The mirrors on NASA's $108 million FUSE telescope won't stay put, and that is delaying astronomers' plans to study chemical clues to the origin, evolution and fate of the universe.
Two of the four mirrors aboard the Hopkins-built Far Ultraviolet Spectroscopic Explorer satellite keep drifting out of alignment, apparently in response to temperature changes, mission leaders say.
It's only a tiny movement; the starlight reflected by the mirror drifts by barely the width of a human hair.
But it is forcing ground controllers to stop and re-align the mirrors for each new observation, and it has postponed work on a final focusing of the telescope.
Astronomers have been confined to looking mostly at faint objects outside our galaxy, delaying some of FUSE's most important observations.
"We are concerned, but not terrified," said Hopkins physics professor Warren Moos, principal investigator on the project. "We see potential solutions."
Moos was quick to wave off any comparison between FUSE's mirror troubles and those of the Hubble Space Telescope.
Hubble was launched in 1990 with an embarrassing and costly manufacturing flaw in its mirror. Hubble's view was blurry until corrective hardware was installed in 1993.
"We're talking about alignment," Moos said of FUSE. "We have seen no evidence of a focus problem. We haven't attempted to focus it yet."
The mirror troubles are the latest in a series of problems that have kept the FUSE team at Hopkins working overtime since the satellite's launch in June.
Most have been resolved, but the three-year mission is now two months behind schedule.
"I underestimated how long it was going to take to get this job done," said Moos.
He compared the orbital checkout process to sausage-making: "Those who love space science instruments should never watch during checkout."
FUSE scientists plan to measure the density of intergalactic gases, seeking clues to whether the universe holds enough matter and mass to eventually brake or reverse its expansion.
They also hope to trace the circulatory systems that disperse new elements throughout the galaxy -- elements such as carbon, iron and silicon that are forged in stellar explosions, and from which planets and life have evolved.
And despite its problems, FUSE's preliminary findings are being described by its science team as "tremendously exciting."
FUSE is already far more powerful and sensitive than earlier ultraviolet observatories, Moos said, and early observations have yielded more than 25 scientific papers. They will be presented at the American Astronomical Society's January meeting in Atlanta.
Blair D. Savage, an astronomer at the University of Wisconsin-Madison, said, "FUSE has already produced more high-resolution spectra in the ultraviolet of extra-galactic sources than Hubble Space Telescope has done over its entire life."
University of Colorado astrophysicist J. Michael Shull, also a FUSE team member, agreed that FUSE is doing publishable science.
"But these are not the projects that require the highest resolution," he said. "I would be very disappointed in six months if we didn't have considerably better resolution than we have now."
FUSE needs all its mirrors aligned and focused to nail its No. 1 objective -- measuring the amount of deuterium in the universe.
Deuterium is a form of hydrogen created in the early universe. Scientists believe that if they can measure the relative amounts of deuterium, hydrogen and helium out there, they can infer the conditions present in the early universe, and find clues to its ultimate fate.
Savage, a member of the FUSE science team, said its start-up problems are not unique, and they might be a consequence of NASA's "better, faster, cheaper" approach to space science.
"There's less time available to fully test these devices on the ground before they're launched," he said. "So in almost every situation in-orbit checkout is going to take longer."
"At least we have a spacecraft that works," said Shull, an allusion to NASA's $125 million Mars Climate Orbiter -- another "better, faster, cheaper" mission that crashed on Mars last month because someone failed to convert guidance commands from English to metric measure.
Downsized and academic
FUSE was downsized from an earlier, $300 million plan and completed in just three years. It is the first NASA mission if its kind to be designed, built and operated by a university academic department. The FUSE control room is in the Bloomburg Center for Physics and Astronomy on Hopkins' Homewood campus.
FUSE does not take pictures. It is a spectroscopic observatory, designed to split the ultraviolet portion of starlight into its constituent wavelengths, or "spectra."
Analyses of the spectra can reveal the chemical and physical conditions in stars or galaxies, or in the dust and gas between them.
But getting started "has been a struggle," Moos said.
Within weeks of its launch June 24, controllers noticed that data in memory chips in FUSE's ultraviolet detectors were being corrupted every few days by high-energy radiation in Earth's radiation belts.
"It's a ferocious radiation environment up there," said Moos, especially over the South Atlantic Ocean, where the radiation belts dip to the altitude of FUSE's orbit, 477 miles up.
The problem is compounded by rising solar activity as the sun nears a peak in its 11-year cycle.
The bad data can be rewritten, but only during the satellite's infrequent, 13-minute passes over its ground antenna in Puerto Rico. Sometimes, the detectors have to be shut down and restarted, a laborious, daylong process.
Not the best chip
In retrospect, Moos said, "I think the choice of chip was not the best." Tougher, "radiation-hardened" chips could shield memory data for decades.
But deputy project scientist Kenneth Sembach said FUSE engineers "chose the best they could given the costs and the time constraints we were under."
By mid-November, Moos said, new software will enable the spacecraft to make the memory fixes on its own.
The repair efforts have been complicated by a series of antenna failures in Puerto Rico. The glitches have disrupted Hopkins' already limited access to its satellite and forced controllers to buy time on a commercial antenna in Hawaii.
The antenna is working better now. But George Sonneborn, NASA's FUSE project scientist, said engineers still "don't really understand why it's fixed."
A "tiger team" of outside experts has been assembled by NASA's Goddard Space Flight Center to study the problem.
Amid all these other headaches, FUSE controllers have also had to trouble-shoot and rewrite computer commands for the spacecraft's guidance system, which had trouble finding its guide stars, and to clean up a cluttered onboard memory disk.
But the mission's big problem is its mirrors.
"We don't fully understand what's going on," Moos said. A painstaking and systematic effort to analyze the trouble is expected to take another month. The fix could take several months more.
Early data suggest that two of the four mirrors -- the ones on the sunward side of the satellite -- drift out of alignment with respect to the other two.
The movement is tiny -- a fraction of the width of a human hair. But it's enough that starlight reflected off the mirrors misses the pinholes it must pass through to reach the ultraviolet detectors.
Moos said the drift seems to occur when the angle of sunlight shining on the telescope changes. So, whenever FUSE is aimed at a new target, the mirrors move out of alignment. FUSE's orbit through "day" and "night" every 100 minutes may complicate things further.
Realigning the mirrors for each new observation eats up valuable telescope time, Moos said.
"You can do it," he said. "You just don't do it on a routine basis and go from target to target easily."
FUSE's designers knew the telescope would face a harsh thermal environment, Sonneborn said. But "it's extremely difficult to simulate the dynamic environment FUSE would be in. You would have to put it on a spit."
A fix could involve reprogramming onboard heaters. Or observations might have to be scheduled so that the telescope's movements are kept very small.
"The important thing," said Savage, "is that there are no known problems that are insurmountable, as far as we can tell right now."