Y2K-like glitch thought responsible for Deep Impact's demise

Once a week, scientists at the University of Maryland, College Park and NASA aimed satellites across the solar system to where they figured the spacecraft known as Deep Impact would have drifted since seven days earlier.

They would ping it for its observations of passing comets or of planets it spied outside our solar system, and it would send the data back.


But on Aug. 14, there was no response. Scientists tinkered with the direction of their satellites, wondering if Deep Impact had been knocked off its orbit or was just hindered by a temporary glitch, but still, their messages went unanswered for weeks. Earlier this month, they gave up, announcing that they had lost contact with the spacecraft.

They suspect that the problem lies in a software glitch similar to what caused the "Y2K" panic in 1999, when a system is built to accept only limited parameters of time.


It was an unexpected end to a project that had already surpassed its original six-month mission by eight years and yet was still primed to increase understanding of the formation of the universe by watching another comet hurtle past the sun and Earth. Deep Impact was a source of pride and renown for the state's flagship university, and scientists working on the mission say there is still much that can be learned from its findings.

"I would have expected some piece of hardware to fail first," said Michael A'Hearn, a university astronomer and principal investigator on the Deep Impact mission. "It's a great disappointment is all I can say."

The disappointment is perhaps a sign of how far Deep Impact had already exceeded expectations.

The spacecraft launched from Cape Canaveral, Fla., on Jan. 12, 2005, traveling 268 million miles toward a comet known as Tempel I. Two years earlier, cost overruns had threatened to kill the $330 million project.

Six months after the launch, 650 people gathered on Independence Day at the Toll Physics Building on the College Park campus as the spacecraft slammed an 820-pound copper impactor into the comet at 23,000 mph.

"We had grown men in here crying because they had been thinking of this moment for so long. I certainly cried," Lucy McFadden, a university astronomer and a member of NASA's science team, told The Baltimore Sun in 2005.

Data gathered from the excavation showed that the comet was more dusty and fragile than they had expected, with small crystals of ice instead of one large chunk of it. Scientists want to learn about comets because they are thought to be the leftovers from the formation of stars and planets and could give an idea of conditions in the early universe.

The success of the mission was "ammunition" for a second try, said Lori Feaga, an assistant research scientist on the project. When NASA solicited ideas for how to use the spacecraft once Tempel I had passed, the Deep Impact team's plan to get a closer look at a comet known as Boethin was the winner.

When Boethin ended up being too small, the mission shifted to a comet known as Hartley 2, instead, observing it (but not sending an impactor into it) in 2010 from within 600 miles. Deep Impact has since observed other celestial objects of interest while awaiting another mission, including Comet ISON, a comet looping in from outside the solar system and on track to graze the sun in November.

Deep Impact also took pictures of Earth, the moon and Mars. These findings helped confirm the existence of ice on the moon and suggested the presence of methane on Mars, though the Curiosity rover failed to detect it there.

NASA was in the midst of a process to establish a new mission for the spacecraft, and it would have been trained on ISON in the meantime, but the efforts have been abandoned.

"We've been fighting all the way, whenever we had these comets of opportunity," Feaga said.


An investigation into why the spacecraft stopped responding revealed a problematic line of code in, ironically, the software that allows systems to be placed on standby if something unusual is detected, the scientists said.

When the spacecraft sends observations, each is stamped by date and time; the scientists said they believe that when it ran out of possible time stamps, the software sent the spacecraft into a reboot loop that they couldn't stop. It may still be in the loop, or its battery may have died; they don't know.

A'Hearn said he has long been preparing for the mission to end, given the amount of things that could have gone wrong before an extended mission might have been pursued. Still, the team was confident that Deep Impact had enough fuel for an extended mission, he said.

They will continue to learn from the mission for a time as they sift through observations they haven't scrutinized yet. The observations could still drive research projects, prompting proposals for other missions using telescopes to confirm or invalidate hypotheses.

"The mission will live on in the data and the projects we can still do; we just won't be receiving any new data," Feaga said.

Tribune Newspapers reporter Amina Khan contributed to this report.


Recommended on Baltimore Sun