By Duncan Steel 09/05/2019

[avatar user=”duncansteel” size=”thumbnail” align=”right” /]

It’s now scientifically possible to predict potential asteroid impacts years in advance. But knowing that such a calamitous event is going to occur, due to the clockwork of the heavens, presents its own problems. Can we divert it, and if so, how? Similarly, if the impact is inevitable, can we model what is going to happen far ahead of time, and so plan better for this rude intrusion into global affairs?  

In my preceding blog post I described the Planetary Defense [sic] Conference (PDC) that I was attending at the University of Maryland: a biennial meeting about the hazard posed to humankind by asteroids and comets, which we know strike the Earth from time to time with calamitous consequences. Just ask the dinosaurs.

Smaller objects than the 10 km leviathan that saw off the ‘terrible lizards’ and heralded the rise of the mammals (and eventually us) slam into our planet much more often. As part of the PDC an exercise is run in which day-by-day over the Monday to Friday meeting a fictitious scenario is introduced and then updated on the basis is what we could do in terms of investigating the threatening object, including dealing with it in any viable way. Attendees include not only astronomers and space scientists like myself, but also a range of experts who would be involved such as space lawyers (yes, they exist), representatives of international bodies such as the United Nations (most especially staff from the Office for Outer Space Affairs, which is based in Vienna), and disaster responders and planners from institutions such as the U.S. Federal Emergency Management Agency (FEMA). There were about 300 there in all, with just two nations in the southern hemisphere (Uruguay and New Zealand) being represented.

This exercise occupies about one-third of our time across the week (the PDC finished last Friday), with the larger part being involved in the presentation of talks and papers (including many posters) on specific aspects of the science, technology and other matters connected with the NEO (near-Earth object) impact hazard, and what it means for the future of humanity. As the t-shirts say, “Asteroids are nature’s way of asking… How’s that space programme coming along?” Sooner or later we will need to intercede, if we are not to go the way of the dinosaurs – or the many other species that have been slammed into extinction by past asteroid and comet impacts.

This year’s exercise – repeat, exercise – involved a relatively small asteroid being identified earlier in 2019, and on a course to run into our planet at hypervelocity in another three of its orbits around the Sun, in 2027. Its orbital parameters were craftily chosen by Paul Chodas of NASA’s Jet Propulsion Laboratory (JPL) to pose many problems for us to chew over, not the least of which was the fact that the circumstances (a small asteroid moving mainly far from Earth and observable even with very large telescopes during only a few intervals before the impact might occur) made it difficult to say anything definitive until very late in the piece.

In the text (and graphics) that follow I will tell you some of the story of this exercise; if you’d like to see/read more, here is the JPL webpage for the 2019 Planetary Defense Conference Exercise. You might also like to peruse the International Asteroid Warning Network (IAWN) website; the NASA Planetary Defense website; and see also the ESA website.

To say it again, this asteroid is fictitious, and this was all merely an exercise; but the scenario complies with reality in that just such a thing could occur, and the science and technology involved are all ‘correct’.

We start on PDC Day 1 (29 April 2019) with the announcement to the public that an asteroid discovered by a team in Hawaii on 26 March has been found to have an estimated one in 100 chance of hitting Earth precisely eight years later, on 29 April 2027. Our best guess at its size, based on its brightness and an assumed albedo (reflectivity) range, is 100 to 300 metres. This is large enough to cause regional- to continental-scale devastation: the energy released on impacting Earth would be 100 to 2,000 megatonnes (cf. the Hiroshima atomic bomb released about 13 kilotonnes; the largest hydrogen bomb ever tested had a yield of about 60 megatonnes).

With the reasonably-precise orbit available, astronomers realise that this asteroid will only be observable during a handful of intervals over the next eight years, and would need very large talescopes to do so: eight metre apertures or more. Such behemoths are located in Chile, Hawaii, and the Canary Islands. The Hubble Space Telecope could also be used; but it will not remain in orbit for many more years. It is hoped that the new James Webb Space Telescope will be launched within a few years, and so could be pressed into service at some stage to track the asteroid, which is very faint and receding from Earth.

Most often when an asteroid is discovered and found to have a finite likelihood of colliding with our planet within the next few decades, additional positional information gained by observers enables us to say for sure that it will miss. That is, an initial value of one per cent (as here) as the collision probability for some known close appraoch to Earth means that 99 times out of a hundred it will miss, and the accumulation of data over weeks, months or years enables us to exclude it as a near-term risk. In my previous post I wrote about asteroid (99942) Apophis, which will pass very close by Earth in 2029 – but we do know that it will miss us.

