Magellan

Magellan carried one instrument to Venus: a radar that used the spacecraft's 3.7 m communications dish as its antenna. Between September 1990 and September 1992 it imaged 98 percent of the surface through the permanent cloud, according to the Jet Propulsion Laboratory's closing status report, and two further years of radio tracking produced a gravity map covering 95 percent of the planet. Its archive was still yielding new studies in 2026.

The mission grew out of a bigger one. In the early 1970s NASA's Ames Research Center and JPL studied Venus radar orbiters separately, and Louis Friedman and Al Laderman named JPL's project the Venus Orbiting Imaging Radar, VOIR, after the French verb voir, to see, according to Andrew Butrica's NASA history of planetary radar, To See the Unseen (1996). By 1978 its plan drew on parts from Mariner 10, Voyager and Pioneer Venus, with a radar derived from the Seasat satellite's. R. Stephen Saunders, a JPL geologist, was principal study scientist of the VOIR group and later Magellan's project scientist.

VOIR never flew. Congress had already voted it a new start when Ronald Reagan took office in January 1981, Butrica writes, but budget director David Stockman pressed NASA to give up a major project to signal spending cuts, and NASA chose VOIR, priced at 680 million dollars. One risk JPL had listed in 1978 for a late launch, that the Soviet Union would get radar images of Venus first, then came true: Venera 15 and 16 reached the planet in October 1983 and mapped its northern quarter by radar.

Its replacement was cheaper by design. Gordon Pettengill of MIT, later the radar's principal investigator, told Butrica that staff were promised their mission if they could cut its cost in half. The Venus Radar Mapper, VRM, aimed at 200 to 300 million dollars, dropped the four experiments that did not use the radar and became a NASA new start in 1984. An elliptical orbit spared it the aerobraking VOIR's circular orbit would have needed after arrival. In December 1985 NASA Headquarters told JPL the project would be called Magellan, after the Portuguese navigator (Butrica); JPL's 1990 Explorer's Guide dates the official renaming to 1986.

Meanwhile the project's scientists studied the Soviet maps that had beaten them to the planet. Between November 1985 and August 1986 Soviet scientists handed over tapes of digital Venera images, for planning only, and received Viking data of Mars and Phobos in return, according to Butrica.

Challenger was lost on 28 January 1986 while Magellan was on schedule and under budget, Butrica writes, with launch approved for April 1988 (May 1988 in JPL's 1990 Explorer's Guide). After the Shuttle/Centaur upper stage meant to launch it was cancelled in June, NASA settled in October on the shuttle, a two-stage Inertial Upper Stage and an April 1989 slot, which forced a lighter spacecraft and new load tests; Butrica puts the cost of the delay at about 150 million dollars. As loaded into Atlantis it weighed 3,453 kg by the Explorer's Guide, 3,445 kg by Siddiqi.

One radar did three jobs on each pass, according to NSSDCA: side-looking synthetic aperture imaging through the big dish, altimetry from a smaller antenna, and measurement of the microwave heat given off by the ground. Each pass imaged a strip about 20 to 25 km wide and 17,000 km long. The images are greyscale, so any colour has been added: NASA's captions describe the familiar orange globes as simulated colour.

A first countdown stopped 31 seconds before launch on 28 April 1989 over a faulty hydrogen pump and a leaking hydrogen line. On 4 May Atlantis lifted off at 18:46:59 UT (NASA's STS-30 page; Siddiqi rounds it to 18:47:00), and the crew released Magellan at 01:01 UT on 5 May. Butrica and NASA both describe it as the first planetary spacecraft launched by a space shuttle.

The route was long. The October 1989 window first set aside for Magellan went to Galileo, which needed a gravity boost from Venus, so Magellan looped more than one and a half times around the Sun over 15 months instead, according to the Explorer's Guide. It braked into orbit on 10 August 1990 behind Venus as seen from Earth, so the burn could only be confirmed about 30 minutes later. NSSDCA gives the first orbit as 294 by 8,540 km; Siddiqi gives 297 by 8,463 km, and neither explains the difference.

Within a week the mission nearly ended. On 16 August 1990, during a radar test, the spacecraft lost its internal heartbeat signal, protected itself and dropped out of contact for about 15 hours; a second silence of 17 hours followed on 21 August, after which controllers uploaded software to reset the craft in such events (Siddiqi; Butrica).

Mapping began on 15 September and ran in cycles of 243 days, the time Venus takes to turn once beneath an orbit that stays fixed in space. Cycle 1 covered 83.7 percent of the planet according to NSSDCA's cycle page and Siddiqi (84 percent in NSSDCA's catalogue record and on the PDS Magellan page), ending on 15 May 1991 (14 May on the cycle page). Cycles 2 and 3, looking west and then adding stereo views, completed the 98 percent by September 1992. Siddiqi puts the data return at 1,200 gigabits, against 900 gigabits from all NASA planetary missions combined at the time.

