Table of Contents
If it still don’t work, I gets a bigger ‘ammer.
Humpty Dumpty Had a Great Fall
Super Vision and Supervision
Don’t
be bitter, young sir. If you want to go on living you have to be alert.
And you can't be bitter and alert at the same time.
William Tenn
We live in a flyspeck we can hardly conceive, a flyspeck that we have not yet mastered, in a universe we have not yet properly conceived. Depending on how you measure our flyspeck, it is a something like a light day across, in a universe billions of light years across, but that is too big for coherent, coordinated control or occupation.
There are implications for the topic of this essay; control of dangerous collisions of natural bodies in our local space, but even so limited a field is spread over many light seconds, even light minutes, and some of the control functions cannot tolerate latency of more than a small fraction of a second. For personal control we generally want millisecond response, which means that even at lunar distances, some forms of local control are impossible, and demand local human presence or, failing that, automation.
Until recently, local presence was the assumed option; when Arthur C. Clarke proposed satellite communication system, he imagined satellites occupied and operated by humans, but even that was preempted by the first satellites, which were unmanned. Less than a century later, we not only have more advanced direct automation, but are developing AGI for meta‑control, in which devices can rationally control events and processes on teleological bases.
For the alertness aspect of this essay, that demands continuous information coverage and reactions over at least many light hours — light days would be better, but some aspects of control demand latency limited to fractions of a second; one‑second control of a ship might suffice, but in controlling a car or aircraft, or even walking, a one‑second delay could be fatal.
Conversely, human control that demands constant alertness, sometimes exposed over long periods to stress or danger, or problems of toxicity, radiation, or temperature, can increasingly be relegated to AGI systems that do not require bulky, complex life support systems. In increasing numbers of such cases human operation is inappropriate and impracticable.
This is by no means an academic observation; in recent decades it has become increasingly obvious that humanity on our planet is living on borrowed time, that there are threats of destruction, with plenty of evidence of their reality, their repeated occurrence throughout past millions of years, millennia, decades, and even as recently as the Chelyabinsk meteor.
Even ignoring the relevance of dino‑killer asteroids, a single impact on the scale of the Tunguska event striking near a major city, could cause deaths and destruction greater than the cost of a global system of spacecraft and AGI control for centuries. It would manage indefinite continuous supervision of Solar System activity within the Oort cloud and sentinel duty, including permanent watch for hazardous bodies.
Accordingly, this topic is a classic example of the need for reevaluation of the appropriate technology. At first it was the humans, then machines, then programmed computer algorithms. Currently emerging requirements demand currently emerging technologies, and the outstanding example is AGI.
Fusion Skeet Shoot
If it were done when 'tis
done, then 'twere well it were done quickly
Macbeth
Risk management is a vital discipline (literally, and in more senses than one), but it is one that most people, including some who are professional decision theorists, do not properly understand.
Politicians and opportunists are practically invariably mobile disasters, with power to affect outcomes that it is beyond their training to appreciate, and consequences beyond their rationality to assess. When we combine low short-term probability and indefinitely high impact threats, these effects are at their worst. “It hasn’t happened yet, so why worry?” or “HAS happened in the past? So what? We are still here, right?” or “Don’t be ridiculous! That would cost billions! Almost as much as a day’s publicity for the next election!”
Also, one gets the opposite, the “Chicken Little”, and “Golden Fleece” attitudes: “No! Just think if it went wrong! What if we set off the next Big One quake when we launched the bomb, and think of the radioactivity and the plastics pollution! How can you prove that it won’t go wrong?”
For some decades now, I have been interested in the risk of cometary or asteroid collision. It is a textbook case in several ways.
· We know that dino‑killers have struck Earth in the past and will strike again; we do not know when.
· We know that the worse the collision, the lower the probability that it will happen in any given year.
· Most of us think that that is the same as proving that nothing will happen; certainly nothing worth doing anything about.
· We can recognise all sorts of worst cases. A really hefty hit on say, Yellowstone could wipe out all air-breathing vertebrate life on Earth: not from the resulting vulcanism, but if the seismic wave from the impact passed round the planet and focussed on the other side of the planet into deep ocean, causing gross release of dissolved CO2, it could release enough to poison us all. Even if the mass struck nicely in the deep central Pacific, the CO2 could have worse effects than the resulting super‑tsunami. It could be another billion years before intelligent life resurfaced, in no predictable form, and it might never resurface at all. The details we cannot predict, but we can be sure that we would like it not one bit.
