 percent is due to an asteroid smaller than a kilometer striking the ocean and causing a tsunami. Roughly another 10 percent is due to comets, and less than 1 percent is due to small asteroids striking the land.

So it is sensible that NASA finally decided, in 1998, to endorse the Spaceguard Survey goal of emphasizing searches for the NEAs greater than 1 kilometer (though the survey also finds smaller NEAs and comets).

Alan Harris: In considering the relative importance of large versus small impacts, one must keep in mind both the nature and the frequency of the disaster. A large impact from, lets say, an asteroid larger than a mile (1.6 km) in diameter would be a global catastrophe. It would spoil your day, your whole life even, no matter where you live or where it hit.

Such an impact would lead to the equivalent of "nuclear winter," causing agricultural failure worldwide and famine that would undoubtedly lead to a billion or more deaths - a significant fraction of the world's population. Events of this magnitude are expected to occur once or twice in a million years.

Consider the other end of the impact range. The smallest impactor that can penetrate the atmosphere deep enough to cause any damage on the ground is not much smaller than the "Tunguska" bolide that flattened a couple thousand square miles of Siberian forest in 1908. The area flattened is about equal to the area of the greater Washington DC area, inside the beltway. That asteroid was estimated to be about 50 to 70 meters in diameter. The nature of the destruction is pretty much the same as a Hiroshima-style nuclear air burst, but without the radiation after-effects. Its still not a pretty picture if it happened in your neighborhood.

But before rising up and screaming "that's intolerable," we must take a careful look at just how often such a small impact event might be expected in a populated area.

I have analyzed the fatality rate of Tunguska-like small impact events over the entire surface of the Earth, given the present population distribution. If you take the fraction of the Earth devastated by the Tunguska impact (about one-millionth of the world's area) and multiply that by the worlds population, you can conclude that an "average" small impact event kills about 10,000 people. But the real historical event in Russia may have killed one person, at most. If it had happened over the sea, it wouldnt have killed anyone. So even if these small impacts happen every century, really catastrophic events caused by such impacts are much more rare.

I have estimated that the frequency of Tunguska-type impacts worldwide is only about once in a thousand years. That's on the edge of implausible since one happened only a century ago, but I think anything more often than once a century is inconsistent both with historical records and with observations of NEAs in space.

Assuming that such Tunguska events occur once in a millennium, a small impactor that hits an area populated enough to kill 1,000 people is expected only once in about 8,000 years. A small impactor that hits a mid-range population, killing 100,000 people, is expected about every 40,000 years. And a small impactor that directly hits a major population center, killing perhaps a million people, has a chance of occurring only a couple times in a million years.

I don't mean to trivialize the human loss of such disasters. In the case of a 1,000-death disaster, similar disasters happen almost annually from floods, earthquakes, and so forth, or at least several times per decade. The middle example - of 100,000 deaths - is comparable to the loss of life in some of the greatest natural disasters of the last century. But such an event caused by a small asteroid is expected to only occur once in 40,000 years a longer time period than all of recorded history.

Clark Chapman: In facing a hazard, whether as an individual or as collective society, we want to allocate our limited resources as effectively as possible. We would wish to address the core problem, and we would want to work first on what is most easily and cheaply accomplished. NASA's investment, which represents most of the world's investment, has been very modest: only a few million dollars a year. But they also leverage technological investments - for example, using Air Force imaging technology that was developed for other purposes. In addition, unpaid volunteer efforts by amateur and overseas astronomers make up part of this investment. It is a no-brainer that over a decade, the investment of perhaps $10,000 per expected life saved is a real bargain, especially with the added possibility of saving all of civilization. Indeed, a much more ambitious program would be easily justified.

The cost-effectiveness drops as one tries to deal with smaller asteroids or with comets. Smaller asteroids constitute only about 10 percent of the hazard, yet detecting them requires new, larger, more expensive telescopes. Detection of long-period comets may require very expensive, state-of-the-art telescopes in order to give us sufficient warning time to respond. At some point, it becomes prohibitively expensive to protect ourselves from every last near-Earth object (NEO). I don't know where the crossover point is.

Benny Peiser: There seems to be a real paradox with our perception of the impact hazard. While three-quarters of the overall NEO risk is due to large asteroids, the most likely impact to occur in the foreseeable future will be caused by a small asteroid.

