Full marks for effort ...

If he had been proved correct, 13-year-old Nico Marquardt might have embarrassed some of Nasa's finest scientists. The schoolboy astronomer thought the space agency had missed something when calculating whether or not an asteroid it is tracking is likely to hit the Earth in 2036. His own calculation suggested that a collision is hundreds of times more likely than Nasa thinks.

But anyone concerned about an imminent impact should rest easy. Scientists welcomed the German teenager's enthusiasm but have pointed to a number of errors in his work. And reports that Nasa has put its hands up and admitted errors (repeated all over the web today) seem wildly exaggerated.

According to the Potsdamer Neuerste Nachrichten newspaper, Marquardt used data from the Institute of Astrophysics in Potsdam to work out the future orbit of Apophis, a 270m-wide asteroid that will pass close to the Earth in 2029 and again in 2036.

Alarmingly high odds

It's true that when astronomers first noticed the asteroid in 2004, they calculated alarmingly high odds of it hitting the Earth in 2029. At the peak of concern, Apophis scored four out of 10 on the Torino scale - a measure of the threat posed by near-Earth objects, where 10 is a certain collision that could cause a global catastrophe. A score of four was the highest of any asteroid in recorded history and Apophis was given a 1 in 37 chance of hitting the Earth.

Estimates of an impact suggested that it would release more than 100,000 times the energy released in the nuclear blast over Hiroshima. Thousands of square kilometres would be directly affected by the blast, but the whole of the Earth would see the effects of the dust released into the atmosphere.

A collision in 2029 was eventually ruled out but there remained a small possibility that, if the asteroid passed through a specific 600m patch of space that year called the keyhole, the Earth's gravity would change the asteroid's orbit so that when it came back around again in 2036, it would collide with us.

Altered trajectory

Nasa's current estimate is that Apophis has a 1 in 45,000 chance of hitting the Earth in 2036. Marquardt reckoned that this should be raised to 1 in 450. He re-calculated the odds by taking into account whether the asteroid would hit one of the 40,000 artificial satellites in orbit around the Earth during its approach in 2029, which he claimed would change its trajectory and increase its chances of hitting the Earth in 2036.

The Potsdamer Neuerster Nachrichten said that satellites travel around 35,000km above Earth, and it is thought that Apophis will get to within 32,500km of the surface.

Alan Fitzsimmons, an astronomer at Queen's University Belfast said that while he applauded Marquardt's understanding of orbital mechanics, there were some questions over his conclusions. "Most satellites orbit in low-Earth orbit at altitudes of 300km-2000km. These are in no danger from the asteroid. Many of the rest are in geostationary orbits at 42,100km from the centre of the Earth. This is much further out than the pass distance of the asteroid in 2029. Overall the risk to satellites is minuscule."

In addition, to be deflected by gravity into the path of the Earth for a collision in 2036, the asteroid still needs to pass through the keyhole. "It either will or it won't. Hitting a satellite will not deflect it into that point."

Brightest and best minds

A spokesperson for Nasa told the Guardian that, though no one at the agency had seen Marquardt's calculations, scientists there were confident of their figures. "We have some of the brightest and best minds to calculate these things, they make careers out of this - I don't see how they could make a big mistake like that."

Monica Grady, a planetary scientist at the Open University in the UK said that calculating the orbits of asteroids way into the future was notoriously difficult and Nasa's own calculations already had huge margins of error. Since motion is governed by gravity, even a slight error in the size or motion of local planets can multiply into a wide range of possibilities, and Marquardt's calculations were within the error bars already calculated by scientists.

In a paper on Apophis published last year in the journal Icarus, Nasa planetary scientists considered the risk of the asteroid hitting an orbiting satellite, but they decided it was not significant. "What they say in the paper is the errors are still so big in the assumptions we make on how we're going to calculate its orbit that we won't be able to get any better data until 2011 to actually be able to say whether it's going to hit a satellite or hit the Earth," said Grady.

She isn't worried by Marquardt's calculations. "Not until I see some more evidence. It's really interesting that it should come to light in this particular way and it shows there is always room for people to check up on things and it's very valuable that they do. But it's not something that I'll read and think, oh crikey, I better start laying in the baked beans."

Fitzsimmons added: "I'm not any more worried about this particular asteroid than I was yesterday."