· We are inclined to play numbers games of the type one uses in real‑life games theory of the type that works in casinos, without accepting that those games are for numerically acceptable pay‑offs. When it comes to risks of losing whole countries or even all of humanity, pay‑offs simply are not worth calculation. We know that there always is a chance that a rogue planet the size of Mercury could be on its way to Earth already, and that that really could be the end, because we could blast it with all the nuclear devices on the planet without deflecting it enough. No species would survive the resulting fireball, but we also know that dino killers, too big for us to deflect with conventional explosives, and almost as deadly, could indeed be deflected with existing technology — technology that never existed before on this planet.
· We know that there are no short‑term serious threats from nearby regions in our ecliptic, because the relevant scientific authorities managed to get mandates for such surveys, but we also know that there also can be threats for other quarters, such as from outside the ecliptic, and dark bodies from the Oort cloud or beyond.
· In a choice between a dino‑killer rock and a bit of radioactivity released in a launch accident, there is no comparison. Look at Chernobyl and Fukushima, two of the worst, and most irresponsible, nuclear accidents in history so far, either of them worse than the worst that could come of a failed launch of nuclear material into space: neither of them made a dent in the respective country’s traffic accidents or deaths from booze or smoking. I am no fan of traffic accidents and the like, but if you lack the perspective on such comparisons, you disqualify yourself from rational discussion of the issues. The dino‑killers are our likeliest, most unacceptable, avoidable catastrophe in history.
· We know that the deadlier the threat may be, that is to say, the bigger and the faster, the rarer it will be: and by that very fact, the harder to anticipate. But rarity is no consolation when it happens. What is more, the worse the catastrophe, the earlier we must detect it if we are to do anything to prevent or mitigate it. Furthermore, the worse it may be, the earlier it must be deflected if we are to minimise the harm. This is obvious, because the target area (that is to say: you and I) is over 12000 km in diameter. That means that, depending on the direction in which it is displaced, the necessary distance of redirection to make the missile miss our planet could be anything up to some 12000 kilometres. If we were to catch the invader to deflect it say, a year ahead of the impact time, then pushing it aside at an easy walking pace would be adequate. But if we only detected it in time for our projectile to hit it about a month before impact, we would need to accelerate many millions of tonnes at quite a smart pace. The problem of detecting the invader and calculating whether it is on its way to us, is bad enough, but what is worse, is to organise our defences to reach it in time, and so far the lead time seems to be a few years.
· DART, the Double Asteroid Redirection Test, was a brilliant exercise, and serendipitous to boot. It not only showed that interception was possible, though it took nearly a year for the projectile to reach its target, Dimorphos, after taking a few years to launch the project. In fairness, that lead time to launch could be shortened to less than a year if the necessary infrastructure had been in place when the target had been detected. In fact, if the infrastructure were along the lines of those described in this article, the craft should have been waiting in space in advance, and the launch could have been begun within days of detection of the target, while the course and time of arrival were still being refined.
Trajectories
Even if
you’re on the right track, you’ll get run over if you just sit there.
Will Rogers
Although the target we must defend (our planet) seems very big, a 12000 ‑kilometre globe in a 450000000‑kilometre orbital path has only one part in about 30000 of being struck accidentally by any random shot at that path, which sounds either like very good odds, or very poor, depending on your viewpoint. Of course Earth is being struck all the time, though mainly by dust and pebbles.
But there is one reassuring aspect to that reflection: we are by now amazingly precise at predicting the path of any particularly large ballistic object that we can observe. And we can tell how far we need to deflect it from its path, and how fast, if we are to protect ourselves if we find it aimed at us. In short, the incoming object is, as it were, on a tightrope to reach us where we happen to be in our planetary orbit when it arrives. Almost any disturbance sufficiently early and well directed will push it off its tightrope.
That principle is vitally important.
Not only will any sufficient push across the trajectory of the incoming mass make it leave the tightrope altogether, but any sufficient push along the trajectory will equally effectively make it miss because it will arrive too early or too late to hit the planet.
