For the fifth time in two years, a report of an Earth-threatening asteroid was proven wrong within days of being announced.

The asteroid was stated to be on a possible collision course with the Earth with the impact date set for 2022. It rapidly became clear that the asteroid would miss the planet by millions of kilometres.

Some scientists fear the public may become desensitised to the warnings.

"Someday, we're going to find something that will have a 1 in 1,000 or 1 in 100 chance of impacting Earth," said James Scotti, who discovered the asteroid last month at Kitt Peak National Observatory. "When that happens, I'd rather us be taken seriously."

One in a million

Dr Scotti did not know about the celestial rock's possible trajectory until Monday, when Italian researcher Andrea Milani posted an internet message warning of a 1 in a million chance of a collision and asking other astronomers to track it carefully.

A day later, Dr Milani announced that the new observations allowed him to make more precise calculations. The asteroid, named 2000 BF19, would come no closer than nine million kilometres (5.6 million miles) to Earth over the next 50 years, he said.

"This change is the result of computation I did today from the response of my call to arms yesterday," Dr Milani said. He added that it had taken about four hours to compute the course using the new observations from around the world.

The object also was being followed at the Near Earth Object Program at the Nasa's Jet Propulsion Laboratory, but it quickly became apparent that the 800-metre- (half-mile-) wide rock posed no real threat. The laboratory has never issued an asteroid collision warning.

"In almost all of the five cases, we're the ones who came back and said it won't happen," said the program's manager, Donald Yeomans. "We're the nay-sayers."