What the radar revealed was a volcanic world. NSSDCA's summary counts at least 85 percent of the surface as volcanic flows and the rest as highly deformed mountain belts, with no evidence of Earth-like plate tectonics. Gerald Schaber of the US Geological Survey compiled the project's inventory of 921 impact craters as of 30 July 1993: scattered at random, mostly pristine, and only 4 percent embayed by lava. His group read that as a planet-wide resurfacing about 300 million years ago followed by a steep decline in activity, while Roger Phillips and colleagues proposed a steady equilibrium resurfacing (Magellan Final Science Reports, 1993). NSSDCA gives a mean surface age of about 500 million years, and its mission page, last updated in 2005, lists sudden against gradual resurfacing as a major unanswered question. Later work on the same images has been read as evidence of eruptions during the mission itself, a reading that is disputed (see below); Volcanoes on Venus and the surface of Venus take it further.

With imaging complete, Cycle 4 (14 September 1992 to 23 May 1993) turned the dish toward Earth, and Doppler shifts in its radio signal traced the mass beneath mountains and basins (NSSDCA). From an elliptical orbit, though, the field blurred toward the poles, since resolution degrades quickly with altitude.

Aerobraking was the answer. The propellant aboard was at least an order of magnitude too small to round off the orbit with thrusters, Douglas Griffith, Saunders and Daniel Lyons wrote, so from late May 1993 the team dipped the low point into the upper atmosphere and let drag do the work on a spacecraft never designed for atmospheric flight. Over 70 days and 730 orbits the high point shrank from 8,467 km to 541 km. JPL called it a first-of-a-kind manoeuvre in 1994, and David Doody describes Magellan as the first spacecraft to use aerobraking to change the shape of its orbit significantly at another planet.

That qualifier matters. Japan's Hiten had already skimmed Earth's upper atmosphere on 19 March 1991, and Siddiqi's chronicle lists it as the first spacecraft to use aerobraking to reduce velocity; the Institute of Space and Astronautical Science describes the experiment as a success.

From the new orbit, Cycle 5 (3 August 1993 to 29 August 1994) mapped gravity at high resolution over about 95 percent of the planet. NSSDCA sums up the outcome: gravity on Venus follows its topography closely, a sign that the planet holds up its highlands differently from Earth, possibly through processes deep in the interior.

By August 1994 thermal stress had degraded the solar panels' cell connections, JPL said, and the project chose to spend the last weeks on experiments. In the windmill experiment, under way by 7 September, the panels were turned in opposite directions like windmill sails, and the torque needed to stop the spacecraft spinning gave a measure of the drag of the upper atmosphere. Betsy Beyer, the program manager at NASA Headquarters, called it "a race to the finish", since the power might fail before the planned final experiment.

The termination experiment began on 11 October 1994. JPL's status report of 12 October counts four thruster sequences, the project's weekly report the next day five orbit trim manoeuvres; together they took the low point to 139.7 km, a little above the 136 km planned in September. With power running low the team set up a last windmill run and expected two more orbits. Magellan managed three, and at 10:02 UT on 12 October, on orbit 15032, its signal was lost just as it was about to pass behind Venus as seen from Earth.

No one saw the end. The project's weekly report expected Magellan to land on Venus by 20:00 UT on 13 October, JPL's press texts said it would burn up, NSSDCA's cycle page allows that some parts may have reached the surface, and NSSDCA's catalogue says it presumably burned up on 13 or 14 October. Jonathan McDowell's catalogue of landings and impacts puts its fall at 18:00 UT on 12 October. NASA's current mission page and Siddiqi put the loss of contact a day later than JPL's reports of the time, after 10:05 UT on 13 October, with burn-up about ten hours later. None of these sources gives an entry or impact point. JPL's mission page calls it the first time an operating planetary spacecraft had been intentionally crashed; what we left at Venus gathers the other hardware.

The final bill, by NSSDCA's count, came to roughly 680 million dollars: 551 million to 29 October 1991 plus about 130 million to the end. That matches the headline price Butrica gives for VOIR in 1981, although the two figures come from different years and neither source states a price year.

Thirty years on, researchers are still mining the data, now online through the Planetary Data System. In March 2023 Robert Herrick of the University of Alaska Fairbanks and Scott Hensley of JPL reported in Science that a vent on Maat Mons, under 2.2 square km in February 1991, had doubled in size and appeared filled to the rim with lava eight months later. In May 2024, in Nature Astronomy, Davide Sulcanese, Giuseppe Mitri and Marco Mastrogiuseppe read radar brightness changes between cycles on the flank of Sif Mons and in western Niobe Planitia as new lava flows; Mitri told the Italian agency ANSA that the team now had certainty of activity. In April 2026 the same journal published a challenge by S. S. Bhiravarasu and colleagues, who argue the changes fit different viewing geometries, alongside a reply maintaining that geometry alone cannot produce the patterns. Neither side conceded in that April 2026 exchange, so the question stays open.

Magellan's gravity data have a second life too. In May 2025 Gael Cascioli of NASA's Goddard Space Flight Center and colleagues compared models of coronae with Magellan gravity and topography in Science Advances and, as INAF's news service reported, found signs of recent and perhaps continuing upwelling beneath 52 of 75 coronae. In February 2026 Leonardo Carrer, Elena Diana and Lorenzo Bruzzone of the University of Trento reported in Nature Communications an open conduit in the Nyx Mons region, which they interpret as a lava tube about 1 km across with a void at least 375 m high (a depth of at least 375 m in the university's Italian release). Radar confirms the conduit for at least 300 m; aligned openings suggest it may run for 45 km or more, a hypothesis that, Bruzzone said in the university's release, needs sharper radar data from missions such as ESA's EnVision and NASA's VERITAS.