Every particle of the incoming body gets affected by the interception event, and accordingly each one gets pushed off its tightrope, no matter in which direction.
It might help to visualise the effect with an imaginary model. Suppose one drops an unlit sparkler on a trajectory that ends on an exposed square inch or so of ground. Mark the spot. Then light the sparkler and throw it up into the air. Of the thousands of sparks flying backward, forward, and sideways, hardly any will land on that same square inch where the stick had fallen, and any stray spark that does happen to land there, would be minute compared to the stick.
The same principle applies to a disintegrated asteroid: the debris does not flock to the target; it scatters away from the original trajectory into the void, striking hardly anything corresponding to that first trajectory. And necessarily of course, each fragment would be small in comparison to the parent body.
With any incoming asteroidal mass that gets fragmented, the effect is still more marked: the fragments all are moving apart; and every one is moving away from any trajectory that ends where Earth will be, no matter whether those fragments were driven by collision, blast, or fracture. We probably would not even have a meteor show worth staying up for.
If our explosion hits early enough, particles will be so scattered that the planet can pass harmlessly between them.
This article deals mainly with bodies sufficiently urgent to deal with by dramatic measures such as explosives and impacts, but it is important to remember that there are alternatives, in particular if the trajectory is determined long enough before any anticipated impact, for instance with a few orbits around the sun before threatening Earth, one could use gentler means, such as attaching a reaction motor, or several kilometres of conductive wire, leaving it to the solar wind to drag the object off trajectory. Those two could be combined, so that say, the reaction motor could be attached by a conductive spring many kilometres long; it would continue to work long after the fuel were exhausted.
Such items need to be borne in mind for perspectives on real, but rare events that need attention.
The Road Least Travelled
Experts have
their expert fun
ex cathedra
telling one
just how nothing
can be done.
Piet Hein
Many schemes for asteroid impact avoidance have been proposed; some wish to push or pull the mass in some direction or other, some to explode it into vapour, some by directing lasers to divert it by vaporising just its surface with lasers or nuclear bombs.
And some spend all their time and breath explaining just how nothing can be done, and therefore why nothing should be done.
Well, the beautiful DART experiment showed that one class of thing really could be done, and surprisingly effectively at that. So everyone seems to have decided prematurely that the problem was solved and we could relax.
This article was inspired by DART, which has pioneered a practical technology for deflecting asteroidal bodies of up to about a kilometre in diameter. The vehicle used in the test was inert apart from its propellent, and it was directed to strike the incoming body head on. This meant that if it were used in that way in a genuine emergency, it would have retarded the incoming, and if it did so early enough, then the body would have missed Earth because it would not have crossed the Earth orbit till after the planet had passed. (The complexities are greater than that suggests, but they do not affect the relevant point: that the body still would miss Earth.)
People unfamiliar with the principle would prefer to have hit the body side‑on to make it pass by on one side, but actually, the head-on impact is the most efficient in converting the energy of its momentum into the necessary acceleration of the incoming body.
Possibly unexpectedly, it turned out that the effect of the DART collision was vastly greater than expected — something like thirty times greater. The reason was not hard to guess, and showed beautifully in the footage of the collision: at the speeds involved, the missile literally splashed out a huge gout of the asteroidal material in the direction that the missile had come from. The dust and gravel of the splash was necessarily far greater than the missile itself, and the reaction, the “kick” of the splash, acted like rocket propulsion.
This was a broad hint. No other principle of deflection could have been more efficient.
But there was room for more. Although this method is estimated to be good for bodies up to a limited size, it would not in itself be suitable for far bigger bodies.
However, the associated implications are not reassuring.
If you could intercept the incoming body enough years before it hits Earth, the DART technique still would work for much larger bodies, but we cannot plan on intercepting bodies light years away. Even within the solar system we might not be able to be sure whether a very distant body will hit the Earth at all; in fact, a bad shot theoretically could turn a near miss into a hit. And by the time that we were sure, it might be too late for us to deflect the incoming body with a DART‑like missile.
So the question is: what else could we do?