According to traditional risk analysis, it simply does not make any sense to fund a search for smaller NEOs. The cost of such a search is exceedingly disproportionate to the economic cost caused by small and medium-scale impacts. In other words, as the price tag for the search goes up, the extent of the damage you prevent goes down. If we stringently stick to this line of argumentation, we might just as well stop funding any NEO searches beyond, say, objects smaller than 200 meters. In a nutshell, this seems to be what Al and Clark are suggesting.

The logical conclusion of this simplistic cost-benefit analysis is straightforward: the estimated 100,000 NEOs in the 50 to 200 meter class should be ignored altogether, because they pose no greater risk than the other major disasters that we have come to accept.

Now the societal and political problem with such an attitude is that we are constantly bombarded by smallish NEOs. In contrast to more familiar natural disasters, impacts are totally random in time, location, and degree, and therefore are much more petrifying than anything else nature is throwing at us. The more our astronomical and space technologies advance, the more we become aware of the considerable number of small impacts that occur each year in the Earths atmosphere.

From time to time, a small object hits the ground with a boom. Nobody knows when or where this is going to happen, but happen it will. Thus there is the realistic risk that NASA -- and much more so those space agencies that are inactive regarding NEO searches -- will be brought to task for failing to pay attention to small NEOs. Apart from the monetary, social, and military risk small impacts pose to our fragile societies, there also potentially seems to be a political cost for inaction.

Clark Chapman: Maybe society should spend the same resources per life saved on mitigating small asteroid impacts that we do on airline safety or safety of nuclear power plants, in which case we need to do much more. Or maybe, when analyzing the facts, our political leaders will choose instead to give small asteroids the same priority they give to protecting agricultural workers and miners from the hazards they face, or the priority given to protecting susceptible people from the flu in other words, next to nothing.

This is a country where more attention was given (in autumn 2001) to half-a-dozen deaths due to anthrax than to the more than 30,000 preventable deaths due to the flu. It is up to the citizens of the world, through their political processes, to decide how to deal with the impact hazard.

For every impact-produced tsunami that might kill hundreds of thousands or a million people, there will be hundreds of equally deadly earthquakes, floods, and other natural disasters. That doesn't mean that we should do nothing about asteroid tsunamis, but it puts the problem into perspective.

Alan Harris: As Clark says, it's a "no-brainer" to make the case for finding the majority of asteroids greater than 1 kilometer in diameter. It's more questionable whether it makes sense economically to find much smaller ones. But as Clark points out, what societies demand and what policy makers choose are not always rationally justifiable.

It is no more than coincidence that the maximum risk is from the largest objects. It could just as well be the other way around. In that case, discovering the larger asteroids would only modestly reduce risk. The biggest risk reduction would have to wait for more capable surveys. But luckily in the case of the impact hazard, the greatest risk happens to reside in the easiest (largest) bodies to discover. In this case we face a very rapidly diminishing return.

Joe Veverka: While a survey of objects 1 kilometer in diameter or larger can be carried out in a moderately short time, I think cataloging objects in the 100-meter category is much more important. First of all, these smaller objects have about a hundred times greater chance of causing mischief by interacting with Earth. Also, it is more useful for us to worry about 100-meter objects, since we can imagine potentially effective and affordable defenses against such impactors. However, when it comes to bodies 1 kilometer in diameter - which on average will be one thousand times more massive - the idea of diverting them or blowing them up in the foreseeable future still borders on the fantastic.

Alan Harris: It is only logical to start with the easiest measures, and then if resources are available, advance on to the more challenging measures. In this case, find the big objects first and then work down to smaller sizes. Of course, the present surveys dont just search for large bodies. That's just a natural consequence of optical surveys that the larger objects are easier to find. It's like fishing: you catch what you catch. And we certainly are not "throwing the little ones back." We catalog everything we can find.

Benny Peiser: The Spaceguard search program should and most likely will evolve gradually. Pragmatically speaking, the goal for the next 15 years should be to extend the terrestrial searches to smaller objects. This is more or less in line with the far-sighted proposals made by the UK Task Force on near-Earth objects. More importantly, we should make every effort to exploit the decreasing cost of satellites launches in order to establish the first tier of an effective space-based NEO detection system. This would guarantee that we could gradually address the potential threat of cometary impacts, as well.