At about 3am on 8 October last year, an asteroid the size of a small house smashed into the Earth’s atmosphere over an isolated part of Indonesia. The asteroid disintegrated in the atmosphere causing a 50 kiloton explosion, about four times the size of the atomic bomb used to destroy Hiroshima. The blast was picked up by several infrasound stations used by the Comprehensive Nuclear-Test-Ban Treaty Organization to monitor nuclear tests.

No-one was injured in blast but the incident highlights the threat that planet faces from near Earth asteroids. Astronomers expect a strike like this once every 2-12 years. And the US congress has given NASA the task of sweeping the skies to identify anything heading our way. So far NASA has looked for objects of a kilometre or more in size and determined that none of these is on track to hit Earth in the foreseeable future.

But what of smaller objects? Various estimates show that an impact with an asteroid just 50 metres across would cause some 30,000 deaths (compared with 50 million deaths from an impact with a 1 kilometre-sized object).

This raises two important questions. The first is how best can astronomers monitor the skies for these smaller objects. The second is what to do should we find something heading our way.

Today, John Tonry from the University of Hawaii gives us his view. He believes the goal of finding 90 per cent of city-destroying asteroids in time to deflect them is extremely challenging. He says a more realistic target is to evacuate the area under threat. And for that we’ll need just three week’s notice.

Given these constraints, he reckons he can do the job with an array of eight 25cm aperture telescopes with a wide field of view that simultaneously scan the visible sky twice a night. He calls this early warning system the Asteroid Terrestrial-impact Last Alert System or ATLAS.

This is certainly a cost effective solution. The telescopes should be easy and cheap to construct. “We believe that a fully equipped telescope with focusser, fifilters, shutter, camera adapter and mount should cost about $50k,” says Tonry (although who he means by ‘we’ isn’t clear). There’d be an additional cost of $50k for the CCDs plus operational costs.

That’s chickenfeed compared to the damage an asteroid could do. So it certainly looks as if the proposal is affordable. It also offers a useful additional benefit: the ability to monitor other rapidly changing events in the heavens, such as supernovas and gravitational lensing events.

A more difficult question to answer is whether three weeks is long enough to evacuate a city that may be home to several million people. It may be only a matter of time before we are forced to try.

Ref: arxiv.org/abs/1011.1028: An Early Warning System for Asteroid Impact