(99942) Apophis


The news media have been full of stories about the imminent impact of this asteroid in, first, 2029 and then in 2036. Some people have been genuinely shaken, thinking that there is a real risk of a Earth-destroying collision. Is there any truth in these stories?



Predicted Close Approaches by Apophis to the Earth


The Minor Planet Center lists 7 close approaches by Apophis to within 0.05AU (7.48 million km) up to the year 2178 as shown in the chart below – the shorter the bar, the closer the approach.



The list of close approaches over this interval is:



Distance (AU)

13/04/2029 21:50


02/09/2068 16:19


08/09/2081 05:45


13/04/2087 11:16


21/04/2114 07:40


09/09/2121 03:50


17/04/2160 20:09


Note that the close approaches will always take place either in mid-April or in early September when the orbit of Apophis crosses the Earth’s orbit.


There is an extremely close approach on April 13th 2029. The distance given is 34 677km from the centre of the Earth. That is 28 306km above the surface of the Earth. The orbit is extremely uncertain after this approach because of the perturbations that the Earth will cause in the orbit of Apophis, which depend critically on the exact distance at closest approach.



What is the fuss about?


The asteroid was discovered by Australian amateur, Gordon Garradd, on December 18th 2004 in Siding Spring Survey observations with the 0.5-m Uppsala Schmidt and placed on the Minor Planet Center’s NEO Confirmation Page (often abbreviated to “NEOCP” by astronomers). This object was evidently unusual because of its rapid rate of movement. The object was confirmed on December 19th by J.E. McGaha from Tucson and then, a few hours later, by Mount Saint John Observatory in New Zealand.


Up to now, everything was quite normal. It showed that the object was indeed passing very close to the Earth and would pass at 0.0964AU around 09:30UT on December 21st.


On December 20th the object was identified with a very faint object observed from Kitt Peak on June 19th and 20th 2004 and designated 2004 MN.


The orbit calculation, using observations on 5 nights over 6 months, showed that it was a low eccentricity, low inclination orbit with a period of just 0.89 years and demonstrated that this object was exceptionally interesting. It was also rather bright, with an absolute magnitude of 19.7, corresponding to a diameter that could be as much as 700-m if it were a dark object.


On December 23rd the JPL “Current Impact Risks” web page and the University of PisaNEODys” web page flagged this object as giving an impact solution for April 13th 2029. What is more, the risk level on the 0-10 Torino Scale was a record level of 4, with a 1-in-40 risk of impact and, with the estimated size and the rather high probability of impact, the event was close to hitting a risk level of 5, “threatening events”.


Only the fact that this news broke over Christmas and coincided with the catastrophic Indonesian earthquake and tsunami stopped this from being a major news story.


On December 27th the asteroid were found in images from Stewart Observatory taken on March 19th. These significantly increased the length of the arc of the orbit available to astronomers and completely eliminated the impact risk for 2029. The new orbit gave a predicted miss distance in 2029 of 65 000km; still unusually close for such a large object, but 10.1 times the radius of the Earth and so far enough away to represent no danger whatsoever.



Why then do we still worry?


Taking advantage of the fact that 2004 MN was still very close to the Earth in January 2005, the Arecibo radar in Puerto Rico observed the asteroid. It was detected on January 27th, 29th and 30th, giving a quite strong, clear radar return.


As the plot shows, the signal from the asteroid is clear and well above the noise.


The measurements showed that the asteroid was slightly closer than expected. The difference, 294km, seems tiny, but it meant that the 2029 approach would be to just 36 700km, or 5.7 Earth radii, little more than half of what had been predicted previously. In other works, a 1km error in the calculated position in 2005 gave a 100km error in the predicted position in 2029.


However, the approach in 2029 is so close that the asteroid’s orbit with be significantly changed by the Earth’s gravity. Effectively, the asteroid’s direction of movement in space will change by about 30 degrees after the encounter. This would lead to the asteroid passing very close again in 2036, with a calculated distance of 0.005AU (750 000km, or only about twice the distance of the Moon). This raised the possibility of an impact in 2036, with the asteroid possibly being deflected by its 2029 encounter into a potential collision course on April 13th 2036.



Correction to calculated position/velocity

Correction to distance in 2029

Miss distance 2029

Miss distance 2036

Feb 4th 2005


-28 000km

36 790km


Sep. 2nd 2005



37 370km


May 16th 2006



37 820km


Current MPC



34 680km



Note how the current orbit solution has Apophis not even making a close approach to the Earth in 2036: a tiny difference in the position in 2006 leads to a small difference in its position in 2029, which makes a huge difference to its position in 2036.



How reliable is the orbit?


The observations of the orbit cover 885 days, from March 15th 2004 to August 16th 2006. This is rather less than 3 orbits (exactly 2.73), so the error in the orbit would normally be relatively large.  Also included are 7 radar observations, which fix the orbit with an exceptional precision. However, even a very small error in position now has a huge influence in the actual position in 2036. To rule out an impact the orbit has to be determined to much higher precision still, something that should be achieved in the 2012 close approach.



Could Apophis hit the Earth in 2029?


No. Despite passing so close, it most certainly cannot. Even the largest possible error in position in 2029 will have the asteroid passing safely outside the Earth’s atmosphere.



What then is the danger?


Were Apophis to pass through a corridor just 200-m in diameter in 2029 – in other words, at precisely the right (or wrong) distance from the Earth to +/-100-m – the Earth’s gravity would deflect it by exactly the amount necessary to impact in 2036.