In the case of this fictitious asteroid, however, as more observations are obtained the derived collision probability goes up. By 29 July 2019 such observations result in an upgrading of the formal collision probability estimate to one in ten: a 10 per cent chance that it’s going to strike our home, and cause havoc.

A note on how the probabilities are derived. Despite there being perhaps hundreds or thousands of positional measurements of an asteroid spread in time, there is still a set of finite uncertainties in its orbital parameters and this results in an ‘error ellipse’ that may be drawn around the location of the Earth at the time that the asteroid is due to come by us. It may seem paradoxical that we can say very accurately when this approach will occur, and yet we are unable to say for sure whether it will hit or miss. But that’s the way it is.

In the diagram below the largest ellipse is the initial one, and the point is that the collision cross-section of the Earth (its geometrical cross-section enhanced due to gravitational focussing being able to ‘suck in’ a passing asteroid which would otherwise miss) is about one per cent of the area of that initial, large ellipse. With more tracking data the ellipse reduces in size (‘later, more accurate prediction’) and now the green dot representing Earth is supposedly about ten per cent of that reduced ellipse’s area. With yet more data the ellipse collapses even further (‘Still more accurate prediction’) and the green circle of Earth is no longer within it: we now know that the asteroid will miss the Earth, just as Apophis will in ten years’ time.


At the end of July 2019, then, just four months post-discovery, we know that there is a worrying one-in-ten chance of major cosmic fireworks in 2027. From the anxiety perspective, astronomers also know that more data collection enabling a definite yes or no – hit or miss – will not be feasible until very late in 2020. It would really be helpful if some radar data were possible (a rule of thumb in the field is that one radar detection can be worth several years of optical telescope tracking, in terms of improving our asteroidal orbit determinations), but currently the only functional planetary radars are located in the northern hemisphere, at Arecibo in Puerto Rico and at Goldstone in California, and neither can access the problematic asteroid.

The reality of the situation with regard to asteroid paths which we know are going to come close by the Earth and possibly hit us is that our uncertainties reduce until there is a narrow line which we know will contain the asteroid as it reaches the Earth’s orbit… the problem is that as of yet we don’t know where across that line the asteroid will be. In the PDC exercise the red line in this diagram shows that part closest to Earth. As we get better knowledge of the asteroid’s orbit either the remaining viable part of the line is off the planet (i.e. it will miss), or else the line shortens and eventually is entirely on the Earth (i.e. it’s going to hit).
A side-view illustrating the same thing, for the instant on 29th April 2027 when the fictitious asteroid will either strike the Earth, or else pass safely by. The red line is composed of discrete dots because it is drawn up using a Monte Carlo simulation with the statistical nature of the uncertainties being incorporated.









The ‘risk corridor’ – the line along which the impact might occur – initially stretches from near Hawaii, across the contiguous U.S. and the Atlantic, and then various nations in west Africa.  


Whilst the risk corridor is narrow, the blast that will occur will spread the damage transversely, so that a large number of people are in severe danger.


For the identified risk corridor as at 29th July 2019 (when the impact probability was still 10 per cent) it is feasible to estimate how many people would be affected by an impact at any particular point along that corridor.
The U.S. is by far the most-prepared nation with regard to the possibility of an asteroid or comet impact anytime soon.

An aside. I did not know in advance that the dates chosen for the fictitious asteroid impact was to be 29th April in 2027, but I did know that the first day of the PDC on 29th April coincided with a peculiarity in Scandinavia. Once-upon-a-time I worked for the European Space Agency, and was living in Sweden. In that country each day has a name associated with it, such that people with that appellation get to celebrate what might be thought of as being a second birthday. And April 29th is Tycho (or Tyko) day. Now, one of the most prominent (and youthful) impact craters on the Moon is called Tycho, for Tycho Brahe, the Danish astronomer who in the late sixteenth century made a vast number of accurate measurements of the positions of stars and planets. After Tycho moved to Prague his catalogue of observational data was used by Johannes Kepler in deriving his eponymous laws of planetary motion.

Tycho’s observatory was on the island of Ven (or Hven), in  the strait between Denmark and Sweden. The southernmost province of modern Sweden (Skåne or Scania) was then, four centuries ago, part of Denmark. Anyhow, there is another thing about how particular dates are regarded in Scandinavia… Tycho Brahe days (there are about three dozen of them spread over the year, and, yes, 29th April is one of them) are considered to be unlucky.

