"The initial response of many students when they hear I'm going to Hawaii to make observations is, 'Wow! That's cool!'" laughs Proty Jiun-huei Wu, an associate professor in the Department of Physics at National Taiwan University (NTU) and a scientist on the AMiBA project. "But you should see the pained expressions when I ask those who've been with me before to come again."

Researchers must drink water regularly to avoid becoming dehydrated in the extremely dry air at the summit.

Issue 1: A vampire's schedule

Most of the AMiBA team's observational work takes place at night, because clouds and moisture are more likely to interfere with reception during daylight hours. At nightfall, the inversion layer drops to about 2,800 meters, where it is less likely to affect their research.

The sun itself is another reason why they work at night. The telescope's precision electronics are just as easily damaged as a naked eye when inadvertently pointed at the sun. The sun's radiation also affects observations.

"There's an 18-hour time difference between Hawaii and Taiwan," says Wu. "Since we work at night, we don't bother adjusting our schedules--we just stay on Taiwan time." To ensure that they're alert during their work hours, they hide indoors with the curtains drawn during the day, sleeping while most people are up and about. "The sun and sand, the swimming and surfing.... We don't do any of the fun things you're supposed to do in Hawaii," laments Wu.

The SMA, located near the summit of the 4,205-meter Mauna Kea, sits well above cloud and water-vapor interference, making it a destination of choice for international astronomers. The volcano's summit offers a majestic view of the seemingly endless sea of clouds below.

Issue 2: Unstable systems, noise

Astronomical observations are different from those in other scientific fields in that scientists can't just bring the object of their research into the controlled environment of a laboratory for analysis. Instead, they must use telescopes to make remote observations, recording whatever images or signals they can get.

"Clouds in the atmosphere, signals from other stellar bodies, or even simply instabilities in the telescope's systems can all generate noise that interferes with observations," explains Wu. "The moon, for example, reflects solar energy as a strong microwave signal. When making observations, we have to steer clear of the moon to prevent its noise from corrupting our data."

In addition, AMiBA is a new telescope, and not all of the bugs have been worked out of its systems. These occasional glitches are a headache for the team's scientists. "But you can encounter noise from system instabilities on almost any telescope," says Umetsu Keiichi, an assistant research fellow at Academia Sinica's Institute for Astronomy and Astrophysics (ASIAA) and member of the AMiBA team. "When it occurs, you just have to be patient, track down the cause of the problem, and continue with your observations."

The SMA research center on Mauna Kea is very well equipped. It has devices that add oxygen and pressure to the indoor air, as well as a small kitchen for the scientists who are making longer-term observations. The dishes that make up the array and the magnificent mountain-top scenery can be seen through the window in the research area. Life here is a far cry from the spartan conditions at the AMiBA base station on Mauna Loa.

Issue 3: Cold, strong winds, altitude

At elevations of nearly 4,000 meters, daytime and nighttime temperatures can vary by well over 10°C and the winds often blow like a typhoon. Then there's the thin, low-pressure atmosphere. These are all serious issues for astronomers not used to living at altitude.

AMiBA project manager and ASIAA research fellow Chen Ming-tang says that lack of oxygen results in headaches, nausea, a racing heart, and difficulty concentrating, all of which are symptoms of altitude sickness. Researchers visiting for the first time must usually spend some time adapting.

Even those who go up the mountain every day never become completely accustomed to the work environment. "It's really cold here at night, maybe only about 0°C," says AMiBA engineer Johnson Han. "Plus the winds are strong. You hole up in the research offices with a heater, but it's terribly dry. You usually end up chugging down water and running to the bathroom all night long."

Hawaii may be one of the world's top tourist destinations, but the Taiwanese astronomers are not here to play. They work all night and sleep the days away, never enjoying the sun, sand, and water sports for which the islands are known.

Issue 4: Lack of amenities

The AMiBA base was built only six months ago, and the facilities are still pretty rudimentary. Life is very different at the SMA observation station. The Mauna Kea location was developed many years ago and features far more amenities. There is even a dormitory and restaurant at 2,800 meters jointly built by international research groups working at Mauna Kea's many telescopes. In contrast, when the AMiBA team leaves their control room and offices, the only place they can go to rest or relax is a converted shipping container that's been outfitted with simple beds. They don't even really have a kitchen, just a small microwave oven.

This lack of amenities, coupled with the difficulty of getting restful sleep at altitude, means that the team usually drives up the mountain at sunset, makes observations all night, then drives back down at dawn to get some sleep. "The road up the mountain to AMiBA is very rough and winds around a lot," says Umetsu Keiichi. "There've been small accidents on the way down. With that in mind, we always go up the mountain in groups of at least two, so that if something were to happen, there'd be someone there to help."

While sitting on the top of a Hawaiian volcano conversing with stars hundreds of millions of light years away may sound like a romantic pastime, the reality can be harsh. But, as grueling as it can be, these scientists believe that finding an answer to one of their questions in the vastness of the sky makes it all worthwhile.