However, the uncertainly in the exact distance at which Apophis will pass by the Earth in 2029 is still about +/-1000km, a huge margin compared to the very tiny corridor that would lead to a later impact. For this reason the probability of an impact is currently estimated as an almost negligible 1 in 45 000 (it was 1 in 40 at Christmas 2004 and 1 in 5560 in October 2005, so the trend is also for the danger to recede with each new measurement of its position).


However, there is also the extremely remote possibility that Apophis could be deflected into a course for an impact in 2037; this is currently given as 1 in 12.3 million.



What does a 1 in 45 000 risk really mean?


A good way of thinking about it is to compare it with the odds of winning the Spanish National Christmas lottery. If you buy tickets for two numbers you have about the same odds of getting the winning number as Apophis does of hitting the Earth in 2036.



If Apophis has by far the largest risk of an impact is it rated as “Torino 0”?


The Torino Scale rates four basic classes of asteroid when assessing risks:


  • Very small bodies (under 40-m diameter)

The kinetic energy of impact is under 1Megatonne. These bodies are likely to break up in the atmosphere and not reach the Earth’s surface anyway. An airburst at high altitude is unlikely to have major consequences unless it occurred over a heavily populated area.


These bodies can never have a non-zero Torino threat value.


  • Small bodies (40-200-m diameter)


The kinetic energy of impact is up to 100Megatonnes. These bodies would cause significant local devastation, although posing no real threat to the Earth as a whole. The amount of damage caused by a 100Mt explosion depends critically on the altitude. A ground impact would produce a crater 1.8km across and total destruction over a diameter of 25km on the ground. An airburst at 10km though would increase the diameter of total destruction to 110km, with damage across a 300km diameter.


Such a body must have a minimum of 1 in 10 000 probability of impact to be rated Torino 1 (with 1 in 500 the minimum for a 20-m body). For a 1% chance of impact the Torino Scale value is 3 and, if greater than 99% (i.e. less than a 1% chance that it will miss) climbs to 8.


However, bodies smaller than 300-m in diameter are not classed as Potentially Hazardous Asteroids.


  • Medium-sized bodies (200-m to 1km diameter)


These are events capable of causing devastation over an area from the size of a country to that of a small continent, classed as “regional destruction”. A 1km asteroid would be capable of forming a crater 15-20km across or, if it were to explode in the atmosphere, to cause damage over an area about half the diameter of the United States.


As these events are far more threatening, even a 1 in 10 000 chance of impact is enough to elevate them to Torino 2 and a 1 in 100 probability of impact raises them to Torino 4 or 5, according to size. With a diameter that was estimated initially to be as large as 400-m and a 1 in 40 probability of impact, Apophis was rated as Torino 4.


  • Large bodies (more than 1km diameter)


These events are extremely rate, but the most devastating of all. They would have consequences on a global scale. For a 5km asteroid the energy of impact would be around 107Mt, the crater approximately 85km in diameter and around 1012 tonnes of mass would be excavated by the explosion.




Thus, for Apophis to have a non-zero value on the Torino Scale with its diameter of 250-m it would thus need to have at least a 1 in 10 000 probability of impact; it is some way short of this.


Another way of looking at what a 1 in 10 000 probability means is that if you try to HIT the Earth with the asteroid, throwing it at us 10 000 times, 9999 of those 10 000 you will miss.



What do we know about Apophis?


Radar observations have shown that Apophis has a much smaller diameter than was first thought – only 250-m – and is thus highly reflective (albedo 33%). This suggests that it is probably rocky rather than carbonaceous.


As Apophis has a low eccentricity orbit that crosses the Earth’s orbit from behind, the relative encounter velocity is low. Apophis passes the Earth at just 5.9km/s. Even with the acceleration caused by the Earth’s field of gravity its velocity at impact would only be 12.6km/s.



What are the potential consequences of an impact?


Given the known size and velocity of Apophis and its estimated density (2.6g/cm3), the kinetic energy of the asteroid is equivalent to 400Mt.


The consequences of an impact depend on whether it hits the ground or explodes in the air. If an asteroid hits the ground it will produce a crater, but much of the explosive force is used up in digging the crater, so it causes less damage overall. If the asteroid penetrates deep into the atmosphere and then explodes, there will be no crater, but the effects of the blast wave will be more serious. This was the case in the famous Tunguska impact in 1908. It left no crater, but its effects were far more devastating because it exploded at an altitude of some 8km.


Ground impact:


It would produce a crater approximately 2.9km in diameter and 140-m deep, expelling around 4000 million tonnes of rock and earth. The blast wave would destroy everything in a radius of 20km around the impact point. Some damage will be caused (windows blown out, trees uprooted, etc.) more than 50km from the impact point.


Explosion at 10km altitude:


No crater would de formed. There would be total devastation out to 90km from ground zero, with damage out to 250km from the centre of the explosion.



Should we worry?


Definitely no.


The danger of an impact in 2036 is so remote as to be almost negligible. However, the consequences of an impact are so grave that it would be foolish to ignore them completely, which is why scientists are watching this asteroid closely. You insure your house not because it WILL burn down, but because there is a remote chance that one day it MIGHT burn down; rarely though does it ever happen, which is why insurance companies make money. 


(Versión en castellano)


     - The asteroid impact risk reviewed. Has it been greatly overestimated.
        Mark Kidger - INSA y Herschel Science Centre. European Space Astronomy Centre. European Space Agency - Villafranca del Castillo
        Satellite Tracking Station (Madrid)