In the next few weeks, radar astronomers hope to obtain the best images to date of 4179 Toutatis, the asteroid that makes its closest approach to Earth tomorrow. They hope these pictures will show details as small as 160m (525ft) across.
New techniques and modern equipment have allowed a revolution in the astronomical use of ground-based radar since the first tentative efforts in the late Fifties. Scientists at Goldstone in California and at Arecibo in Puerto Rico are now able to detect features as small as 70m across (250ft) on Mars, 120km (74 miles) on Mercury, and 1.5km (1 mile) on Venus.
Although Earth-based observations of Venus have now been surpassed by the incredibly detailed radar images from the Magellan orbiter, the same cannot be said of the other planets. Often hidden in the glare of the Sun, Mercury has always been a difficult world to explore and has long been recognised as one of the most inhospitable planets in the solar system. With a surface temperature of 430C, it would hardly seem to be the ideal place to find water ice. Yet this is just what American scientists announced at the end of last year.
One spacecraft, Mariner 10, has so far visited Mercury, but it was only able to send back pictures of the hemisphere it passed over. Until recently little was known about the planet's other side. In August last year a team led by Professor Duane Muhleman, of the California Institute of Technology, succeeded in obtaining the first detailed images of Mercury's 'hidden' hemisphere.
They used the 70-metre radio dish at Goldstone to transmit a 500kW signal over a period of eight hours. The faint echoes were then picked up by 27 linked radio telescopes in New Mexico. The resulting images were dominated by an elliptical feature 300km- 600km (186-370 miles) across, centred around Mercury's north pole.
Comparisons with similar studies of the Martian ice-caps and the frozen surfaces of Jupiter's satellites led to the startling conclusion that these enhanced echoes are caused by a highly fractured layer of ice.
Further studies this spring indicated a similar, though much smaller, feature close to the planet's south pole. According to Dr Martin Slade of the Jet Propulsion Laboratory (JPL), the location of the radar-bright patch corresponds to a 100km-wide (62-mile) crater named Chao Meng-Fuin, which can be seen in the Mariner 10 pictures.
Dr Slade is the first to agree that polar ice-caps are not very likely on a waterless, airless, Sun- blistered ball of rock such as Mercury. 'If the ice could be continually replicated, there would be no difficulty in explaining its presence, but it has to be an almost permanent feature. For this to be the case, there needs to be a temperature of around minus 160C.'
Such low temperatures possibly exist at Mercury's poles. The planet spins on its axis like an upright top, so the Sun never appears far above the polar horizon. If the poles are regions of rough topography, there may be a zone of permanent shadow.
But where did the ice come from? Volatile substances such as water could be driven from the rocks as they boil under the midday Sun. Such gases would then either escape into space or settle as 'snow' in the cold polar regions. Further supplies of water and carbon dioxide ice could be brought at irregular intervals by comets and asteroids which vaporise on impact with the planet.
Dr Slade explains how such ice might remain intact for a considerable period of time: 'Since radar can see through loose surface debris, it is possible that the ice is buried.' A blanket of material, perhaps 50cm (19in) thick, could protect the ice-cap from melting and from erosion by interstellar radiation.
The new radar maps also reveal huge impact basins on Mercury's hitherto unknown hemisphere. Two of these have diameters of about 800km (500 miles) and lie at the same latitude but on opposite sides of the planet's equator.
Not all discoveries are the result of bright radar echoes. In recent studies of Mars, some parts of the planet are invisible to radar. One huge blank on the maps, 2,000km (1,240 miles) across, displays no noticeable echo. Christened 'Stealth' by Professor Muhleman's team, the region seems to be covered in loosely consolidated ash, blown down slopes from nearby volcanoes.
The greatest challenge for radar astronomy is Titan, the giant moon of Saturn, whose surface is permanently covered by a dense orange smog. The vast distances involved result in a return echo of a mere 10-22 (a decimal point followed by 21 noughts and a one) watts being picked up by the New Mexico telescopes - very close to the limits of detection by existing equipment. Observations over the past three years have led Professor Muhleman to the conclusion that Titan does not copy our Moon by always keeping the same hemisphere towards its home planet, as previously believed.
Even more significant are variations in radar reflectivity, which suggest that Titan's surface consists of icy continents, perhaps coated in tars or hydrocarbons, with scattered lakes of ethane, a liquid rather like paraffin. Astronomers can also use radar to learn about relatively near, fast-moving objects. In New Zealand radar is used to track shooting stars, the tiny interstellar dust particles that burn up in the Earth's upper atmosphere.
Earth-approaching asteroids are the quarry of Dr Steven Ostro, of JPL. He has already had some spectacular success. Last year, his team announced that an asteroid, which had been discovered some five years before, called 1986 DA, was made of iron and nickel mixed with 10 parts per million of platinum and gold. In a body the size of 1986 DA, this would add up to 100,000 tons of precious metals.
Three years ago he obtained the most detailed radar images yet of an asteroid. This 3km-wide (1.8 miles) chunk of rock, now known as 4769 Castalia, turned out to comprise two irregular boulders, welded together like beads on a string and tumbling in a chaotic pirouette across the sky.
Dr Ostro is convinced that many more surprises are in store if sufficient funds can be provided. 'Radar promises a revolution in astronomy,' he says.