However, scientists have created a new model for predicting the impact of such collisions that shows the devastation wouldn’t be anywhere near as severe as previously thought.
Scientists from the University of Munich argue that previous predictions of impact damage assume the Earth is smooth and spherical – but it’s actually elliptical with huge peaks and troughs, which would help to diffuse the resulting seismic waves.
Research leader Matthias Meschede said: ‘After a meteorite impact, seismic waves travel outward across the Earth's surface like after a stone is thrown in water.
'For the Earth, these calculations are usually made using a smooth, perfect sphere model, but we found that the surface features of a planet or a moon have a huge effect on the aftershock a large meteorite will have, so it's extremely important to take those into account.’
Mr Meschede’s team used their model to produce a fresh simulation of the meteorite crash that wiped out the dinosaurs 65million years ago and caused the huge Chicxulub crater, which can be seen from space.
The space rock was the mass of Mountt Everest – when its tip hit Mexico the tail was still 35,000 feet up.
It was the equivalent of two million hydrogen bombs going off – yet the Munich team have downplayed its effects, claiming that the tsunamis and earthquakes that resulted were enough to wipe out the dinosaurs, but the shock-waves' that caused them would have been weakened by the planet’s landscape.
Mr Meschede added: 'Our results go beyond Chicxulub. We can, in principle, now estimate how large a meteorite would have to have been to cause catastrophic events. Our model can be used to estimate the magnitude and effect of other major impacts in Earth's past.'
Tomorrow an asteroid the size of four football pitches called 2005 YU55 will pass between the Earth and Moon.
If something that size was to slam into the Earth it would cause a 2.5mile-wide crater, a magnitude 7.0 earthquake, and if it hit water a 70-foot high tsunami would spread within 60 miles of the crash site.