We could send a more massive, faster missile, a javelin instead of a dart, so to speak. This would work if done properly, but in fact, that is not as simple as it sounds. On the other hand, if you could instead give your missile an extra mass of fuel and aim it to hit at a much higher velocity, we could exploit the fact that the energy of a missile increases with its mass, but increases with the square of its speed.
But even that sort of design is limited.
Another improvement might be to add a payload of as much mass as is practical, of a suitable high‑explosive bomb to the missile’s payload. That would give you the same maximal payload as an inert mass, but if the explosive, perhaps a tonne of RDX or TKX-50, were timed to explode when it had penetrated deep into the target after a few microseconds, it would expel a far larger reaction mass than an inert missile would, and expel it far more energetically than any passive impact could have done. We would have increased the mass of ejected material by tonnes, but the effect of the increase in the speed of the ejected material would be squared.
That bomb need not be the only option either; you could follow the first missile with another one if the body was still on target and mainly intact.
Which of these routes of loss of forward momentum is most important, depends on the parameters of the impact. As long as the vector sum of the impact is opposed to the vector of the incoming object, there will be a matching loss of the incoming momentum. There also may be a matching loss of energy in diverting momentum to accelerate material transversely to the incoming body’s trajectory. This was vividly visible in footage of the DART impact.
The effectiveness of the impact of the intercepting missile depends largely on how efficiently energy is diverted from the momentum of the incoming body, and how optimally the internal explosive effect can be placed inside the target. Ideally an explosion at the centre of mass maximises the forward propulsion of part of the mass of the incoming in its original direction, but in a harmless form such as dust. But the mass that the missile has passed on the way in, amounts to reaction mass in reducing the momentum of the hinder part of the incoming body.
The optimal effect probably would be if the missile penetrates to the centre of mass of the incoming, and explodes there, as a combination of the impact energy plus any explosive payload of the missile. Ideally the front half of the incoming then would act as the reaction mass, generally harmlessly pulverised. Then the rear half, whether shattered or not, but in any case greatly reduced, would be maximally slowed, removing its trajectory tightrope far from planet Earth. In any case, there would be a considerable mass of shattered material scattered radially, and any lumps in that cloud would never get near to the original trajectory again.
The optimal allocation of the mass of the missile to fuel, rather than to an explosive payload, is not a simple matter. Ideally every joule of the fuel that went into accelerating the missile would go into the impact, and the fuel mass greatly exceeds the payload mass. But it gets worse.
Suppose that we have a choice of adding to the missile, a tonne of payload in the form of say, TKX‑50, or adding to the momentum of the missile by increasing its fuel load. The latter would be cheaper and possibly safer and would avoid the problem of the extremely tight timing of the detonation, not to mention the possibility of something going wrong on the way into the centre of the incoming mass, such as hitting a large lump of nickel‑iron that wrecks the warhead.
But that too is not a compelling choice; if the extra fuel in lieu of the explosive payload pays off in the missile being too fast, and accordingly too penetrating for the nature of the target, it could go more than halfway through, even right through the pile. That certainly would cause considerable slowing to the residual mass, and probably would disperse much of the rubble at right angles to the trajectory, but still would direct hardly any of the reaction mass in the preferred direction. It might even accelerate the main mass forward, still very likely to miss Earth, but at best it would be an ineffectual outcome, if not actually excessively dangerous.
And, even if a tonne of TKX‑50 would be less energetic than an extra tonne of rocket fuel, it would be more brisant, would have more shattering power placed more precisely where desired, more likely to break apart the target body as required, than mechanically pulverising rock all the way in.
We can solve all those riddles systematically as they arise, but there is one aspect that must not lose sight of, even though it is increasingly rare, as the parameter increases: the size of the incoming object. Suppose we were not dealing with a ping pong 1‑km diameter asteroid, but a real dino‑killer 10 km or more in diameter.
Teratonnes. Your real dino‑killers.
By then it is time to up the ante seriously.
What I prescribe for much larger targets instead of a chemical explosive, is to use a nuclear explosion, and not a toy nuclear bomb burning plutonium either; such bombs are both dirty in the isotopes they produce, and expensive in fissile materials. So what would be better would be a (relatively) clean fusion bomb (H‑bomb); it uses just enough fission to set off the fusion explosion, producing a far larger explosion, and far larger acceleration than any comparable chemical could provide, whether as propellent, or excavation agent.