So, the PDC starting on 29th April with a chosen date for a fictitious asteroid impact being that day eight years hence would seem to be felicitous. Or not, depending on your attitude to superstitions, and catastrophic asteroid arrivals.

It takes time – years, generally – to design, build and prepare a satellite for launch, especially for a deep space mission. With a 10 per cent chance of an impending asteroid impact, do you build and launch (in a rapid fashion) reconnaissance space probes to be sent to visit the threatening object and collect information? We need to know our enemy, and soon. Such matters were debated and decided on days 2 and 3 of the PDC. By Presidential decree the U.S. space agency NASA is given two billion dollars for an emergency spacecraft programme, and US$500 million for ground-based tracking and other observations. Various satellites already in Earth orbit are used to study the asteroid; for example thermal infra-red data can narrow down our estimate of its size.

As the end of 2020 approaches it had been hoped that with more positional measurements and a longer time-base our shrinking uncertainty in the asteroid orbit would result in it being proven that it would miss in 2027. The opposite occurs. It is now (at the end of 2020) certain that the asteroid will hit Earth, at a location near Denver on 29 April 2027. (Now it’s getting personal: for some years I lived in Colorado, working on NASA’s Pioneer Venus programme.)

Stepping forward to 30 December 2021, the first reconnaissance spacecraft mission flies by the asteroid. Images show the asteroid to be about 260 metres long and 140 wide, apparently a contact binary (we already know of several asteroids like that, being two-lobed). Its colours indicate it to be an S-type (stony; but it’s a bit more complicated than that). The impact energy estimate is now between about 150 and 600 megatonnes, being uncertain because we don’t know its mass. This asteroid could be a solid rock, in which case its density might be around 3 grams per cubic centimetre, or it could be a ‘rubble pile’: an agglomeration of boulders held together by self-gravity, with many voids within it. Whichever it is would make a huge difference to what will happen as it plunges into our atmosphere at about 69,000 kph.

As more observations of the asteroid accumulate, by the end of 2021 it is known that the impact will be near Denver, Colorado.
Given definitive knowledge of the impact location, it is is feasible to map the most-severely affected areas. The above pertains only to ground impact damage; the intensity of the radiation from the blast would be expected to start forest wildfires over a far larger region in the Rockies, for example.

There are just five days in the PDC, so we need to step ahead at pace. Various international agreements prohibit the deployment of nuclear weapons in space. Likely most of the people who understand the asteroid problem would prefer, in a real-life situation similar to this exercise, that stand-off nuclear explosions would be used to deflect or disrupt a threatening asteroid, but in this scenario we exclude such tactics. This means that the only means at our disposal is a kinetic impactor: slam a high-mass satellite as fast as you can into the asteroid, and try to knock it off course.

The preference is for a straightforward deflection: the asteroid remains intact but is diverted by enough to achieve a miss of our planet. It is known, though, that a weak asteroid might be disrupted. If it were indeed a rubble pile then shattering it into myriad lumps would be a useful outcome, in that although most or all would still hit the Earth they would be spread out, and smaller rocks explode higher up in the atmosphere. As it is, modelling of the entry physics of a 200-metre solid stony asteroid at the speed in question indicates that the asteroid would release most of its energy in an airburst at an altitude of about 1,500 metres (5,000 feet) above Denver. This is not a good thing. Such an airburst would cause more widespread damage than a projectile reaching the ground intact and then exploding, excavating a crater about 3 km across.

In the event (or, at least, in this fictional exercise) the asteroid is hit by three intercept spacecraft at the end of August 2024, and the first days of September. Six such spacecraft were launched by NASA, the European Space Agency, and other nations participating in a global effort to push the asteroid off-course, but half of them fail to make it.

On 3rd September 2024 it is announced that the three interceptors which hit the asteroid have successfully deflected the main mass, but split off the smaller lobe which is still on a collision course with Earth. The impact location is not yet defined. This fragment is estimated to be 50-80 metres in size. The last time something of this size hit Earth was in 1908, when the Tunguska event occurred in Siberia, laying waste to thousands of square kilometres of the taiga. No scientific expedition reached the region until 1927, so much is still unknown (or unknowable), but estimates of the Tunguska explosion’s energy range between 3 and 15 megatonnes. Apparently it was either a small asteroid around 50 metres in size, or perhaps a slightly-larger fragment of a comet.

Three kinetic impactors ‘attack’ the asteroid near the end of August 2024. These succeed in deflecting the main portion of the asteroid such that it will miss the Earth in April 2027, but a 50-80 metre fragment is knocked off and continues on a path intercepting our planet on 29th April 2027.
Following the kinetic interceptors hitting the asteroid in 2024, a fragment is calved off and this it is found will hit somewhere along the red risk corridor shown above.


