Bill Gordon, and the Arecibo Observatory. Credit: NAIC-ARECIBO OBSERV., FACILITY OF NSF

Bill Gordon, who died on 16 February at the age of 92, was a pioneer in studies of the ionized part of the upper atmosphere (the ionosphere, above about 50 kilometres). He conceived of and built a gigantic steerable radar telescope with a diameter of 300 metres, nestled among the mountains of Puerto Rico. Although this instrument was originally designed to study the ionosphere, today it is also used for radar astronomy, exploring objects as distant as a moon of Saturn and as near as potentially dangerous small asteroids. The antenna also serves as a huge telescope for radio astronomy, which, in contrast to radar astronomy, relies on radio waves emitted rather than reflected by an object. It is unsurpassed in the study of pulsars; one such study led to the discovery of the first planets outside the Solar System, and the authors of another won the 1993 Nobel Prize in Physics.

In the late 1950s, Gordon was a professor of electrical engineering at Cornell University in Ithaca, New York, working with Henry Booker on developing radio scatter communication links, which were used commercially in the Caribbean islands and by NATO in Europe. In 1958, Gordon came up with the idea of using radio signals backscattered from individual electrons in the ionosphere to study the 'weather' above 50 km — that is, space weather. This scatter is exceedingly weak, so studying it is rather like trying to detect a one-millimetre sphere at a range of 300 km. But Gordon showed that this might actually be done with existing 1958 radar technology using a 300-metre-diameter antenna and a megawatt or two of pulsed transmitter power. From the power and shape of the Doppler spectrum of this incoherent scatter, as it is now called, most of the properties of the ionosphere to altitudes beyond 1,000 km can be determined.

Gordon had his seminal idea at a propitious time. In 1958 the cold war was at its height, and surplus military radar equipment was available for the asking, along with military funding for interesting ideas — even unclassified ones. In a remarkably short time, Gordon raised US$10 million or so (in 1960s dollars, remember), designed the radar with colleagues at Cornell and in industry, and moved to Puerto Rico for about five years with his family and four other families from Cornell to supervise the radar's construction.

How do you build a 300-metre steerable radar? Steering the whole antenna is obviously impractical. Fixing the antenna's reflector and moving its feed is the only alternative, but then the reflector must be spherical, not parabolic, and with a line, not a point, focus. Designing a good feed is surprisingly difficult. But first, where do you locate the radar? You want it to be as far south as possible (so that it can observe planets as well as the ionosphere), and in a politically stable location. The karst region of Puerto Rico has ideal topography, and a natural sinkhole of about the right shape was found near the city of Arecibo. Somehow, all the necessary steel, concrete and heavy equipment were transported over the narrow, winding, rural roads to the site of the observatory. This remarkable radar was dedicated in 1963, shortly after making its first observations. It is now a US national facility operated by Cornell.

At the fortieth anniversary of this dedication, the Institute of Electrical and Electronic Engineers declared the Arecibo Observatory to be a “Milestone in Electrical Engineering and Computing”, and the American Society of Mechanical Engineers declared it to be a “Mechanical Engineering Landmark”. This double award made the observatory one of the top ten engineering achievements of all time.

At the ceremony, Gordon remarked: “When we were talking about building the radar back in the late '50s, we were told by eminent authorities it couldn't be done. We were in the position of trying to do something that was impossible, and it took a lot of guts and we were young enough that we didn't know we couldn't do it.” He also advised the audience: “If you dream, have big dreams, [and] talented supporters.” And to graduate students he added: “May each of you experience the passion of creation as you discover something new as you do your doctoral research, and may you experience the same deep passion a few more times in your professional careers.”

On a lighter note, the Arecibo Observatory featured in the 1995 James Bond movie Goldeneye and in the 1997 movie Contact, starring Jodie Foster, which is about the search for extraterrestrial intelligence, an activity carried out by the observatory.

In 1966, Gordon moved to Rice University in Houston, Texas, where he served in various capacities, including as professor of space sciences and electrical engineering, provost and vice-president. He retired in 1986 as a Rice distinguished professor emeritus, one of only two professors to be so honoured. While at Rice he remained active in Arecibo research, especially in a programme involving heating the ionosphere with high-power radio waves.

Bill received many scientific awards, and was a man of great humanity. He was superb at picking a good team and leading it effectively, as he showed in building Arecibo. He was widely sought after to chair important committees. Once, when a rather fractious group with strong opinions arrived for their first meeting, arguing heatedly, Bill, as chairman, began the meeting with: “Let's make a list of those things on which we can all agree.” As the list grew, the outline of the report of a committee that had reached consensus emerged. Another Gordon committee rescued a Peruvian incoherent-scatter observatory when its funding was threatened.

Bill enjoyed 61 years of marriage to his first wife Elva, and is survived by their two children, four grandchildren and three great grandsons. After Elva's death, he married his long-time friend Elizabeth Bolgiano, who survives him. At a dinner with Liz and Bill in February 2009, Bill remarked to me: “Remember that every day is Valentine's Day.” That was Bill. He will be missed.