If it still don’t work, I gets a bigger ‘ammer.
When the only tool you have is a hammer, every problem looks
like a nail.
Abraham Maslow
The political problems with nuclear devices far outweigh the technical problems.
For example, try asking random members of the public in any nation, about the difference between fission and fusion. If they even know what you are talking about, you probably are in a University town, and most likely in one of the physics departments.
Most of the laity would object violently to any use of nuclear devices off Earth anyway. If you pressed them, some would object to radiation contamination of outer space — an absurdity, but very difficult to explain. Other members of the voting public, better informed, but still by no means competent in the field, would object to sending such things up into space because of the risk of the rocket crashing down and scattering radioactivity all over.
Rockets do certainly fail now and then, and that is why the inert bomb components, including the lithium deuteride etc, would be sent up separately together with most of the bomb body components. If the launch failed, they would do no more harm than any plane crash, and that only where they landed. Expensive anyway, so you could be sure that such a crash would be as rare as anyone could make it.
As for the radioactive fissile component, it is a relatively small part of the bomb, the match so to speak, intended to set off the main fusion explosion. But though it is not inert, it still cannot explode nor even present much of a radioactive threat unless it is set up to do so, any more than a car engine’s parts can start unless they are assembled.
The problem is by no means new, and techniques for dealing with it are well established, either on launch failure, or on returning to Earth. Unused parts of fissionable material can be packed into steel containers sufficiently shock‑proofed so that even if the rocket comes apart and dumps its load in a failed launch, or the failed assembly has to be dumped on re-entry, those lumps of steel will not release their content, no matter when or how they land.
Once in orbit, and with no problems, the bomb can be assembled ready for use in the missile; this should be done by robots with advanced AI, but supervised by online inspection of humans on the ground. It would make no sense to send humans up to do jobs that machines can do better, faster, more cheaply and more healthily. Even if such a bomb were somehow ruined beyond recovery, its inert parts could be jettisoned to be destroyed on re-entry, and its fissile materials, which could not cause any nuclear event on their own, could be directed to the moon or Venus or the sun, depending on practicalities, or returned in their impregnable delivery containers.
Some people argue that if you use a fusion bomb, you need not make contact with your target at all: simply explode your bomb at the right distance away, and the flash will ablate its surface, vaporising it explosively, producing a thrust with far less chance of the target flying apart.
It is an interesting approach, but its advantages are arguable, even if it were conceded that avoiding fragmentation were desirable. Besides, its inefficiency would be appalling: if it could be shown that the thrust from such an explosion amounted to using 0.01% of that bomb’s power, I would be shocked. Major arguments in favour of the idea are that the blast would not disrupt fragile targets, and that in such a project what counts is not efficiency, but effectiveness: if we can afford to send out a megatonne bomb, we can afford some inefficiency, so why do we need such efficiency?
This may be true as far as it goes, but it ignores pertinent considerations. Firstly, there is no relevant limit to the size of possibly incoming bodies, but there are realistic limits to how big a body we can deflect, even using nukes. In contrast to the external fusion bomb, techniques based on the rocket reaction‑mass impact, or even more, an explosive or even a nuclear explosive deep inside the mass, would give enormous efficiency. A deep impact nuclear bomb could shift even a real dino‑killer if we caught it soon enough. The efficiency of the measure sets the limit to the size of threat we can avert. If by deep impact with a nuclear bomb we could improve an efficiency of 0.01% to 1%, that would suggest an ability to protect Earth from targets a hundred times more massive, using the same bomb.
That ratio is far more important than the numbers immediately suggest: the frequency of a threat from an incoming body of a given size varies inversely to the mass, according to power laws with various parameters, but all scales of estimation agree that increases in efficiency that enable us to avert threats a hundred to a thousand times greater, would increase the life expectancy of our civilisations to geological periods rather than millennia. We could rival the dinosaurs.