The number of people directly affected by the asteroid fragment varies depending on precisely where it arrives, and this is unknown until a couple of months prior to the impact on 29th April 2027.


Back to the Planetary Defense Conference exercise scenario. It is now ten days before the asteroid arrives, 19th April 2027. For two months it has been known that the region around New York City would be struck, but it has only been with the availability of radar data from the past couple of days that the epicentre has been determined as being Central Park. The impact energy will be 5 to 20 megatonnes – still uncertain because we have no firm determination of its mass.

For the finale in this asteroid impact exercise, the 50-80 metre fragment knocked off the complete asteroid carries on and is predicted to explode above Central Park in Manhattan, causing modelled blast areas as above; the overall damage would be much more widespread that this, however.

The obliterated area will be more than a thousand times that of the 11 September 2001 attack. By ‘obliterated’ we mean nothing much left standing. There is no chance of anyone surviving if they are within the central area where the overpressure caused by the asteroid explosion will kill everyone, unless some choose to remain in pressurized underground bunkers.

It’s at that stage that the space scientists and physicists such as myself hand over to the disaster planners. How to evacuate over ten million people, many of them in hospitals and nursing homes, and where to shift them, is not something that I’d like to think too hard about. Indeed it’s not an evacuation, as such, but a removal: they are never going back. This is, however, something that is seriously considered and planned for in the U.S., and for me it was illuminating to listen to the thoughts of the personnel from FEMA and various international agencies in this regard.

Joel Marks, professor emeritus at the University of New Haven, wrote an article almost a decade ago about NEO defence entiled “First, Do No Harm” – a phrase often quoted as part of the Hippocratic Oath of the medical profession. So, how does one judge the intervention pondered in this asteroid impact scenario? Denver and surrounds were saved from a cataclysm, but as a consequence New York City received a lesser (but still calamitous) blow. This is one of the reasons we run such exercises: to uncover problems, and chew them over. What if the initial diversion using the kinetic impact spacecraft had left the asteroid intact, but shifted its path away from Colorado just enough such that it then impacted a densely-populated part of Nigeria or Cameroon?

A more-complete assessment of the asteroid impact exercise is available on this NASA webpage; there are many reports of the outcomes in the mass media, which I’ll leave you to track down yourselves.

The PDC wound up last Friday, with one of the last announcements being the selection of Vienna to host the next such meeting in 2021. There’s lots still to talk about, and plan for, because all sorts of things – not all of them beneficial – can come out of space.

Post scriptum: After the PDC ended, I had a weekend free to do some touring about. I chose to drive eastwards, past Annapolis and into Maryland’s Eastern Shore: the east side of Chesapeake Bay. There I stayed in the historic town of Easton, and drove from Oxford to Cambridge in less than thirty minutes; continuing south I whizzed through Vienna, Salisbury, Princess Anne and Pocomoke before passing into Virginia (just), arriving in Chincoteague to have an ice-cream and also visit NASA’s Wallops Island launch complex on the Atlantic coast.

The next day I visited Saint Michaels, and Tilghman Island, on the banks of Chesapeake Bay. Why did I mention Chesapeake Bay (again)? Well, it’s an impact crater, you see, formed when a massive asteroid or comet smacked into the area about 35 million years ago. This huge event is associated (along with the bigger Popigai crater in Siberia) with the end of the Eocene epoch, and the Eocene-Oligocene extinction event. It’s not only the dinosaurs that have had their reigns ended by cosmic impacts. Will we?

0 Responses to “Imagine an asteroid impact due in 2027: How would you tackle it?  ”

  • I would get to as high ground as possible. There are dead forests up north on the keri keri peninsula that have all the trees in a line. Knocked over by massive waves. Waves from a space rock get massive in the 100’s of metres, not 10’s like ones from an earthquake.

  • Thanks Derek. Good points. For an airburst caused by a ‘small’ asteroid like the one in question, it is anticipated that the tsunami would be small-ish. On the other hand, bigger asteroids/comets (which reach the ground/ocean largely intact) are expected to cause phenomenal tsunamis – just as you indicate. It has been suggested that various huge boulders lifted over headlands on the NSW coast got where they are now located through asteroid-provoked tsunamis.

    Some years (almost twenty) ago I was asked to write an article for an adventure sports magazine published in London and entitled Adrenalin, this article being about asteroid-impact-induced tsunamis and whether you could surf them. That article can be seen here.