Seiichi's major research field is radio astronomy and interstellar physics. He participated in the planning and construction of the Atacama Large Millimeter/submillimeter Array (ALMA). Seiichi currently works for JAXA as an astronomer and Director for Space Science Outreach at the Institute of Space and Astronautical Science (ISAS).

Professional observatories are our windows to the Universe, but where to build these colossal structures is a science in itself. In this feature article, Seiichi Sakamoto talks through the process of site selection, using the high-tech Atacama Large Millimeter/submillimeter Array as a case study.

How sites for observatories are selected

Terrestrial atmosphere absorbs, scatters, and scintillates the signals from celestial bodies. Space will be an ideal site for observations if budgetary and technical constraints are cleared. Practically speaking, however, instruments much heavier than a few tons have so far been difficult to launch, and so the roles of ground-based facilities are vitally important. There is much literature on the importance of the scientific instruments; I will focus on the practical selection procedures of the site where the instruments are to be installed.

Astronomers love dry, high sites far away from cities

Conditions to be a "good" site for astronomical observations depend on what you would like to observe. In general, a site for an astronomical project should be good -- at least not critically bad -- in terms of atmosphere and geography, as well as other non-technical conditions. The atmospheric factors include atmospheric transparency, seeing, meteorological conditions (e.g. surface wind, snowfall, near-surface temperature, lightning safety), and sky brightness/interference. Geographical and geological conditions include local topography (i.e., slope, roughness), seismicity, mechanical/electric/thermal properties of rocks/soils, and source availability (i.e. latitude and skyline).

Non-technical issues such as existing infrastructure, accessibility, status of the host country (e.g. political status, environmental issues, satisfactory agreement, and labour level), construction cost, and ease of getting funds sometimes play major roles in determination of the site. The scores and relative weighting of these conditions depend on the scientific objectives, specifications, and cost of the instrument, as well as who is to promote and who is to fund. Even more complicated, they are often time variable and need to be modified within the boundaries of given budget, schedule, and manpower of the project.

The case of ALMA site testing

For optical telescopes, optical seeing and the fraction of photometric nights are of concern, and there are many sites; e.g. Hawaii, northern Chile, Canary Islands, and South Africa. For submillimeter-wave and mid-infrared observations, in which wavelengths the atmospheric absorption is critical, astronomers prefer even drier and higher sites compared to optical observers. I will describe the selection of the site for ALMA, the Atacama Large Millimeter/submillimeter Array.

ALMA is an international project organised as a merger of three progenitors: the Japanese LMSA (Large Millimeter and Submillimeter Array), the American MMA (Millimeter Array), and the European LSA (Large Southern Array). All these required a site with high transparency and seeing in millimeter and submillimeter wavelengths, and with flat and wide areas for installation of large arrays.

In Japan, soon after the completion of the Nobeyama Millimeter Array (NMA) in the 1980s, radio astronomers started to think about an even larger array. The first concept was to fill up all 30 stations of NMA with element antennas for short-term success, but the majority preferred going abroad for better sites. Our site survey started at this time.

Initial surveys and pinpointing

After examination of archived precipitation and topographic data, we picked up dry and high sites in Hawaii, northern Chile, western China, and northern India as the candidate sites that meet the demands for the large millimeter and submillimeter arrays. The Antarctic plateau was discarded because of poor stability of the icy surface. We organised a caravan of measurements with weather stations and radiometers, and surveyed several tens of the sites as a joint effort among the Nobeyama Radio Observatory (NRO), National Radio Astronomy Observatory (NRAO), and European Southern Observatory (ESO).

Through this joint venture, we learned that northern Chile has superb conditions. We then concentrated our efforts to pinpoint the site with the same instruments. We installed solar panels and containers as a refuge, and then weather stations, radiometers, and radio seeing monitors for statistical characterisation of the sites. After taking measurements over a few years, we made a quantitative comparison of candidate sites and finally selected a 5000 m altitude area near Cerro Chascon in northern Chile.

Site characterisation for design works

Once the site is fixed, we need to learn in more detail about the atmosphere above the site and the terrain for practical design works for instruments and infrastructure. The items we monitored include atmospheric temperature, humidity, wind speed and direction, wind gust, atmospheric transparency and seeing in radio waves, solar radiation in visible and UV rays, cosmic ray intensity, lightning, and soil resistivity. We also launched radiosondes from the high site. Virtually all of the initial measurements were carried out by astronomers and observatory staff.

Site remoteness, reduced oxygen, and safety

During the site testing, safety was my greatest concern. We usually drive four-wheel drive trucks to go off road. Additional spare tires, extra fuel tanks, and satellite communication system were necessary during the early survey stage. For safety at the remote and high site, the joint site testing team established the "two-car rule" and "two-person rule", quite similar to the buddy system applied by scuba divers. In addition, each car is equipped with a walky-talky to ensure communication.

Work at very high (> 4000 m) altitude is challenging. The brain consumes plenty of oxygen while there is only a half of what we are used to. We wear pulse oxymeters to continuously monitor our arterial oxygen saturation. I determined a work hour limit at the high site -- four hours for the fist day, six hours for the second, and eight hours from the third day on.

Now the site is almost ready

In November 2003, a ground breaking for site construction was held at the site. A few years later, I left ALMA to start my new career at JAXA, and I hear that the site is now almost ready for installation of the new instruments. ALMA will be fully operational in 2012, and the first call-for-proposals is expected soon. I hope I will be among those lucky enough to use it!

This feature article was written as part of the Cosmic Diary Cornerstone project for the International Year of Astronomy 2009. To find out more, check out www.cosmicdiary.org and www.astronomy2009.org.