Furthermore, we can expect our technology to improve over time, enabling us for example to detect and intercept incoming bodies from far greater distances. Our ability to deflect larger targets than anything currently foreseeable would increase disproportionately. We could extend the probable survival of our heritage beyond geological periods, to when planet Earth itself could no longer support life because our sun would no longer have a stable habitable zone. For example, imagine having detected an incoming rogue planet the size of Pluto, large enough to reduce Earth to a glowing ember, and too large to attack directly. By efficient use of penetrating nuclear bombs, we might be able to steer some billion-tonne bodies from the Oort cloud to collide with the incoming rogue a light‑year out. At such a distance a tiny deviation to the rogue’s course could save our planet. The last time Earth encountered such a threat was some four billion years ago. By proper attention to efficiency we could face threats orders of magnitudes greater than by wasting the potential of our tools.
Discussion of those details is beyond the scope of this essay, but not to be dismissed as academic; prospects for extending human longevity to similar orders of magnitude, not only are realistic, but essential to the survival of our species and heritage and the quality of our future.
Humpty Dumpty Had a Great Fall
The only thing to prevent what's past
is to put a stop to it before it happens
Boyle Roche
Bigger bombs, harder hammers, are all well, the sceptic insists, but what if your bomb, or even your DART‑type impact blows your target apart? Then where you had one dino‑killer, you now have a whole flock of them.
That sounds ominous of course; for one thing, not every asteroid has the cohesive strength of a solid lump of nickel‑iron. More often it will be a relatively fragile chunk of rock or even a rubble pile or a muddy snowball with no considerable cohesion at all.
Rubble piles in particular, as in fact was the case with Dimorphos, have been posed as a great concern to interceptors of incoming bodies, because they seem likely to absorb impacts much as massive marshmallows might. This view however, is unrealistically pessimistic; we now have seen vividly that if the target is solid, or a rubble-pile that maintains its cohesion after the impact, it may react like a rocket, with much of its own body material serving as reaction mass. Dimorphos in fact did not split, though it did scatter a lot of harmless dust and gravel.
Even a marshmallow must obey the conservation of momentum, and that implies that the vectors of the various components of the impact all contribute to the outcome. The incoming rubble pile must get slowed down by the reverse scattering from the impact, and it also loses its momentum towards the target, Earth, to particles expelled by transverse scattering. It simultaneously gains momentum from material ejected from the rear of the body, but that of course is trivial in most circumstances.
We already have seen that the path of any distant object on the way to our planet, is tightrope-precise, and that even a slight disturbance, say one part in tens of thousands, either in direction or in time of arrival, will cause it to miss. This has a very important implication for our concern about the consequences of shattering the incoming body: those particles share common attributes, no matter what the nature of their disturbance.
Accordingly the dread of creating a swarm of dino‑killers whenever we shatter an incoming body, is deluded. Not only can we dismiss any fear of the swarm; we want the swarm, the shattered invader, because after the original contact, the bullseye of the original target (typically planet Earth) is the safest place to be.
An incidental concept is that it might be practicable to hit a large, known rubble pile off‑centre with a projectile too small to shatter it, but setting it spinning rapidly enough to shed large amounts of loose rock, small enough or far enough off‑trajectory to ignore. After such an exercise, the remnant’s trajectory would be reassessed to see whether it was still unsafe. If so, the remnant could be assumed to be solid, and probably suitable for a DART‑type deflection.
And so. . .
Most people miss Opportunity because it is dressed in overalls and looks
like work.
Thomas A. Edison
And so... We need to act, and we need to act before the next disaster. Proper protection of our biosphere from threats that range from city-busters to dino-killers, would ultimately work out cheaper than ad‑hoc reactions; how does one price a vaporized city or a Pacific‑wide tsunami?
We need to establish a permanent infrastructure of autonomous, AGI‑controlled detection and interception craft in space. They would pay for themselves anyway by utility functions such as communication and navigation infrastructure, astronomic observation, and mapping the Kuiper belt.
By utilizing the sheer efficiency of deep nuclear impacts, we can shift the life expectancy of our civilization to geological epochs. But that survival is contingent on having these indifferent sentinels already waiting in the void, ready to secure the environment. We cannot afford to start building the hammer only after we see the nail.
Acknowledgement: Though this work is my own in conception and authorship, I used the services of Google Gemini in editing and in colleague mode for criticism, and am gratified at the effect.
No comments:
Post a Comment